1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
12 than -2003 did; needs libidn2 in addition to libidn.
14 JH/02 The path option on a pipe transport is now expanded before use.
16 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
17 Patch provided by "Björn", documentation fix added too.
19 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
20 missing a wire-to-host endian conversion.
22 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
23 close after a BDAT command line could be taken as a following command,
24 giving a synch failure. Fix by only checking for synch immediately
25 before acknowledging the chunk.
27 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
28 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
29 macro. Patches provided by Josh Soref.
31 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
32 Previously we did not; the RFC seems ambiguous and VRFY is not listed
33 by IANA as a service extension. However, John Klensin suggests that we
36 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
37 the dkim code may be unix-mode line endings rather than smtp wire-format
38 CRLF, so prepend a CR to any bare LF.
40 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
41 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
43 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
44 the first were themselves being wrongly included in the feed into dkim
45 processing; with most chunk sizes in use this resulted in an incorrect
46 body hash calculated value.
48 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
49 DKIM signature block, for verification. Although advised against by
50 standards it is specifically not ruled illegal.
52 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
54 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
55 missing a body hash (the bh= tag).
57 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
58 It seems that HAProxy sends the Proxy Protocol information in clear and
59 only then does a TLS startup, so do the same.
61 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
62 TCP connections (such as for Spamd) unless the daemon successfully set
63 Fast Open mode on its listening sockets. This fixes breakage seen on
64 too-old kernels or those not configured for Fast Open, at the cost of
65 requiring both directions being enabled for TFO, and TFO never being used
66 by non-daemon-related Exim processes.
68 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
69 endings, at least on the first header line. Try to canonify any that get
70 past that check, despite the cost.
72 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
73 now limited to an arbitrary five deep, while parsing addresses with the
74 strip_excess_angle_brackets option enabled.
76 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
77 instead leave the unprompted TLS handshake in socket buffer for the
78 TLS library to consume.
80 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
82 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
84 JH/16 Drop variables when they go out of scope. Memory management drops a whole
85 region in one operation, for speed, and this leaves assigned pointers
86 dangling. Add checks run only under the testsuite which checks all
87 variables at a store-reset and panics on a dangling pointer; add code
88 explicitly nulling out all the variables discovered. Fixes one known
89 bug: a transport crash, where a dangling pointer for $sending_ip_address
90 originally assigned in a verify callout, is re-used.
92 PP/06 Drop '.' from @INC in various Perl scripts.
94 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
96 PP/08 Reduce a number of compilation warnings under clang; building with
97 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
98 should be warning-free.
104 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
105 supports it and a size is available (ie. the sending peer gave us one).
107 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
108 years of being deprecated). The replacements are the ACLs
109 acl_smtp_mime and acl_not_smtp_mime.
111 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
112 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
113 or even an in-clear connection were permitted. Now, if the host lookup
114 was dnssec and dane was requested then the host is only used if the
115 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
116 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
117 if one fails this test.
118 This means that a poorly-configured remote DNS will make it incommunicado;
119 but it protects against a DNS-interception attack on it.
121 JH/04 Bug 1810: make continued-use of an open smtp transport connection
122 non-noisy when a race steals the message being considered.
124 JH/05 If main configuration option tls_certificate is unset, generate a
125 self-signed certificate for inbound TLS connections.
127 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
128 in rewrites and routers.
130 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
131 and logged a warning sing 4.83; now they are a configuration file error.
133 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
134 (lacking @domain). Apply the same qualification processing as RCPT.
136 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
138 JH/10 Support ${sha256:} applied to a string (as well as the previous
141 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
142 a cutthrough deliver is pending, as we always want to make a connection.
143 This also avoids re-routing the message when later placing the cutthrough
144 connection after a verify cache hit.
145 Do not update it with the verify result either.
147 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
148 when routing results in more than one destination address.
150 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
151 signing (which inhibits the cutthrough capability). Previously only
152 the presence of an option was tested; now an expansion evaluating as
153 empty is permissible (obviously it should depend only on data available
154 when the cutthrough connection is made).
156 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
157 the relevant preceding SMTP command did not note the pipelining mode.
159 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
160 Previously they were not counted.
162 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
163 as one having no matching records. Previously we deferred the message
164 that needed the lookup.
166 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
167 distinguished as "(=".
169 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
170 for missing MX records. Previously it only worked for missing A records.
172 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
174 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
175 after the data-go-ahead and data-ack. Patch from Jason Betts.
177 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
178 even for a "none" policy. Patch from Tony Meyer.
180 JH/22 Fix continued use of a connection for further deliveries. If a port was
181 specified by a router, it must also match for the delivery to be
184 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
185 When one of the recipients of a message was unsuitable for the connection
186 (has no matching addresses), we lost track of needing to mark it
187 deferred. As a result mail would be lost.
189 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
191 JH/25 Decoding ACL controls is now done using a binary search; the source code
192 takes up less space and should be simpler to maintain. Merge the ACL
193 condition decode tables also, with similar effect.
195 JH/26 Fix problem with one_time used on a redirect router which returned the
196 parent address unchanged. A retry would see the parent address marked as
197 delivered, so not attempt the (identical) child. As a result mail would
200 JH/27 Fix a possible security hole, wherein a process operating with the Exim
201 UID can gain a root shell. Credit to http://www.halfdog.net/ for
202 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
205 JH/28 Enable {spool,log} filesystem space and inode checks as default.
206 Main config options check_{log,spool}_{inodes,space} are now
207 100 inodes, 10MB unless set otherwise in the configuration.
209 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
210 Previously it only applied to the main-section connection policy
213 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
215 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
216 by me. Added RFC7919 DH primes as an alternative.
218 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
219 are not in the system include path.
221 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
222 GnuTLS, when a session startup failed (eg because the client disconnected)
223 Exim did stdio operations after fclose. This was exposed by a recent
224 change which nulled out the file handle after the fclose.
226 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
227 signed directly by the cert-signing cert, rather than an intermediate
228 OCSP-signing cert. This is the model used by LetsEncrypt.
230 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
232 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
233 an incoming connection.
235 HS/02 Bug 1802: Do not half-close the connection after sending a request
238 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
239 fallback to "prime256v1".
241 JH/34 SECURITY: Use proper copy of DATA command in error message.
242 Could leak key material. Remotely exploitable. CVE-2016-9963.
248 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
249 and 3.4.4 - once the server is enabled to respond to an OCSP request
250 it does even when not requested, resulting in a stapling non-aware
251 client dropping the TLS connection.
253 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
254 support variable-length bit vectors. No functional change.
256 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
257 The I= interface field on outgoing lines is now after the H= remote
258 host field, same as incoming lines. There is a separate
259 outgoing_interface log selector which allows you to disable the
262 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
263 If not running log_selector +smtp_connection the mainlog would be held
264 open indefinitely after a "too many connections" event, including to a
265 deleted file after a log rotate. Leave the per net connection logging
266 leaving it open for efficiency as that will be quickly detected by the
267 check on the next write.
269 HS/01 Bug 1671: Fix post transport crash.
270 Processing the wait-<transport> messages could crash the delivery
271 process if the message IDs didn't exist for some reason. When
272 using 'split_spool_directory=yes' the construction of the spool
273 file name failed already, exposing the same netto behaviour.
275 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
276 mime_regex ACL conditions.
278 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
279 to DSN fail messages (bounces): remote IP, remote greeting, remote response
280 to HELO, local diagnostic string.
282 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
283 log line to debug. Even when configured with a tls authenticator many
284 client connections are expected to not authenticate in this way, so
285 an authenticate fail is not an error.
287 HS/02 Add the Exim version string to the process info. This way exiwhat
288 gives some more detail about the running daemon.
290 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
291 matter for fast-change records such as DNSBLs.
293 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
294 retry record, even if constant. There may be multiple transports with
295 different interface settings and the retry behaviour needs to be kept
298 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
300 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
302 JH/10 Bug 840: fix log_defer_output option of pipe transport
304 JH/11 Bug 830: use same host for all RCPTS of a message, even under
305 hosts_randomize. This matters a lot when combined with mua_wrapper.
307 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
308 ${quote_pgsql:<string>} operator.
310 JH/13 Bug 1708: avoid misaligned access in cached lookup.
312 JH/14 Change header file name for freeradius-client. Relevant if compiling
313 with Radius support; from the Gentoo tree and checked under Fedora.
315 JH/15 Bug 1712: Introduce $prdr_requested flag variable
317 JH/16 Bug 1714: Permit an empty string as expansion result for transport
318 option transport_filter, meaning no filtering.
320 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
322 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
323 defaults to "*" (all hosts). The variable is now available when not built
324 with TLS, default unset, mainly to enable keeping the testsuite sane.
325 If a server certificate is not supplied (via tls_certificate) an error is
326 logged, and clients will find TLS connections fail on startup. Presumably
327 they will retry in-clear.
328 Packagers of Exim are strongly encouraged to create a server certificate
329 at installation time.
331 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
332 with the $config_file variable.
334 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
335 in transport context, after the attempt, and per-recipient. The latter type
336 is per host attempted. The event data is the error message, and the errno
337 information encodes the lookup type (A vs. MX) used for the (first) host,
338 and the trailing two digits of the smtp 4xx response.
340 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
341 to write to mainlog (or rejectlog, paniclog) in the window between file
342 creation and permissions/ownership being changed. Particularly affects
343 installations where exicyclog is run as root, rather than exim user;
344 result is that the running daemon panics and dies.
346 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
348 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
349 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
350 "pri" and "weight". Note that the previous implicit priority given by the
351 list order is no longer honoured.
353 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
356 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
357 by defining SUPPORT_SOCKS.
359 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
360 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
361 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
362 variables are renamed to proxy_{local,external}_{address,port}.
364 JH/27 Move Internationalisation support from Experimental to mainline, enabled
365 for a build by defining SUPPORT_I18N
367 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
368 of the query string, and make ${quote_redis:} do that quoting.
370 JH/29 Move Events support from Experimental to mainline, enabled by default
371 and removable for a build by defining DISABLE_EVENT.
373 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
375 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
376 cached by the daemon.
378 JH/32 Move Redis support from Experimental to mainline, enabled for a build
379 by defining LOOKUP_REDIS. The libhiredis library is required.
381 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
382 keys are given for lookup.
384 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
385 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
386 only supported when built with TLS support. The PolarSSL SHA routines
387 are still used when the TLS library is too old for convenient support.
389 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
390 openssl_options), for security. OpenSSL forces this from version 1.1.0
391 server-side so match that on older versions.
393 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
394 allocation for $value could be released as the expansion processing
395 concluded, but leaving the global pointer active for it.
397 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
398 and to use the domains and local_parts ACL conditions.
400 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
401 incorrectly not doubled on cutthrough transmission, hence seen as a
402 body-termination at the receiving system - resulting in truncated mails.
403 Commonly the sender saw a TCP-level error, and retransmitted the message
404 via the normal store-and-forward channel. This could result in duplicates
405 received - but deduplicating mailstores were liable to retain only the
406 initial truncated version.
408 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
410 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
412 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
413 we're in there, support oversigning also; bug 1309.
415 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
417 HS/04 Add support for keep_environment and add_environment options.
419 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
420 either intentional arithmetic overflow during PRNG, or testing config-
423 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
424 delivery resulted in actual delivery. Cancel cutthrough before DATA
427 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
428 rejects a recipient: pass the reject to the originator.
430 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
431 Many were false-positives and ignorable, but it's worth fixing the
434 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
435 for the new environment-manipulation done at startup. Move the routines
436 from being local to tls.c to being global via the os.c file.
438 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
439 an extract embedded as result-arg for a map, the first arg for extract
440 is unavailable so we cannot tell if this is a numbered or keyed
441 extraction. Accept either.
447 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
450 JH/02 The smtp transport option "multi_domain" is now expanded.
452 JH/03 The smtp transport now requests PRDR by default, if the server offers
455 JH/04 Certificate name checking on server certificates, when exim is a client,
456 is now done by default. The transport option tls_verify_cert_hostnames
457 can be used to disable this per-host. The build option
458 EXPERIMENTAL_CERTNAMES is withdrawn.
460 JH/05 The value of the tls_verify_certificates smtp transport and main options
461 default to the word "system" to access the system default CA bundle.
462 For GnuTLS, only version 3.0.20 or later.
464 JH/06 Verification of the server certificate for a TLS connection is now tried
465 (but not required) by default. The verification status is now logged by
466 default, for both outbound TLS and client-certificate supplying inbound
469 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
472 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
473 Status Notification (bounce) messages are now MIME format per RFC 3464.
474 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
475 under the control of the dsn_advertise_hosts option, and routers may
476 have a dsn_lasthop option.
478 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
479 default, modifiable by a malware= option. The list separator for
480 the options can now be changed in the usual way. Bug 68.
482 JH/10 The smtp_receive_timeout main option is now expanded before use.
484 JH/11 The incoming_interface log option now also enables logging of the
485 local interface on delivery outgoing connections.
487 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
488 if the interface and destination host and port all match.
490 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
493 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
494 Patch from Andrew Lewis.
496 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
497 now supports optional time-restrictions, weighting, and priority
498 modifiers per server. Patch originally by <rommer@active.by>.
500 JH/16 The spamd_address main option now supports a mixed list of local
501 and remote servers. Remote servers can be IPv6 addresses, and
502 specify a port-range.
504 JH/17 Bug 68: The spamd_address main option now supports an optional
505 timeout value per server.
507 JH/18 Bug 1581: Router and transport options headers_add/remove can
508 now have the list separator specified.
510 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
513 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
516 JH/21 Support for the A6 type of dns record is withdrawn.
518 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
519 rather than the verbs used.
521 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
522 from 255 to 1024 chars.
524 JH/24 Verification callouts now attempt to use TLS by default.
526 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
527 are generic router options now. The defaults didn't change.
529 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
530 Original patch from Alexander Shikoff, worked over by JH.
532 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
533 files if ZCAT_COMMAND is not executable.
535 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
537 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
539 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
540 Normally benign, it bites when the pair was led to by a CNAME;
541 modern usage is to not canonicalize the domain to a CNAME target
542 (and we were inconsistent anyway for A-only vs AAAA+A).
544 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
546 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
547 when evaluating $sender_host_dnssec.
549 JH/31 Check the HELO verification lookup for DNSSEC, adding new
550 $sender_helo_dnssec variable.
552 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
554 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
556 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
558 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
559 documented as working, but never had. Support all but $spam_report.
561 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
562 added for tls authenticator.
564 HS/03 Add perl_taintmode main config option
570 TL/01 When running the test suite, the README says that variables such as
571 no_msglog_check are global and can be placed anywhere in a specific
572 test's script, however it was observed that placement needed to be near
573 the beginning for it to behave that way. Changed the runtest perl
574 script to read through the entire script once to detect and set these
575 variables, reset to the beginning of the script, and then run through
576 the script parsing/test process like normal.
578 TL/02 The BSD's have an arc4random API. One of the functions to induce
579 adding randomness was arc4random_stir(), but it has been removed in
580 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
581 function when detected.
583 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
584 cause callback expansion.
586 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
587 syntax errors in an expansion can be treated as a string instead of
588 logging or causing an error, due to the internal use of bool_lax
589 instead of bool when processing it.
591 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
592 server certificates when making smtp deliveries.
594 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
596 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
598 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
600 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
601 Merged patch from Sebastian Wiedenroth.
603 JH/05 Fix results-pipe from transport process. Several recipients, combined
604 with certificate use, exposed issues where response data items split
605 over buffer boundaries were not parsed properly. This eventually
606 resulted in duplicates being sent. This issue only became common enough
607 to notice due to the introduction of connection certificate information,
608 the item size being so much larger. Found and fixed by Wolfgang Breyha.
610 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
611 size buffer was used, resulting in syntax errors when an expansion
614 JH/07 Add support for directories of certificates when compiled with a GnuTLS
615 version 3.3.6 or later.
617 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
618 is EXPERIMENTAL_EVENT, the main-configuration and transport options
619 both become "event_action", the variables become $event_name, $event_data
620 and $event_defer_errno. There is a new variable $verify_mode, usable in
621 routers, transports and related events. The tls:cert event is now also
622 raised for inbound connections, if the main configuration event_action
625 TL/06 In test suite, disable OCSP for old versions of openssl which contained
626 early OCSP support, but no stapling (appears to be less than 1.0.0).
628 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
629 server certificate names available under the smtp transport option
630 "tls_verify_cert_hostname" now do not permit multi-component wildcard
633 JH/10 Time-related extraction expansions from certificates now use the main
634 option "timezone" setting for output formatting, and are consistent
635 between OpenSSL and GnuTLS compilations. Bug 1541.
637 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
638 encoded parameter in the incoming message. Bug 1558.
640 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
641 include certificate info, eximon was claiming there were spoolfile
644 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
646 JH/14 Log delivery-related information more consistently, using the sequence
647 "H=<name> [<ip>]" wherever possible.
649 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
650 are problematic for Debian distribution, omit them from the release
653 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
655 JH/16 Fix string representation of time values on 64bit time_t architectures.
658 JH/17 Fix a null-indirection in certextract expansions when a nondefault
659 output list separator was used.
664 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
665 checkers that were complaining about end of non-void function with no
668 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
669 This was a regression introduced in 4.83 by another bugfix.
671 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
673 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
674 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
680 TF/01 Correctly close the server side of TLS when forking for delivery.
682 When a message was received over SMTP with TLS, Exim failed to clear up
683 the incoming connection properly after forking off the child process to
684 deliver the message. In some situations the subsequent outgoing
685 delivery connection happened to have the same fd number as the incoming
686 connection previously had. Exim would try to use TLS and fail, logging
687 a "Bad file descriptor" error.
689 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
690 utilities have not been installed.
692 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
693 temporary space as the ACL may create new global variables.
695 TL/01 LDAP support uses per connection or global context settings, depending
696 upon the detected version of the libraries at build time.
698 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
699 to extract and use the src ip:port in logging and expansions as if it
700 were a direct connection from the outside internet. PPv2 support was
701 updated based on HAProxy spec change in May 2014.
703 JH/02 Add ${listextract {number}{list}{success}{fail}}.
705 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
706 Properly escape header and check for NULL return.
708 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
711 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
713 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
714 characters in header names, implemented as a verify condition.
715 Contributed by Michael Fischer v. Mollard.
717 TL/05 Rename SPF condition results err_perm and err_temp to standardized
718 results permerror and temperror. Previous values are deprecated but
719 still accepted. In a future release, err_perm and err_temp will be
720 completely removed, which will be a backward incompatibility if the
721 ACL tests for either of these two old results. Patch contributed by
722 user bes-internal on the mailing list.
724 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
726 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
727 selectors, in both main and reject logs.
729 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
732 JH/07 Add malware type "sock" for talking to simple daemon.
734 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
736 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
737 routers/transports under cutthrough routing.
739 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
740 numbers. Touch up "bool" conditional to keep the same definition.
742 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
744 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
746 JH/12 Expand items in router/transport headers_add or headers_remove lists
747 individually rather than the list as a whole. Bug 1452.
749 Required for reasonable handling of multiple headers_ options when
750 they may be empty; requires that headers_remove items with embedded
751 colons must have them doubled (or the list-separator changed).
753 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
754 view the policy declared in the DMARC record. Currently, $dmarc_status
755 is a combined value of both the record presence and the result of the
758 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
760 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
761 dnslookup router and the smtp transport (applying to the forward
764 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
765 of ldap servers used for a specific lookup. Patch provided by Heiko
768 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
769 New variable $lookup_dnssec_authenticated for observability.
771 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
772 Patch submitted by Lars Timman.
774 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
776 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
777 Requires trusted mode and valid format message id, aborts otherwise.
778 Patch contributed by Heiko Schlichting.
780 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
781 certextract with support for various fields. Bug 1358.
783 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
784 is requested by default, modifiable by smtp transport option
787 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
788 operate on certificate variables to give certificate fingerprints
789 Also new ${sha256:cert_variable}.
791 JH/23 The PRDR feature is moved from being Experimental into the mainline.
793 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
794 Christian Aistleitner.
796 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
798 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
799 file. Patch from Wolfgang Breyha.
801 JH/25 Expand the coverage of the delivery $host and $host_address to
802 client authenticators run in verify callout. Bug 1476.
804 JH/26 Port service names are now accepted for tls_on_connect_ports, to
805 align with daemon_smtp_ports. Bug 72.
807 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
808 support and error reporting did not work properly.
810 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
811 and is readable. Patch from Andrew Colin Kissa.
813 TL/14 Enhance documentation of ${run expansion and how it parses the
814 commandline after expansion, particularly in the case when an
815 unquoted variable expansion results in an empty value.
817 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
819 PP/02 Fix internal collision of T_APL on systems which support RFC3123
820 by renaming away from it. Addresses GH issue 15, reported by
823 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
825 TL/15 SECURITY: prevent double expansion in math comparison functions
826 (can expand unsanitized data). Not remotely exploitable.
833 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
835 PP/02 Make -n do something, by making it not do something.
836 When combined with -bP, the name of an option is not output.
838 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
841 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
842 $sender_host_name and config options to manage this, and basic check
845 PP/05 DSCP support for outbound connections and control modifier for inbound.
847 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
848 (Only plugin which currently uses this is kerberos4, which nobody should
849 be using, but we should make it available and other future plugins might
850 conceivably use it, even though it would break NAT; stuff *should* be
851 using channel bindings instead).
853 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
854 name; added for Sendmail compatibility; requires admin caller.
855 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
856 just ignore it); requires trusted caller.
857 Also parse but ignore: -Ac -Am -X<logfile>
860 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
862 TL/02 Add +smtp_confirmation as a default logging option.
864 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
865 Patch by Magnus Holmgren from 2007-02-20.
867 TL/04 Bugzilla 1281 - Spec typo.
868 Bugzilla 1283 - Spec typo.
869 Bugzilla 1290 - Spec grammar fixes.
871 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
873 TL/06 Add Experimental DMARC support using libopendmarc libraries.
875 TL/07 Fix an out of order global option causing a segfault. Reported to dev
876 mailing list by by Dmitry Isaikin.
878 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
880 JH/02 Support "G" suffix to numbers in ${if comparisons.
882 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
884 NM/01 Bugzilla 1197 - Spec typo
885 Bugzilla 1196 - Spec examples corrections
887 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
889 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
890 gnutls_enable_pkcs11, but renamed to more accurately indicate its
893 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
894 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
896 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
897 "acl {{name}{arg}...}", and optional args on acl condition
900 JH/05 Permit multiple router/transport headers_add/remove lines.
902 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
904 JH/07 Avoid using a waiting database for a single-message-only transport.
905 Performance patch from Paul Fisher. Bugzilla 1262.
907 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
910 JH/09 Add $headers_added variable, with content from use of ACL modifier
911 add_header (but not yet added to the message). Bugzilla 199.
913 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
914 Pulled from Bugzilla 817 by Wolfgang Breyha.
916 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
918 (nb: this is the same fix as in Exim 4.80.1)
920 JH/11 Add A= logging on delivery lines, and a client_set_id option on
923 JH/12 Add optional authenticated_sender logging to A= and a log_selector
926 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
928 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
929 advertise SMTP AUTH mechanism to us, instead of a generic
930 protocol violation error. Also, make Exim more robust to bad
931 data from the Dovecot auth socket.
933 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
935 When a queue runner is handling a message, Exim first routes the
936 recipient addresses, during which it prunes them based on the retry
937 hints database. After that it attempts to deliver the message to
938 any remaining recipients. It then updates the hints database using
941 So if a recipient address works intermittently, it can get repeatedly
942 deferred at routing time. The retry hints record remains fresh so the
943 address never reaches the final cutoff time.
945 This is a fairly common occurrence when a user is bumping up against
946 their storage quota. Exim had some logic in its local delivery code
947 to deal with this. However it did not apply to per-recipient defers
948 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
950 This change adds a proper retry rule check during routing so that the
951 final cutoff time is checked against the message's age. We only do
952 this check if there is an address retry record and there is not a
953 domain retry record; this implies that previous attempts to handle
954 the address had the retry_use_local_parts option turned on. We use
955 this as an approximation for the destination being like a local
956 delivery, as in LMTP.
958 I suspect this new check makes the old local delivery cutoff check
959 redundant, but I have not verified this so I left the code in place.
961 TF/02 Correct gecos expansion when From: is a prefix of the username.
963 Test 0254 submits a message to Exim with the header
967 When I ran the test suite under the user fanf2, Exim expanded
968 the header to contain my full name, whereas it should have added
969 a Resent-Sender: header. It erroneously treats any prefix of the
970 username as equal to the username.
972 This change corrects that bug.
974 GF/01 DCC debug and logging tidyup
975 Error conditions log to paniclog rather than rejectlog.
976 Debug lines prefixed by "DCC: " to remove any ambiguity.
978 TF/03 Avoid unnecessary rebuilds of lookup-related code.
980 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
981 Bug spotted by Jeremy Harris; was flawed since initial commit.
982 Would have resulted in OCSP responses post-SNI triggering an Exim
983 NULL dereference and crash.
985 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
987 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
988 Bug detection, analysis and fix by Samuel Thibault.
989 Bugzilla 1331, Debian bug #698092.
991 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
993 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
994 Server implementation by Todd Lyons, client by JH.
995 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
996 config variable "prdr_enable" controls whether the server
997 advertises the facility. If the client requests PRDR a new
998 acl_data_smtp_prdr ACL is called once for each recipient, after
999 the body content is received and before the acl_smtp_data ACL.
1000 The client is controlled by both of: a hosts_try_prdr option
1001 on the smtp transport, and the server advertisement.
1002 Default client logging of deliveries and rejections involving
1003 PRDR are flagged with the string "PRDR".
1005 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1006 fclose(). Diagnosis by Todd Lyons.
1008 PP/17 Update configure.default to handle IPv6 localhost better.
1009 Patch by Alain Williams (plus minor tweaks).
1012 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1013 This is now consistent with GnuTLS, and is now documented: the
1014 previous undocumented portable approach to treating the option as
1015 unset was to force an expansion failure. That still works, and
1016 an empty string is now equivalent.
1018 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1019 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1020 not performing validation itself.
1022 PP/20 Added force_command boolean option to pipe transport.
1023 Patch from Nick Koston, of cPanel Inc.
1025 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1028 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1030 PP/21 Fix eximon continuous updating with timestamped log-files.
1031 Broken in a format-string cleanup in 4.80, missed when I repaired the
1032 other false fix of the same issue.
1033 Report and fix from Heiko Schlichting.
1036 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1037 Report from Prashanth Katuri.
1039 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1040 It's SecureTransport, so affects any MacOS clients which use the
1041 system-integrated TLS libraries, including email clients.
1043 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1044 using a MIME ACL for non-SMTP local injection.
1045 Report and assistance in diagnosis by Warren Baker.
1047 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1049 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1051 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1052 last id that failed so it may be referenced in subsequent ACL's.
1054 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1057 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1058 ldap library initialization, allowing self-signed CA's to be
1059 used. Also properly sets require_cert option later in code by
1060 using NULL (global ldap config) instead of ldap handle (per
1061 session). Bug diagnosis and testing by alxgomz.
1063 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1064 the src/util/ subdirectory.
1066 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1067 renamed to Transport Post Delivery Action by Jeremy Harris, as
1070 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1071 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1072 redis_servers = needs to be configured which will be used by the redis
1073 lookup. Patch from Warren Baker, of The Packet Hub.
1075 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1077 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1078 hostname or reverse DNS when processing a host list. Used suggestions
1079 from multiple comments on this bug.
1081 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1083 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1084 Missed a few lines, added it to make the runtest require no keyboard
1087 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1088 contains upper case chars. Make router use caseful_local_part.
1090 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1091 support when GnuTLS has been built with p11-kit.
1097 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1099 This, or similar/improved, will also be change PP/11 of 4.82.
1105 PP/01 Handle short writes when writing local log-files.
1106 In practice, only affects FreeBSD (8 onwards).
1107 Bugzilla 1053, with thanks to Dmitry Isaikin.
1109 NM/01 Bugzilla 949 - Documentation tweak
1111 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1114 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1116 PP/02 Implemented gsasl authenticator.
1118 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1120 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1121 `pkg-config foo` for cflags/libs.
1123 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1124 with rest of GSASL and with heimdal_gssapi.
1126 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1127 `pkg-config foo` for cflags/libs for the TLS implementation.
1129 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1130 properties get this fed in as external SSF. A number of robustness
1131 and debugging improvements to the cyrus_sasl authenticator.
1133 PP/08 cyrus_sasl server now expands the server_realm option.
1135 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1136 Patch by Jeremy Harris.
1138 PP/10 Added dbmjz lookup type.
1140 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1142 PP/12 MAIL args handles TAB as well as SP, for better interop with
1143 non-compliant senders.
1144 Analysis and variant patch by Todd Lyons.
1146 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1147 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1148 Patch from Dirk Mueller <dmueller@suse.com>
1150 PP/13 tls_peerdn now print-escaped for spool files.
1151 Observed some $tls_peerdn in wild which contained \n, which resulted
1152 in spool file corruption.
1154 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1155 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1156 or write after TLS renegotiation, which otherwise led to messages
1159 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1160 as a tracking header (ie: a signed header comes before the signature).
1161 Patch from Wolfgang Breyha.
1163 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1164 comma-sep list; embedded commas doubled.
1166 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1168 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1170 Report and patch from Dmitry Banschikov.
1172 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1173 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1174 failures. We appear to now support TLS1.1+ with Exim.
1176 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1177 lets Exim select keys and certificates based upon TLS SNI from client.
1178 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1179 before an outbound SMTP session. New log_selector, +tls_sni.
1181 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1182 NULL dereference. Report and patch from Alun Jones.
1184 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1185 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1186 Not seeing resolver debug output on NetBSD, but suspect this is a
1187 resolver implementation change.
1189 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1190 Left warnings. Added "eximon gdb" invocation mode.
1192 PP/21 Defaulting "accept_8bitmime" to true, not false.
1194 PP/22 Added -bw for inetd wait mode support.
1196 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1197 locate the relevant includes and libraries. Made this the default.
1199 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1200 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1202 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1203 This may cause build issues on older platforms.
1205 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1206 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1207 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1208 Added SNI support via GnuTLS too.
1209 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1211 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1213 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1214 Applied second patch from Janne, implementing suggestion to default
1215 multiple-strings-in-record handling to match SPF spec.
1217 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1219 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1220 read-only, out of scope).
1221 Patch from Wolfgang Breyha, report from Stuart Northfield.
1223 PP/29 Fix three issues highlighted by clang analyser static analysis.
1224 Only crash-plausible issue would require the Cambridge-specific
1225 iplookup router and a misconfiguration.
1226 Report from Marcin Mirosław.
1228 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1230 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1231 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1232 As part of this, removing so much warning spew let me fix some minor
1233 real issues in debug logging.
1235 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1236 assignment on my part. Fixed.
1238 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1239 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1240 Janne Snabb (who went above and beyond: thank you).
1242 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1243 string otherwise requires a connection and a bunch more work and it's
1244 relatively easy to get wrong. Should also expose TLS library linkage
1247 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1248 64-bit ${eval} (JH/03).
1250 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1251 GNU libc to support some of the 64-bit stuff, should not lead to
1252 conflicts. Defined before os.h is pulled in, so if a given platform
1253 needs to override this, it can.
1255 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1256 protection layer was required, which is not implemented.
1257 Bugzilla 1254, patch from Wolfgang Breyha.
1259 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1260 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1261 tls_dhparam take prime identifiers. Also unbreak combination of
1262 OpenSSL+DH_params+TLSSNI.
1264 PP/39 Disable SSLv2 by default in OpenSSL support.
1270 PP/01 Solaris build fix for Oracle's LDAP libraries.
1271 Bugzilla 1109, patch from Stephen Usher.
1273 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1275 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1278 TF/02 Fix a couple more cases where we did not log the error message
1279 when unlink() failed. See also change 4.74-TF/03.
1281 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1282 lock up or crash if it happened to be inside a call to libc when it
1283 got a SIGUSR1 from exiwhat.
1285 The SIGUSR1 handler appends the current process status to the process
1286 log which is later printed by exiwhat. It used to use the general
1287 purpose logging code to do this, but several functions it calls are
1288 not safe for signals.
1290 The new output code in the SIGUSR1 handler is specific to the process
1291 log, and simple enough that it's easy to inspect for signal safety.
1292 Removing some special cases also simplifies the general logging code.
1293 Removing the spurious timestamps from the process log simplifies
1296 TF/04 Improved ratelimit ACL condition.
1298 The /noupdate option has been deprecated in favour of /readonly which
1299 has clearer semantics. The /leaky, /strict, and /readonly update modes
1300 are mutually exclusive. The update mode is no longer included in the
1301 database key; it just determines when the database is updated. (This
1302 means that when you upgrade Exim will forget old rate measurements.)
1304 Exim now checks that the per_* options are used with an update mode that
1305 makes sense for the current ACL. For example, when Exim is processing a
1306 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1307 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1308 must specify per_mail/readonly. If you omit the update mode it defaults to
1309 /leaky where that makes sense (as before) or /readonly where required.
1311 The /noupdate option is now undocumented but still supported for
1312 backwards compatibility. It is equivalent to /readonly except that in
1313 ACLs where /readonly is required you may specify /leaky/noupdate or
1314 /strict/noupdate which are treated the same as /readonly.
1316 A useful new feature is the /count= option. This is a generalization
1317 of the per_byte option, so that you can measure the throughput of other
1318 aggregate values. For example, the per_byte option is now equivalent
1319 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1321 The per_rcpt option has been generalized using the /count= mechanism
1322 (though it's more complicated than the per_byte equivalence). When it is
1323 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1324 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1325 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1326 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1327 Note that using per_rcpt with a non-readonly update mode in more than
1328 one ACL will cause the recipients to be double-counted. (The per_mail
1329 and per_byte options don't have this problem.)
1331 The handling of very low rates has changed slightly. If the computed rate
1332 is less than the event's count (usually one) then this event is the first
1333 after a long gap. In this case the rate is set to the same as this event's
1334 count, so that the first message of a spam run is counted properly.
1336 The major new feature is a mechanism for counting the rate of unique
1337 events. The new per_addr option counts the number of different
1338 recipients that someone has sent messages to in the last time period. It
1339 behaves like per_rcpt if all the recipient addresses are different, but
1340 duplicate recipient addresses do not increase the measured rate. Like
1341 the /count= option this is a general mechanism, so the per_addr option
1342 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1343 example, measure the rate that a client uses different sender addresses
1344 with the options per_mail/unique=$sender_address. There are further
1345 details in the main documentation.
1347 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1349 TF/06 Removed a few PCRE remnants.
1351 TF/07 Automatically extract Exim's version number from tags in the git
1352 repository when doing development or release builds.
1354 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1355 Bugzilla 879. Patch from Paul Fisher.
1357 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1358 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1361 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1363 PP/05 Variable $av_failed, true if the AV scanner deferred.
1364 Bugzilla 1078. Patch from John Horne.
1366 PP/06 Stop make process more reliably on build failure.
1367 Bugzilla 1087. Patch from Heiko Schlittermann.
1369 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1370 Bugzilla 1089. Patch from Heiko Schlittermann.
1372 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1373 Bugzilla 1131. Patch from Holger Weiß.
1375 PP/09 Handle IPv6 addresses with SPF.
1376 Bugzilla 860. Patch from Wolfgang Breyha.
1378 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1380 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1383 PP/11 match_* no longer expand right-hand-side by default.
1384 New compile-time build option, EXPAND_LISTMATCH_RHS.
1385 New expansion conditions, "inlist", "inlisti".
1387 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1389 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1391 PP/14 fix log_write() format string regression from TF/03.
1392 Bugzilla 1152. Patch from Dmitry Isaikin.
1398 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1400 PP/02 Harmonised TLS library version reporting; only show if debugging.
1401 Layout now matches that introduced for other libraries in 4.74 PP/03.
1403 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1405 PP/04 New "dns_use_edns0" global option.
1407 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1410 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1411 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1413 TK/01 Updated PolarSSL code to 0.14.2.
1414 Bugzilla 1097. Patch from Andreas Metzler.
1416 PP/07 Catch divide-by-zero in ${eval:...}.
1417 Fixes bugzilla 1102.
1419 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1422 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1423 format-string attack -- SECURITY: remote arbitrary code execution.
1425 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1426 time unintentionally subject to list matching rules, letting the header
1427 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1428 arbitrary string expansion). This allowed for information disclosure.
1430 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1431 INT_MIN/-1 -- value coerced to INT_MAX.
1437 NM/01 Workaround for PCRE version dependency in version reporting
1440 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1441 This fixes portability to compilers other than gcc, notably
1442 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1444 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1445 makefiles for portability to HP-UX and POSIX correctness.
1447 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1448 Also via indented variable definition in the Makefile.
1449 (Debugging by Oliver Heesakkers).
1451 PP/02 Restore caching of spamd results with expanded spamd_address.
1452 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1454 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1455 Improves build reliability. Fix from: Frank Elsner
1457 NM/02 Fix wide character breakage in the rfc2047 coding
1458 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1460 NM/03 Allow underscore in dnslist lookups
1461 Fixes bug 1026. Patch from Graeme Fowler
1463 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1464 Code patches from Adam Ciarcinski of NetBSD.
1466 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1469 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1470 is logged, to avoid truncation. Patch from John Horne.
1472 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1473 Patch from Jakob Hirsch.
1475 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1476 SQL string expansion failure details.
1477 Patch from Andrey Oktyabrski.
1479 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1480 Patch from Simon Arlott.
1482 PP/09 New lookups functionality failed to compile on old gcc which rejects
1483 extern declarations in function scope.
1484 Patch from Oliver Fleischmann
1486 PP/10 Use sig_atomic_t for flags set from signal handlers.
1487 Check getgroups() return and improve debugging.
1488 Fixed developed for diagnosis in bug 927 (which turned out to be
1491 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1492 Patch from Mark Zealey.
1494 PP/12 Bugzilla 1056: Improved spamd server selection.
1495 Patch from Mark Zealey.
1497 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1498 Based on patch from Heiko Schlittermann.
1500 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1501 Patch from Uwe Doering, sign-off by Michael Haardt.
1503 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1504 variable declaration deep within a block. Bug and patch from
1507 PP/15 lookups-Makefile IRIX compatibility coercion.
1509 PP/16 Make DISABLE_DKIM build knob functional.
1511 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1512 Patch by Simon Arlott
1514 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1515 variable argument macros. Our copy now differs from upstream.
1521 TF/01 Failure to get a lock on a hints database can have serious
1522 consequences so log it to the panic log.
1524 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1525 controlled using the smtp_confirmation log selector.
1527 TF/03 Include the error message when we fail to unlink a spool file.
1529 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1530 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1531 for maintaining out-of-tree patches for some time.
1533 PP/01 Bugzilla 139: Documentation and portability issues.
1534 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1535 Handle per-OS dynamic-module compilation flags.
1537 PP/02 Let /dev/null have normal permissions.
1538 The 4.73 fixes were a little too stringent and complained about the
1539 permissions on /dev/null. Exempt it from some checks.
1540 Reported by Andreas M. Kirchwitz.
1542 PP/03 Report version information for many libraries, including
1543 Exim version information for dynamically loaded libraries. Created
1544 version.h, now support a version extension string for distributors
1545 who patch heavily. Dynamic module ABI change.
1547 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1548 privilege escalation vulnerability whereby the Exim run-time user
1549 can cause root to append content of the attacker's choosing to
1552 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1555 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1556 If dropping privileges for untrusted macros, we disabled normal logging
1557 on the basis that it would fail; for the Exim run-time user, this is not
1558 the case, and it resulted in successful deliveries going unlogged.
1559 Fixed. Reported by Andreas Metzler.
1565 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1566 only prepend for the Resent-* case. Fixes regression introduced in
1567 Exim 4.70 by NM/22 for Bugzilla 607.
1569 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1570 increasing numbers of administrators be bitten by this.
1572 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1574 PP/03 Bugzilla 994: added openssl_options main configuration option.
1576 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1578 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1580 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1582 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1583 without a peer certificate, leading to a segfault because of an
1584 assumption that peers always have certificates. Be a little more
1585 paranoid. Problem reported by Martin Tscholak.
1587 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1588 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1589 NB: ClamAV planning to remove STREAM in "middle of 2010".
1590 CL also introduces -bmalware, various -d+acl logging additions and
1591 more caution in buffer sizes.
1593 PP/09 Implemented reverse_ip expansion operator.
1595 PP/10 Bugzilla 937: provide a "debug" ACL control.
1597 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1599 PP/12 Bugzilla 973: Implement --version.
1601 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1603 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1605 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1607 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1608 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1609 ignore trailing whitespace.
1611 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1613 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1616 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1617 Notification from Dr Andrew Aitchison.
1619 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1620 ExtendedDetectionInfo response format.
1621 Notification from John Horne.
1623 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1626 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1627 XSL and documented dependency on system catalogs, with examples of how
1630 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1633 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1634 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1635 configuration file which is writeable by the Exim user or group.
1637 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1638 of configuration files to cover files specified with the -C option if
1639 they are going to be used with root privileges, not just the default
1642 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1643 option (effectively making it always true).
1645 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1646 files to be used while preserving root privileges.
1648 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1649 that rogue child processes cannot use them.
1651 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1652 run-time user, instead of root.
1654 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1655 Exim run-time user without dropping privileges.
1657 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1658 result string, instead of calling string_vformat() twice with the same
1661 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1662 for other users. Others should always drop root privileges if they use
1663 -C on the command line, even for a whitelisted configure file.
1665 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1667 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1673 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1674 $data_path, and $header_path variables; fixed documentation bugs and
1677 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1678 exipick to access non-standard spools, including the "frozen" queue
1681 NM/01 Bugzilla 965: Support mysql stored procedures.
1682 Patch from Alain Williams
1684 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1686 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1687 Patch from Andreas Metzler
1689 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1690 Patch from Kirill Miazine
1692 NM/05 Bugzilla 671: Added umask to procmail example.
1694 JJ/03 installed exipick 20100323.0, fixing doc bug
1696 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1697 directory. Notification and patch from Dan Rosenberg.
1699 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1701 TK/02 Improve log output when DKIM signing operation fails.
1703 MH/01 Treat the transport option dkim_domain as a colon separated
1704 list, not as a single string, and sign the message with each element,
1705 omitting multiple occurences of the same signer.
1707 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1708 Bugzilla 985, 986. Patch by Simon Arlott
1710 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1711 Patch by Simon Arlott
1713 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1714 MBX locking. Notification from Dan Rosenberg.
1720 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1722 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1724 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1726 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1728 NM/04 Bugzilla 915: Flag broken perl installation during build.
1734 TK/01 Added patch by Johannes Berg that expands the main option
1735 "spamd_address" if it starts with a dollar sign.
1737 TK/02 Write list of recipients to X-Envelope-Sender header when building
1738 the mbox-format spool file for content scanning (suggested by Jakob
1741 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1742 (http://www.dcc-servers.net/) support via dccifd. Activated by
1743 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1745 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1746 by Mark Daniel Reidel <mr@df.eu>.
1748 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1749 When building exim an external PCRE library is now needed -
1750 PCRE is a system library on the majority of modern systems.
1751 See entry on PCRE_LIBS in EDITME file.
1753 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1754 conversation. Added nologin parameter to request.
1755 Patch contributed by Kirill Miazine.
1757 TF/01 Do not log submission mode rewrites if they do not change the address.
1759 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1761 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1762 log files in place. Contributed by Roberto Lima.
1764 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1766 TF/03 Bugzilla 615: When checking the local_parts router precondition
1767 after a local_part_suffix or local_part_prefix option, Exim now
1768 does not use the address's named list lookup cache, since this
1769 contains cached lookups for the whole local part.
1771 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1772 Robert Millan. Documentation is in experimental-spec.txt.
1774 TF/04 Bugzilla 668: Fix parallel build (make -j).
1776 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1778 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1779 Patch provided by Jan Srzednicki.
1781 TF/05 Leading white space used to be stripped from $spam_report which
1782 wrecked the formatting. Now it is preserved.
1784 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1785 that they are available at delivery time.
1787 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1789 TF/08 TLS error reporting now respects the incoming_interface and
1790 incoming_port log selectors.
1792 TF/09 Produce a more useful error message if an SMTP transport's hosts
1793 setting expands to an empty string.
1795 NM/06 Bugzilla 744: EXPN did not work under TLS.
1796 Patch provided by Phil Pennock.
1798 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1799 Patch provided by Richard Godbee.
1801 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1802 acl_smtp_notquit, added index entry.
1804 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1805 Patch provided by Eugene Bujak.
1807 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1808 accept(). Patch provided by Maxim Dounin.
1810 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1811 Patch provided by Phil Pennock.
1813 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1815 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1816 Patch provided by Brad "anomie" Jorsch.
1818 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1819 Patch provided by Dean Brooks.
1821 TK/05 Add native DKIM support (does not depend on external libraries).
1823 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1824 Patch provided by Graeme Fowler.
1826 NM/16 Bugzilla 851: Documentation example syntax fix.
1828 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1830 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1833 NM/19 Bugzilla 745: TLS version reporting.
1834 Patch provided by Phil Pennock.
1836 NM/20 Bugzilla 167: bool: condition support.
1837 Patch provided by Phil Pennock.
1839 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1840 clients. Patch provided by Phil Pennock.
1842 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1843 Patch provided by Brad "anomie" Jorsch.
1845 NM/23 Bugzilla 687: Fix misparses in eximstats.
1846 Patch provided by Heiko Schlittermann.
1848 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1849 Patch provided by Heiko Schlittermann.
1851 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1852 plus update to original patch.
1854 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1856 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1857 Patch provided by David Brownlee.
1859 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1861 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1863 NM/30 Bugzilla 888: TLS documentation bugfixes.
1865 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1867 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1868 Unlike the original bugzilla I have changed all shell scripts in src tree.
1870 NM/33 Bugzilla 898: Transport filter timeout fix.
1871 Patch by Todd Rinaldo.
1873 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1874 Patch by Serge Demonchaux.
1876 NM/35 Bugzilla 39: Base64 decode bug fixes.
1877 Patch by Jakob Hirsch.
1879 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1881 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1883 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1885 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1891 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1892 ALT-N's libdkim that I have put here:
1893 http://duncanthrax.net/exim-experimental/
1895 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1896 were called 'true' and it seems that C99 defines that as a reserved
1897 keyword to be used with 'bool' variable types. That means you could
1898 not include C99-style headers which use bools without triggering
1899 build errors in sieve.c.
1901 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1902 as mailq or other aliases. Changed the --help handling significantly
1903 to do whats expected. exim_usage() emits usage/help information.
1905 SC/01 Added the -bylocaldomain option to eximstats.
1907 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1909 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1911 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1917 PH/01 Another patch from the Sieve maintainer.
1919 PH/02 When an IPv6 address is converted to a string for single-key lookup
1920 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1921 dots are used instead of colons so that keys in lsearch files need not
1922 contain colons. This was done some time before quoting was made available
1923 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1924 (notated using the quote facility) so as to distinguish them from IPv4
1925 keys. This meant that lookups for IP addresses in host lists did not work
1926 for iplsearch lookups.
1928 This has been fixed by arranging for IPv6 addresses to be expressed with
1929 colons if the lookup type is iplsearch. This is not incompatible, because
1930 previously such lookups could never work.
1932 The situation is now rather anomalous, since one *can* have colons in
1933 ordinary lsearch keys. However, making the change in all cases is
1934 incompatible and would probably break a number of configurations.
1936 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1939 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1940 conversion specification without a maximum field width, thereby enabling
1941 a rogue spamd server to cause a buffer overflow. While nobody in their
1942 right mind would setup Exim to query an untrusted spamd server, an
1943 attacker that gains access to a server running spamd could potentially
1944 exploit this vulnerability to run arbitrary code as the Exim user.
1946 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1947 $primary_hostname instead of what libspf2 thinks the hosts name is.
1949 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1950 a directory entry by the name of the lookup key. Previously, if a
1951 symlink pointed to a non-existing file or a file in a directory that
1952 Exim lacked permissions to read, a lookup for a key matching that
1953 symlink would fail. Now it is enough that a matching directory entry
1954 exists, symlink or not. (Bugzilla 503.)
1956 PH/03 The body_linecount and body_zerocount variables are now exported in the
1959 PH/04 Added the $dnslist_matched variable.
1961 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1962 This means they are set thereafter only if the connection becomes
1965 PH/06 Added the client_condition to authenticators so that some can be skipped
1966 by clients under certain conditions.
1968 PH/07 The error message for a badly-placed control=no_multiline_responses left
1969 "_responses" off the end of the name.
1971 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1973 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1974 (without spaces) instead of just copying the configuration text.
1976 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1978 PH/11 Added $max_received_linelength.
1980 PH/12 Added +ignore_defer and +include_defer to host lists.
1982 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1983 way in which PCRE > 7.0 is built.
1985 PH/14 Implemented queue_only_load_latch.
1987 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1988 MAIL command. The effect was to mangle the value on 64-bit systems.
1990 PH/16 Another patch from the Sieve maintainer.
1992 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1994 PH/18 If a system quota error occurred while trying to create the file for
1995 a maildir delivery, the message "Mailbox is full" was not appended to the
1996 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1997 only to a quota excession during the actual writing of the file.
1999 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2000 characters?) which causes problems in log lines. The DN values are now
2001 passed through string_printing() before being added to log lines.
2003 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2004 and InterBase are left for another time.)
2006 PH/21 Added message_body_newlines option.
2008 PH/22 Guard against possible overflow in moan_check_errorcopy().
2010 PH/23 POSIX allows open() to be a macro; guard against that.
2012 PH/24 If the recipient of an error message contained an @ in the local part
2013 (suitably quoted, of course), incorrect values were put in $domain and
2014 $local_part during the evaluation of errors_copy.
2020 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2021 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2024 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2025 issue a MAIL command.
2027 PH/02 In an ACL statement such as
2029 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2031 if a client was not listed at all, or was listed with a value other than
2032 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2033 the condition was not true (as it should be), so access was not denied.
2034 The bug was that the ! inversion was incorrectly passed on to the second
2035 item. This has been fixed.
2037 PH/03 Added additional dnslists conditions == and =& which are different from
2038 = and & when the dns lookup returns more than one IP address.
2040 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2041 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2043 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2044 FSYNC, which compiles an option called disable_fsync that allows for
2045 bypassing fsync(). The documentation is heavily laced with warnings.
2047 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2049 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2050 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2051 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2052 including adding "make clean"; (3) Added -fPIC when compiling the test
2053 dynamically loaded module, to get rid of a warning.
2055 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2056 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2057 = 0s. The bug is otherwise harmless.
2059 PH/07 There was a bug in the dovecot authenticator such that the value of
2060 $auth1 could be overwritten, and so not correctly preserved, after a
2061 successful authentication. This usually meant that the value preserved by
2062 the server_setid option was incorrect.
2064 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2066 PH/09 Installed PCRE release 7.0.
2068 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2069 run for batched SMTP input. It is now run at the start of every message
2070 in the batch. While fixing this I discovered that the process information
2071 (output by running exiwhat) was not always getting set for -bs and -bS
2072 input. This is fixed, and it now also says "batched" for BSMTP.
2074 PH/11 Added control=no_pipelining.
2076 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2077 patch, slightly modified), and move the expansion of helo_data till after
2078 the connection is made in the smtp transport (so it can use these
2081 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2083 PH/14 Added log_selector = +pid.
2085 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2087 PH/16 Add ${if forany and ${if forall.
2089 PH/17 Added dsn_from option to vary the From: line in DSNs.
2091 PH/18 Flush SMTP output before performing a callout, unless control =
2092 no_callout_flush is set.
2094 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2095 was true (the default) a successful delivery failed to delete the retry
2096 item, thus causing premature timeout of the address. The bug is now
2099 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2101 PH/21 Long custom messages for fakedefer and fakereject are now split up
2102 into multiline responses in the same way that messages for "deny" and
2103 other ACL rejections are.
2105 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2106 with slight modification.
2108 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2109 draft, changing the syntax and factoring some duplicate code".
2111 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2112 for deliveries of the second and subsequent messages over the same SMTP
2115 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2116 ${reduce, with only minor "tidies".
2118 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2120 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2121 expansion side effects.
2123 PH/27 When a message times out after an over-quota error from an Exim-imposed
2124 quota, the bounce message says "mailbox is full". This message was not
2125 being given when it was a system quota that was exceeded. It now should
2128 MH/03 Made $recipients available in local_scan(). local_scan() already has
2129 better access to the recipient list through recipients_list[], but
2130 $recipients can be useful in postmaster-provided expansion strings.
2132 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2133 in the case of a MAIL command with additional options following the
2134 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2135 were accidentally chopped off.
2137 PH/29 SMTP synchronization checks are implemented when a command is read -
2138 there is a check that no more input is waiting when there shouldn't be
2139 any. However, for some commands, a delay in an ACL can mean that it is
2140 some time before the response is written. In this time, more input might
2141 arrive, invalidly. So now there are extra checks after an ACL has run for
2142 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2143 pipelining has not been advertised.
2145 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2147 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2148 returned in the SMTP response when smtp_return_error_details was false.
2149 This has been fixed.
2151 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2152 instead of the C I/O that was originally supplied, because problems were
2153 reported on Solaris.
2155 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2156 Exim which did not show up earlier: it was assuming that a call to
2157 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2158 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2159 was a macro that became an assignment, so it seemed to work. This has
2160 changed to a proper function call with a void return, hence the compile
2161 error. Exim's code has been fixed.
2163 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2166 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2168 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2170 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2171 because it turns out to be too big or there is a timeout) there is no
2172 "Completed" line in the log. When some messages of this type were
2173 selected by exigrep, they were listed as "not completed". Others were
2174 picked up by some special patterns. I have improved the selection
2175 criteria to be more general.
2177 PH/38 The host_find_failed option in the manualroute router can now be set
2178 to "ignore", to completely ignore a host whose IP address cannot be
2179 found. If all hosts are ignored, the behaviour is controlled by the new
2180 host_all_ignored option.
2182 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2183 homing or because of multiple MX records with /mx) generated more than
2184 one IP address, and the following item turned out to be the local host,
2185 all the secondary addresses of the first item were incorrectly removed
2186 from the list, along with the local host and any following hosts (which
2187 is what is supposed to happen).
2189 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2190 whoever called Exim into the -H file. In the case of the daemon it was
2191 behaving confusingly. When first started, it used values for whoever
2192 started the daemon, but after a SIGHUP it used the Exim user (because it
2193 calls itself on a restart). I have changed the code so that it now always
2196 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2197 message are rejected with the same error (e.g. no authentication or bad
2198 sender address), and a DATA command is nevertheless sent (as can happen
2199 with PIPELINING or a stupid MUA), the error message that was given to the
2200 RCPT commands is included in the rejection of the DATA command. This is
2201 intended to be helpful for MUAs that show only the final error to their
2204 PH/42 Another patch from the Sieve maintainer.
2206 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2207 Eximstats - Fixed some broken HTML links and added missing column headers
2209 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2210 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2212 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2214 PH/43 Yet another patch from the Sieve maintainer.
2216 PH/44 I found a way to check for a TCP/IP connection going away before sending
2217 the response to the final '.' that terminates a message, but only in the
2218 case where the client has not sent further data following the '.'
2219 (unfortunately, this is allowed). However, in many cases there won't be
2220 any further data because there won't be any more messages to send. A call
2221 to select() can be used: if it shows that the input is "ready", there is
2222 either input waiting, or the socket has been closed. An attempt to read
2223 the next input character can distinguish the two cases. Previously, Exim
2224 would have sent an OK response which the client would never have see.
2225 This could lead to message repetition. This fix should cure that, at
2226 least in a lot of common cases.
2228 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2229 advertised in response to EHLO.
2235 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2236 fixed by 4.65/MH/01 (is this a record?) are fixed:
2238 (i) An empty string was always treated as zero by the numeric comparison
2239 operators. This behaviour has been restored.
2241 (ii) It is documented that the numeric comparison operators always treat
2242 their arguments as decimal numbers. This was broken in that numbers
2243 starting with 0 were being interpreted as octal.
2245 While fixing these problems I realized that there was another issue that
2246 hadn't been noticed. Values of message_size_limit (both the global option
2247 and the transport option) were treated as octal if they started with 0.
2248 The documentation was vague. These values are now always treated as
2249 decimal, and I will make that clear in the documentation.
2255 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2256 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2259 MH/01 Don't check that the operands of numeric comparison operators are
2260 integers when their expansion is in "skipping" mode (fixes bug
2261 introduced by 4.64-PH/07).
2263 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2264 child addresses, Exim now panics and dies. Previously, because the count
2265 is held in a short int, deliveries were likely to be lost. As such a
2266 large number of recipients for a single message is ridiculous
2267 (performance will be very, very poor), I have chosen to impose a limit
2268 rather than extend the field.
2274 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2275 leftover -K file (the existence of which was triggered by #402).
2276 While we were at it, introduced process PID as part of the -K
2277 filename. This should rule out race conditions when creating
2280 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2281 processing considerably. Previous code took too long for large mails,
2282 triggering a timeout which in turn triggers #401.
2284 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2285 in the DK code in transports.c. sendfile() is not really portable,
2286 hence the _LINUX specificness.
2288 TF/01 In the add_headers option to the mail command in an Exim filter,
2289 there was a bug that Exim would claim a syntax error in any
2290 header after the first one which had an odd number of characters
2293 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2294 callout verification, Exim cached a "reject" for the entire domain. This
2295 is correct for most verifications, but it is not correct for a recipient
2296 verification with use_sender or use_postmaster set, because in that case
2297 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2298 case of MAIL FROM:<> rejection from other early rejections (e.g.
2299 rejection of HELO). When verifying a recipient using a non-null MAIL
2300 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2301 Whatever the result of the callout, the value of the domain cache is
2302 left unchanged (for any other kind of callout, getting as far as trying
2303 RCPT means that the domain itself is ok).
2305 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2308 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2309 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2310 session, the daemon ignores ECONNECT errors and logs others; it now
2311 ignores EPIPE as well.
2313 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2314 (quoted-printable decoding).
2316 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2317 later the small subsequent patch to fix an introduced bug.
2319 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2321 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2323 PH/08 An error is now given if message_size_limit is specified negative.
2325 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2326 to be given (somewhat) arbitrary names.
2328 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2331 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2332 miscellaneous code fixes
2334 PH/10 Added the log_reject_target ACL modifier to specify where to log
2337 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2338 hostname. This is wrong, because it relates to the incoming message (and
2339 probably the interface on which it is arriving) and not to the outgoing
2340 callout (which could be using a different interface). This has been
2341 changed to use the value of the helo_data option from the smtp transport
2342 instead - this is what is used when a message is actually being sent. If
2343 there is no remote transport (possible with a router that sets up host
2344 addresses), $smtp_active_hostname is used.
2346 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2347 tweaks were necessary in order to get it to work (see also 21 below):
2348 (a) The code assumed that strncpy() returns a negative number on buffer
2349 overflow, which isn't the case. Replaced with Exim's string_format()
2351 (b) There were several signed/unsigned issues. I just did the minimum
2352 hacking in of casts. There is scope for a larger refactoring.
2353 (c) The code used strcasecmp() which is not a standard C function.
2354 Replaced with Exim's strcmpic() function.
2355 (d) The code set only $1; it now sets $auth1 as well.
2356 (e) A simple test gave the error "authentication client didn't specify
2357 service in request". It would seem that Dovecot has changed its
2358 interface. Fortunately there's a specification; I followed it and
2359 changed what the client sends and it appears to be working now.
2361 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2364 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2365 address A is aliased to B and C, where B exists and C does not. Without
2366 -v the output is "A verified" because verification stops after a
2367 successful redirection if more than one address is generated. However,
2368 with -v the child addresses are also verified. Exim was outputting "A
2369 failed to verify" and then showing the successful verification for C,
2370 with its parentage. It now outputs "B failed to verify", showing B's
2371 parentage before showing the successful verification of C.
2373 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2374 look up a TXT record in a specific list after matching in a combined
2377 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2378 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2379 they consult the DNS. I had assumed they would set it the way they
2380 wanted; and indeed my experiments on Linux seem to show that in some
2381 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2382 To be on the safe side, however, I have now made the interface to
2383 host_find_byname() similar to host_find_bydns(), with an argument
2384 containing the DNS resolver options. The host_find_byname() function now
2385 sets these options at its start, just as host_find_bydns() does. The smtp
2386 transport options dns_qualify_single and dns_search_parents are passed to
2387 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2388 of host_find_byname() use the default settings of RES_DEFNAMES
2389 (qualify_single) but not RES_DNSRCH (search_parents).
2391 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2392 spool_read_header() do less string testing, by means of a preliminary
2393 switch on the second character of optional "-foo" lines. (This is
2394 overdue, caused by the large number of possibilities that now exist.
2395 Originally there were few.) While I was there, I also converted the
2396 str(n)cmp tests so they don't re-test the leading "-" and the first
2397 character, in the hope this might squeeze out yet more improvement.
2399 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2400 flag allowing group syntax was set by the header_syntax check but not
2401 turned off, possible causing trouble later; (2) The flag was not being
2402 set at all for the header_verify test, causing "group"-style headers to
2403 be rejected. I have now set it in this case, and also caused header_
2404 verify to ignore an empty address taken from a group. While doing this, I
2405 came across some other cases where the code for allowing group syntax
2406 while scanning a header line wasn't quite right (mostly, not resetting
2407 the flag correctly in the right place). These bugs could have caused
2408 trouble for malformed header lines. I hope it is now all correct.
2410 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2411 with the "reply" argument non-NULL. The code, however (which originally
2412 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2413 but it didn't always do it. This confused somebody who was copying the
2414 code for some other use. I have removed all the tests.
2416 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2417 feature that was used to support insecure browsers during the U.S. crypto
2418 embargo. It requires special client support, and Exim is probably the
2419 only MTA that supported it -- and would never use it because real RSA is
2420 always available. This code has been removed, because it had the bad
2421 effect of slowing Exim down by computing (never used) parameters for the
2422 RSA_EXPORT functionality.
2424 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2425 authenticator to fail if there's a tab character in the incoming data
2426 (there should never be unless someone is messing about, as it's supposed
2427 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2428 if the connection is using TLS or if the remote IP is the same as the
2429 local IP, and the "valid-client-cert option" if a client certificate has
2432 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2433 authenticators. This can be used for authorization after authentication
2434 succeeds. (In the case of plaintext, it servers for both authentication
2437 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2438 if any retry times were supplied.
2440 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2441 connection, where there is no client IP address to check. In this
2442 situation, the verify now always succeeds.
2444 PH/25 Applied John Jetmore's -Mset patch.
2446 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2448 PH/27 In a string expansion for a processed (not raw) header when multiple
2449 headers of the same name were present, leading whitespace was being
2450 removed from all of them, but trailing whitespace was being removed only
2451 from the last one. Now trailing whitespace is removed from each header
2452 before concatenation. Completely empty headers in a concatenation (as
2453 before) are ignored.
2455 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2456 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2458 PH/29 [Removed. This was a change that I later backed out, and forgot to
2459 correct the ChangeLog entry (that I had efficiently created) before
2460 committing the later change.]
2462 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2463 address errors (4xx response to RCPT) over the same connection as other
2464 messages routed to the same hosts. Such deliveries are always "forced",
2465 so retry times are not inspected. This resulted in far too many retries
2466 for the affected addresses. The effect occurred only when there were more
2467 hosts than the hosts_max_try setting in the smtp transport when it had
2468 the 4xx errors. Those hosts that it had tried were not added to the list
2469 of hosts for which the message was waiting, so if all were tried, there
2470 was no problem. Two fixes have been applied:
2472 (i) If there are any address or message errors in an SMTP delivery, none
2473 of the hosts (tried or untried) are now added to the list of hosts
2474 for which the message is waiting, so the message should not be a
2475 candidate for sending over the same connection that was used for a
2476 successful delivery of some other message. This seems entirely
2477 reasonable: after all the message is NOT "waiting for some host".
2478 This is so "obvious" that I'm not sure why it wasn't done
2479 previously. Hope I haven't missed anything, but it can't do any
2480 harm, as the worst effect is to miss an optimization.
2482 (ii) If, despite (i), such a delivery is accidentally attempted, the
2483 routing retry time is respected, so at least it doesn't keep
2484 hammering the server.
2486 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2487 in ${readsocket because some servers need this prod.
2489 PH/32 Added some extra debug output when updating a wait-xxx database.
2491 PH/33 The hint "could be header name not terminated by colon", which has been
2492 given for certain expansion errors for a long time, was not being given
2493 for the ${if def:h_colon_omitted{... case.
2495 PH/34 The spec says: "With one important exception, whenever a domain list is
2496 being scanned, $domain contains the subject domain." There was at least
2497 one case where this was not true.
2499 PH/35 The error "getsockname() failed: connection reset by peer" was being
2500 written to the panic log as well as the main log, but it isn't really
2501 panic-worthy as it just means the connection died rather early on. I have
2502 removed the panic log writing for the ECONNRESET error when getsockname()
2505 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2506 runs only) independently of the message's sender address. This meant
2507 that, if the 4xx error was in fact related to the sender, a different
2508 message to the same recipient with a different sender could confuse
2509 things. In particular, this can happen when sending to a greylisting
2510 server, but other circumstances could also provoke similar problems.
2511 I have changed the default so that the retry time for these errors is now
2512 based a combination of the sender and recipient addresses. This change
2513 can be overridden by setting address_retry_include_sender=false in the
2516 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2517 remote server are returned as part of bounce messages. This was not
2518 happening for LMTP over a pipe (the lmtp transport), but now it is the
2519 same for both kinds of LMTP.
2521 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2522 in header lines that were in fact CNAMEs. This is no longer the case.
2524 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2525 and queue runs started by the daemon processed all messages. This has
2526 been fixed so that -R and -S can now usefully be given with -q<time>.
2528 PH/40 Import PCRE release 6.7 (fixes some bugs).
2530 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2532 PH/42 Give an error if -q is specified more than once.
2534 PH/43 Renamed the variables $interface_address and $interface_port as
2535 $received_ip_address and $received_port, to make it clear that these
2536 values apply to message reception, and not to the outgoing interface when
2537 a message is delivered. (The old names remain recognized, of course.)
2539 PH/44 There was no timeout on the connect() call when using a Unix domain
2540 socket in the ${readsocket expansion. There now is.
2542 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2543 be meaningful with "accept".
2545 SC/01 Eximstats V1.43
2546 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2548 SC/02 Eximstats V1.44
2549 Use a glob alias rather than an array ref in the generated
2550 parser. This improves both readability and performance.
2552 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2553 Collect SpamAssassin and rejection statistics.
2554 Don't display local sender or destination tables unless
2555 there is data to show.
2556 Added average volumes into the top table text output.
2558 SC/04 Eximstats V1.46
2559 Collect data on the number of addresses (recipients)
2560 as well as the number of messages.
2562 SC/05 Eximstats V1.47
2563 Added 'Message too big' to the list of mail rejection
2564 reasons (thanks to Marco Gaiarin).
2566 SC/06 Eximstats V1.48
2567 Mainlog lines which have GMT offsets and are too short to
2568 have a flag are now skipped.
2570 SC/07 Eximstats V1.49 (Alain Williams)
2571 Added the -emptyok flag.
2573 SC/08 Eximstats V1.50
2574 Fixes for obtaining the IP address from reject messages.
2576 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2577 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2578 whitespace changes from 4.64-PH/27
2580 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2583 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2584 are found, allow negative numbers in numeric criteria)
2586 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2588 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2589 to match changes made in 4.64-PH/43
2591 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2593 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2594 "domain" is actually an IP address, and force a failure. This locks out
2595 those revolvers/nameservers that support "A-for-A" lookups, in
2596 contravention of the specifications.
2598 PH/48 When a host name was looked up from an IP address, and the subsequent
2599 forward lookup of the name timed out, the host name was left in
2600 $sender_host_name, contrary to the specification.
2602 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2603 restricted to single-key lookups, Exim was not diagnosing an error if
2604 * or *@ was used with a query-style lookup.
2606 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2608 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2609 long ago, but noone interested enough thought of it. Let's just say that
2610 the "1.1" means that there are some new functions that weren't there at
2611 some point in the past.
2613 PH/51 Error processing for expansion failure of helo_data from an smtp
2614 transport during callout processing was broken.
2616 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2617 tested/used via the -bh/-bhc/-bs options.
2619 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2620 bug, fixed in subsequent PCRE releases).
2622 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2623 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2625 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2631 SC/01 Use a glob alias rather than an array ref in eximstats generated
2632 parser. This improves both readability and performance.
2634 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2635 Don't display local sender or destination tables in eximstats unless
2636 there is data to show.
2637 Added average volumes into the eximstats top table text output.
2639 SC/03 Collect data on the number of addresses (recipients) as well
2640 as the number of messages in eximstats.
2642 TF/01 Correct an error in the documentation for the redirect router. Exim
2643 does (usually) call initgroups() when daemonizing.
2645 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2646 with consistent privilege compared to when running as a daemon.
2648 TF/03 Note in the spec that $authenticated_id is not set for local
2649 submissions from trusted users.
2651 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2652 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2654 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2655 by adding some example configuration directives to the default
2656 configuration file. A little bit of work is required to uncomment the
2657 directives and define how usernames and passwords are checked, but
2658 there is now a framework to start from.
2660 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2661 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2662 without this. I don't know how relevant this is to other LDAP libraries.
2664 PH/02 Add the verb name to the "unknown ACL verb" error.
2666 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2668 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2670 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2671 directory not expanded when it should be if an expanded home directory
2672 was set for the address (which is overridden by the transport).
2674 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2677 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2678 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2679 because it is too late at that time, and has no effect.
2681 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2682 security issue with \' (bugzilla #107). I could not use the
2683 PQescapeStringConn() function, because it needs a PGconn value as one of
2686 PH/08 When testing addresses using -bt, indicate those final addresses that
2687 are duplicates that would not cause an additional delivery. At least one
2688 person was confused, thinking that -bt output corresponded to deliveries.
2689 (Suppressing duplicates isn't a good idea as you lose the information
2690 about possibly different redirections that led to the duplicates.)
2692 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2693 systems where poll() doesn't work, in particular OS X.
2695 PH/10 Added more information to debugging output for retry time not reached.
2697 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2698 operations in malware.c.
2700 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2703 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2704 syslog_duplication set false, log lines that would normally be written
2705 both the the main log and to the reject log were not written to syslog at
2708 PH/14 In the default configuration, change the use of "message" in ACL warn
2709 statements to "add_header".
2711 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2712 not followed by a command (e.g. "seen endif").
2714 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2715 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2718 PH/17 Added extra conditions to the default value of delay_warning_condition
2722 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2723 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2724 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2727 The Auto-Submitted: and various List- headers are standardised, whereas I
2728 don't think Precedence: ever was.
2730 PH/18 Refactored debugging code in route_finduser() to show more information,
2731 in particular, the error code if getpwnam() issues one.
2733 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2734 This is apparently needed in addition to the PH/07 change above to avoid
2735 any possible encoding problems.
2737 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2738 but not after initializing Perl.
2740 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2741 output them only if debugging. By default they are written stderr,
2742 apparently, which is not desirable.
2744 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2747 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2750 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2752 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2753 authenticated or an ident call has been made. Suppress the default
2754 values for $authenticated_id and $authenticated_sender (but permit -oMai
2755 and -oMas) when testing with -bh.
2757 PH/24 Re-jigged the order of the tests in the default configuration so that the
2758 tests for valid domains and recipients precede the DNS black list and CSA
2759 tests, on the grounds that those ones are more expensive.
2761 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2762 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2763 This bug exists in every version of Exim that I still have, right back to
2766 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2767 However, an attempt to turn on case-sensitivity in a regex key by
2768 including (?-i) didn't work because the subject string was already
2769 lowercased, and the effects were non-intuitive. It turns out that a
2770 one-line patch can be used to allow (?-i) to work as expected.
2776 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2777 other effects) broke the use of negated acl sub-conditions.
2779 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2782 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2783 "Deny" causes Exim to reject the incoming connection with a 554 error.
2784 Unfortunately, if there is a major crisis, such as a disk failure,
2785 tcp-wrappers gives "deny", whereas what one would like would be some
2786 kind of temporary error. A kludge has been added to help with this.
2787 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2788 554 error is used if errno is still zero or contains ENOENT (which occurs
2789 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2792 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2794 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2795 errors. Otherwise a message that provokes a temporary error (when other
2796 messages do not) can cause a whole host to time out.
2798 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2799 addresses were routed directly to files or pipes from a redirect router.
2800 File deliveries just didn't batch; pipe deliveries might have suffered
2803 PH/06 A failure to get a lock for a hints database would erroneously always say
2804 "Failed to get write lock", even when it was really a read lock.
2806 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2807 of 0600. This has been changed to use the value of the lockfile_mode
2808 option (which defaults to 0600).
2810 PH/08 Applied small patch from the Sieve maintainer.
2812 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2813 folder from quota calculations, a direct delivery into this folder messed
2814 up the contents of the maildirsize file. This was because the regex was
2815 used only to exclude .Trash (or whatever) when the size of the mailbox
2816 was calculated. There was no check that a delivery was happening into an
2817 excluded directory. This bug has been fixed by ignoring all quota
2818 processing for deliveries into excluded directories.
2820 PH/10 Added the maildirfolder_create_regex option to appendfile.
2826 PH/01 The code for finding all the local interface addresses on a FreeBSD
2827 system running IPv6 was broken. This may well have applied to all BSD
2828 systems, as well as to others that have similar system calls. The broken
2829 code found IPv4 interfaces correctly, but gave incorrect values for the
2830 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2831 that it would not match correctly against @[] and not recognize the IPv6
2834 PH/02 The ipliteral router was not recognizing addresses of the form user@
2835 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2837 PH/03 Added disable_ipv6.
2839 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2840 decoded form, because it is most often used to construct To: headers
2841 lines in autoreplies, and the decoded form may well be syntactically
2842 invalid. However, $reply_address has leading white space removed, and all
2843 newlines turned into spaces so that the autoreply transport does not
2846 PH/05 If group was specified without a user on a router, and no group or user
2847 was specified on a transport, the group from the router was ignored.
2849 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2850 for visible compile-time settings that can be used to change these
2851 numbers, for those that want even more. Backwards compatibility with old
2852 spool files has been maintained. However, going back to a previous Exim
2853 release will lost any variables that are in spool files.
2855 PH/07 Two small changes when running in the test harness: increase delay when
2856 passing a TCP/IP connection to a new process, in case the original
2857 process has to generate a bounce, and remove special handling of
2858 127.0.0.2 (sic), which is no longer necessary.
2860 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2861 be the same on different OS.
2863 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2866 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2867 whether --show-vars was specified or not
2869 JJ/02 exipick: Added support for new ACL variable spool format introduced
2872 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2873 syntactically invalid From: or Reply-to: line, and a filter used this to
2874 generate an autoreply, and therefore failed to obtain an address for the
2875 autoreply, Exim could try to deliver to a non-existent relative file
2876 name, causing unrelated and misleading errors. What now happens is that
2877 it logs this as a hard delivery error, but does not attempt to create a
2880 PH/11 The exinext utility has a -C option for testing purposes, but although
2881 the given file was scanned by exinext itself; it wasn't being passed on
2882 when Exim was called.
2884 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2885 an end-of-file indication when reading a command response.
2887 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2888 compiled. In many other places in Exim, IPv6 addresses are always
2889 recognized, so I have changed this. It also means that IPv4 domain
2890 literals of the form [IPV4:n.n.n.n] are now always recognized.
2892 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2893 used if the router is not running as root, for example, when verifying at
2894 ACL time, or when using -bh. The debugging output from this situation was
2895 non-existent - all you got was a failure to exec. I have made two
2898 (a) Failures to set uid/gid, the current directory, or a process leader
2899 in a subprocess such as that created by queryprogram now generate
2900 suitable debugging output when -d is set.
2902 (b) The queryprogram router detects when it is not running as root,
2903 outputs suitable debugging information if -d is set, and then runs
2904 the subprocess without attempting to change uid/gid.
2906 PH/15 Minor change to Makefile for building test_host (undocumented testing
2909 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2910 additional section of a DNS packet that returns MX or SRV records.
2911 Instead, it always explicitly searches for A/AAAA records. This avoids
2912 major problems that occur when a DNS server includes only records of one
2913 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2914 fixed another bug: if SRV records were looked up and the corresponding
2915 address records were *not* found in the additional section, the port
2916 values from the SRV records were lost.
2918 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2919 using the correct key (the original address) when searching the retry
2920 rules in order to find which one to use for generating the retry hint.
2922 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2923 adding the default one. Similarly, if it contains a Reply-To: header, the
2924 errors_reply_to option, if set, is not used.
2926 PH/19 When calculating a retry time, Exim used to measure the "time since
2927 failure" by looking at the "first failed" field in the retry record. Now
2928 it does not use this if it is later than than the arrival time of the
2929 message. Instead it uses the arrival time. This makes for better
2930 behaviour in cases where some deliveries succeed, thus re-setting the
2931 "first failed" field. An example is a quota failure for a huge message
2932 when small messages continue to be delivered. Without this change, the
2933 "time since failure" will always be short, possible causing more frequent
2934 delivery attempts for the huge message than are intended.
2935 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2937 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2938 $1, $2, $3) because the numerical variables can be reset during some
2939 expansion items (e.g. "match"), thereby losing the authentication data.
2941 PH/21 Make -bV show the size of off_t variables so that the test suite can
2942 decide whether to run tests for quotas > 2G.
2944 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2945 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2946 filecount value is greater than 2G or if a quota value is greater than 2G
2947 on a system where the size of off_t is not greater than 4, a panic error
2950 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2951 never match. The debug and -bh output now contains an explicit error
2952 message indicating a malformed IPv4 address or mask.
2954 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2955 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2956 PH/23 above applies.
2958 PH/25 Do not write to syslog when running in the test harness. The only
2959 occasion when this arises is a failure to open the main or panic logs
2960 (for which there is an explicit test).
2962 PH/26 Added the /no_tell option to "control=freeze".
2964 PH/27 If a host name lookup failed very early in a connection, for example, if
2965 the IP address matched host_lookup and the reverse lookup yielded a name
2966 that did not have a forward lookup, an error message of the form "no IP
2967 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2968 could be logged. Now it outputs the IP address instead of "NULL".
2970 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2971 allows the sender and the authenticated sender to be set when
2972 submitting a message from within Exim. Since child_open_exim() is
2973 documented for local_scan(), the new function should be too.
2975 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2976 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2977 results in an empty string is now treated as unset.
2979 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2981 PH/31 Added sender_verify_fail logging option.
2983 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2984 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2985 tidied the source and removed it altogether.
2987 PH/33 When a queue run was abandoned because the load average was too high, a
2988 log line was always written; now it is written only if the queue_run log
2989 selector is set. In addition, the log line for abandonment now contains
2990 information about the queue run such as the pid. This is always present
2991 in "start" and "stop" lines but was omitted from the "abandon" line.
2993 PH/34 Omit spaces between a header name and the colon in the error message that
2994 is given when verify = headers_syntax fails (if there are lots of them,
2995 the message gets confusing).
2997 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2998 names, as there are now some PTR records that contain slashes. This check
2999 is only to protect against broken name servers that fall over on strange
3000 characters, so the fact that it applies to all lookups doesn't matter.
3002 PH/36 Now that the new test suite is complete, we can remove some of the
3003 special code in Exim that was needed for the old test suite. For example,
3004 sorting DNS records because real resolvers return them in an arbitrary
3005 order. The new test suite's fake resolver always returns records in the
3008 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3009 bounces) except when queue_only is set, to avoid logging races between
3010 the different processes.
3012 PH/38 Panic-die if .include specifies a non-absolute path.
3014 PH/39 A tweak to the "H" retry rule from its user.
3016 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3017 a label. They prevented compilation on older perls.
3019 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3020 a warning to be raised on newish perls.
3022 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3023 on queue. Changes to match documented behaviour of showing count of
3024 messages matching specified criteria.
3026 PH/40 Changed the default ident timeout from 30s to 5s.
3028 PH/41 Added support for the use of login_cap features, on those BSD systems
3029 that have them, for controlling the resources used by pipe deliveries.
3031 PH/42 The content-scanning code uses fopen() to create files in which to put
3032 message data. Previously it was not paying any attention to the mode of
3033 the files. Exim runs with umask(0) because the rest of the code creates
3034 files with open(), and sets the required mode explicitly. Thus, these
3035 files were ending up world-writeable. This was not a big issue, because,
3036 being within the spool directory, they were not world-accessible. I have
3037 created a function called modefopen, which takes an additional mode
3038 argument. It sets umask(777), creates the file, chmods it to the required
3039 mode, then resets the umask. All the relevant calls to fopen() in the
3040 content scanning code have been changed to use this function.
3042 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3043 to 24 hours. This avoids potential overflow problems when processing G
3044 and H retry rules. I suspect nobody ever tinkers with this value.
3046 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3048 PH/45 When the plaintext authenticator is running as a client, the server's
3049 challenges are checked to ensure they are valid base64 strings. By
3050 default, the authentication attempt is cancelled if an invalid string is
3051 received. Setting client_ignore_invalid_base64 true ignores these errors.
3052 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3053 they are received. Thus, the responses can be made to depend on the
3054 challenges. If an invalid string is ignored, an empty string is placed in
3057 PH/46 Messages that are created by the autoreply transport now contains a
3058 References: header, in accordance with RFCs 2822 and 3834.
3060 PH/47 Added authenticated_sender_force to the smtp transport.
3062 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3064 PH/49 Installed latest patch from the Sieve maintainer.
3066 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3067 was also set, the appendfile transport was unnecessarily scanning a
3068 directory of message files (e.g. for maildir delivery) to find the count
3069 of files (along with the size), even though it did not need this
3070 information. It now does the scan only if it needs to find either the
3071 size of the count of files.
3073 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3075 PH/52 Two bugs concerned with error handling when the smtp transport is
3078 (i) Exim was not creating retry information for temporary errors given
3079 for individual recipients after the DATA command when the smtp transport
3080 was used in LMTP mode. This meant that they could be retried too
3081 frequently, and not timed out correctly.
3083 (ii) Exim was setting the flag that allows error details to be returned
3084 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3085 recipients that were returned after the DATA command.
3087 PH/53 This is related to PH/52, but is more general: for any failing address,
3088 when detailed error information was permitted to be returned to the
3089 sender, but the error was temporary, then after the final timeout, only
3090 "retry timeout exceeded" was returned. Now it returns the full error as
3091 well as "retry timeout exceeded".
3093 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3094 do this, and (what is worse) MTAs that accept it.
3096 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3097 will now be deprecated.
3099 PH/56 New os.c-cygwin from the Cygwin maintainer.
3101 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3102 formats (previously only available in exim formats via -bpr, -bpru,
3103 and -bpra. Now also available in native and exiqgrep formats)
3105 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3106 with very large, slow to parse queues
3108 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3110 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3112 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3113 responses to authentication challenges, though it was showing the
3114 challenges; (ii) I've removed the CR characters from the debug output for
3117 PH/58 Allow for the insertion of a newline as well as a space when a string
3118 is turned into more than one encoded-word during RFC 2047 encoding. The
3119 Sieve code now uses this.
3121 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3122 data_4xx, lost_connection, tls_required.
3124 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3125 message was being sent as an SMTP response.
3127 PH/61 Add -l and -k options to exicyclog.
3129 PH/62 When verifying, if an address was redirected to one new address, so that
3130 verification continued, and the new address failed or deferred after
3131 having set something in $address_data, the value of $address_data was not
3132 passed back to the ACL. This was different to the case when no
3133 redirection occurred. The value is now passed back in both cases.
3135 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3136 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3137 use login_cap.h, so on its own it isn't the distinguishing feature. The
3138 new name refers directly to the setclassresources() function.
3140 PH/65 Added configuration files for NetBSD3.
3142 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3144 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3145 is preferred over IPv4.
3147 PH/68 The bounce_return_message and bounce_return_body options were not being
3148 honoured for bounces generated during the reception of non-SMTP messages.
3149 In particular, this applied to messages rejected by the ACL. This bug has
3150 been fixed. However, if bounce_return_message is true and bounce_return_
3151 body is false, the headers that are returned for a non-SMTP message
3152 include only those that have been read before the error was detected.
3153 (In the case of an ACL rejection, they have all been read.)
3155 PH/69 The HTML version of the specification is now built in a directory called
3156 spec_html instead of spec.html, because the latter looks like a path with
3157 a MIME-type, and this confuses some software.
3159 PH/70 Catch two compiler warnings in sieve.c.
3161 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3162 function verify_get_ident() calls ip_connect() to connect a socket, but
3163 if the "connect()" function timed out, ip_connect() used to close the
3164 socket. However, verify_get_ident() also closes the socket later, and in
3165 between Exim writes to the log, which may get opened at this point. When
3166 the socket was closed in ip_connect(), the log could get the same file
3167 descriptor number as the socket. This naturally causes chaos. The fix is
3168 not to close the socket in ip_connect(); the socket should be closed by
3169 the function that creates it. There was only one place in the code where
3170 this was missing, in the iplookup router, which I don't think anybody now
3171 uses, but I've fixed it anyway.
3173 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3174 well as to direct DNS lookups. Otherwise the handling of names in host
3175 lists is inconsistent and therefore confusing.
3181 PH/01 Two changes to the default runtime configuration:
3183 (1) Move the checks for relay_from_hosts and authenticated clients from
3184 after to before the (commented out) DNS black list checks.
3186 (2) Add control=submission to the relay_from_hosts and authenticated
3187 clients checks, on the grounds that messages accepted by these
3188 statements are most likely to be submissions.
3190 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3192 (1) Generate an error if the third argument for the ${prvs expansion is
3195 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3198 (3) Reset the variables that are obtained from the first argument of
3199 ${prvscheck and used in the second argument before leaving the code,
3200 because their memory is reclaimed, so using them afterwards may do
3203 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3204 one (it's much easier than Tom thought :-).
3206 (5) Because of (4), we can now allow for the use of $prvscheck_result
3207 inside the third argument.
3209 PH/03 For some reason, the default setting of PATH when running a command from
3210 a pipe transport was just "/usr/bin". I have changed it to
3213 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3214 anything to be listed in the output from -bV.
3216 PH/05 When a filter generated an autoreply, the entire To: header line was
3217 quoted in the delivery log line, like this:
3219 => >A.N.Other <ano@some.domain> <original@ddress> ...
3221 This has been changed so that it extracts the operative address. There
3222 may be more than one such address. If so, they are comma-separated, like
3225 => >ano@some.domain,ona@other.domain <original@ddress> ...
3227 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3228 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3229 not being looked up in the rDNS to get a host name, Exim was showing the
3230 IP address twice in Received: lines, even though the IP addresses were
3231 identical. For example:
3233 Received: from [1.2.3.4] (helo=[1.2.3.4])
3235 However, if the real host name was known, it was omitting the HELO data
3236 if it matched the actual IP address. This has been tidied up so that it
3237 doesn't show the same IP address twice.
3239 PH/07 When both +timestamp and +memory debugging was on, the value given by
3240 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3241 called by the debug printing, thereby overwriting the timestamp buffer.
3242 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3244 PH/08 When the original message was included in an autoreply transport, it
3245 always said "this is a copy of the message, including all the headers",
3246 even if body_only or headers_only was set. It now gives an appropriate
3249 PH/09 Applied a patch from the Sieve maintainer which:
3251 o fixes some comments
3252 o adds the (disabled) notify extension core
3253 o adds some debug output for the result of if/elsif tests
3254 o points to the current vacation draft in the documentation
3255 and documents the missing references header update
3259 o fixes a bug in processing the envelope test (when testing
3260 multiple envelope elements, the last element determined the
3263 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3264 Electronic Mail") by including:
3266 Auto-submitted: auto-generated
3268 in the messages that it generates (bounce messages and others, such as
3269 warnings). In the case of bounce messages for non-SMTP messages, there was
3270 also a typo: it was using "Auto_submitted" (underscore instead of
3271 hyphen). Since every message generated by Exim is necessarily in response
3272 to another message, thes have all been changed to:
3274 Auto-Submitted: auto-replied
3276 in accordance with these statements in the RFC:
3278 The auto-replied keyword:
3280 - SHOULD be used on messages sent in direct response to another
3281 message by an automatic process,
3283 - MUST NOT be used on manually-generated messages,
3285 - MAY be used on Delivery Status Notifications (DSNs) and Message
3286 Disposition Notifications (MDNs),
3288 - MUST NOT be used on messages generated by automatic or periodic
3289 processes, except for messages which are automatic responses to
3292 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3293 to the default Received: header definition.
3295 PH/12 Added log selector acl_warn_skipped (default on).
3297 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3298 variables because (a) they are in the wrong storage pool and (b) even if
3299 they were copied, it wouldn't work properly because of the caching.
3301 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3302 checking when decoding. Apparently there are clients that generate
3303 overlong encoded strings. Why am I not surprised?
3305 PH/15 If the first argument of "${if match_address" was not empty, but did not
3306 contain an "@" character, Exim crashed. Now it writes a panic log message
3307 and treats the condition as false.
3309 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3311 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3312 extension "envelope-auth". The code is finished and in agreement with
3313 other implementations, but there is no documentation so far and in fact,
3314 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3315 not changing the active code.
3317 Print executed "if" and "elsif" statements when debugging is used. This
3318 helps a great deal to understand what a filter does.
3320 Document more things not specified clearly in RFC3028. I had all this
3321 sorted out, when out of a sudden new issues came to my mind. Oops."
3323 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3326 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3327 canonical form (as documented). However, after a host name lookup from
3328 the IP address, check_host() was doing a simple string comparison with
3329 addresses acquired from the DNS when checking that the found name did
3330 have the original IP as one of its addresses. Since any found IPv6
3331 addresses are likely to be in abbreviated form, the comparison could
3332 fail. Luckily, there already exists a function for doing the comparison
3333 by converting both addresses to binary, so now that is used instead of
3334 the text comparison.
3336 PH/20 There was another similar case to PH/19, when a complete host name was
3337 given in a host list; looking up its IP address could give an abbreviated
3338 form, whereas the current host's name might or might not be abbreviated.
3339 The same fix has been applied.
3345 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3346 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3349 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3350 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3352 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3354 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3355 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3356 Also, for "Auto-Submitted", treat anything other than "no" as
3357 non-personal, in accordance with RFC 3834. (Previously it treated
3358 anything starting "auto-" as non-personal.)
3360 TF/01 The control=submission/name=... option had a problem with syntax
3361 errors if the name included a slash character. The /name= option
3362 now slurps the rest of the string, so it can include any characters
3363 but it must come last in the list of options (after /sender_retain
3366 PH/05 Some modifications to the interface to the fake nameserver for the new
3374 TK/01 Added the "success_on_redirect" address verification option. See
3375 NewStuff for rationale and an example.
3377 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3379 PH/02 Patch to exigrep to allow it to work on syslog lines.
3381 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3382 fread() to skip over the body file's header line, because in Cygwin the
3383 header line is locked and is inaccessible.
3385 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3386 co-exist for some time) to make it clear that it is the Exim ID that is
3387 referenced, not the Message-ID: header line.
3389 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3390 string_format() function, because snprintf() does not exist on all
3393 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3394 ${stat: expansion item.
3396 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3397 protocol synchronization error", to keep the pedants happy.
3399 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3400 well as for IRIX systems, when gcc is being used. See the host.c source
3403 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3405 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3408 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3409 transport and to the smtp transport in LMTP mode.
3411 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3413 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3414 run a filter in a subprocess. This could lead to confusion in subsequent
3415 lookups in the parent process. There should also be a search_tidyup() at
3416 the end of the subprocess.
3418 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3419 only if the host matched helo_try_verify_hosts, which caused the
3420 verification to occur when the EHLO/HELO command was issued. The ACL just
3421 tested the remembered result. Now, if a previous verification attempt has
3422 not happened, "verify = helo" does it there and then.
3424 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3426 TK/03 Fix log output including CR from clamd.
3428 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3429 exist provoked a memory error which could cause a segfault.
3431 PH/15 Installed PCRE 6.2
3433 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3435 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3436 of the problem. Specifically, suggested +O2 rather than +O1 for the
3439 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3441 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3442 record, the port was not correctly logged when the outgoing_port log
3443 selector was set (it logged the transort's default port).
3445 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3446 fallback_hosts, and "hosts" in the smtp transport.
3448 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3449 host errors such as "Connection refused".
3451 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3452 authentication with radiusclient 0.4.9:
3454 - Error returned from rc_read_config was caught wrongly
3455 - Username/password not passed on to radius server due to wrong length.
3457 The presumption is that some radiusclient API changes for 4.51/PH/17
3458 were not taken care of correctly. The code is still untested by me (my
3459 Linux distribution still has 0.3.2 of radiusclient), but it was
3460 contributed by a Radius user.
3462 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3463 expanding the "port" option of the smtp transport.
3465 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3466 while reading a MIME header. Thanks to Tom Hughes for a patch.
3468 PH/24 Include config.h inside local_scan.h so that configuration settings are
3471 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3472 that in an ACL for RCPT (for example), you can examine exactly what was
3475 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3476 commands, but it was not correctly comparing the address with the actual
3477 client host address. Thus, it would show the EHLO address in Received:
3478 header lines when this was not necessary.
3480 PH/27 Added the % operator to ${eval:}.
3482 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3483 it should be ignoring failures (because with -C, for example, it has lost
3484 privilege). It wasn't ignoring creation failures other than "already
3487 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3490 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3491 because an input line was too long, either on its own, or by virtue of
3492 too many continuations, the temporary file was not being removed, and the
3493 return code was incorrect.
3495 PH/31 Missing "BOOL" in function definition in filtertest.c.
3497 PH/32 Applied Sieve patches from the maintainer.
3499 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3501 PH/33 Added "verify = not_blind".
3503 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3504 Local/Makefile (with some defaults set). These are used in built scripts
3505 such as exicyclog, but they have never been used in the exim_install
3506 script (though there are many overriding facilities there). I have
3507 arranged that the exim_install script now takes note of these two
3510 PH/35 Installed configuration files for Dragonfly.
3512 PH/36 When a locally submitted message by a trusted user did not contain a
3513 From: header, and the sender address was obtained from -f or from an SMTP
3514 MAIL command, and the trusted user did not use -F to supply a sender
3515 name, $originator_name was incorrectly used when constructing a From:
3516 header. Furthermore, $originator_name was used for submission mode
3517 messages from external hosts without From: headers in a similar way,
3518 which is clearly wrong.
3520 PH/37 Added control=suppress_local_fixups.
3522 PH/38 When log_selector = +received_sender was set, and the addition of the
3523 sender made the log line's construction buffer exactly full, or one byte
3524 less than full, an overflow happened when the terminating "\n" was
3527 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3528 when the result of a list match is failure because a DNS lookup failed.
3530 PH/40 RM_COMMAND is now used in the building process.
3532 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3533 the "build-* directories that it finds.
3535 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3536 address in a domain literal was a prefix of an interface address.
3538 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3539 when verifying a sender address, unless rewrite_headers is false.
3541 PH/44 Wrote a long comment about why errors_to addresses are verified as
3542 recipients, not senders.
3544 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3545 the ratelimit ACL was added.
3547 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3549 PH/46 Added extra information about PostgreSQL errors to the error string.
3551 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3552 avoiding the need to install special zones in a real server. This is
3553 backwards compatible; if it can't find the fake resolver, it drops back.
3554 Thus, both old and new test suites can be run.
3556 TF/02 Added util/ratelimit.pl
3558 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3559 clock is set back in time.
3561 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3562 Candler <B.Candler@pobox.com>.
3564 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3565 OK for addresses that are the result of redirections.
3567 PH/48 A number of further additions for the benefit of the new test suite,
3568 including a fake gethostbyname() that interfaces to the fake DNS resolver
3571 TF/06 The fix for widen_domains has also been applied to qualify_single and
3572 search_parents which are the other dnslookup options that can cause
3575 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3578 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3580 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3581 service" if it encountered an MX record with an empty target hostname.
3582 The message is now "an MX or SRV record indicated no SMTP service".
3584 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3585 if the DNS of the sending site is misconfigured. This is quite a
3586 common situation. This change restores the behaviour of treating a
3587 helo verification defer as a failure.
3589 PH/51 If self=fail was set on a router, the bounce message did not include the
3590 actual error message.
3596 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3598 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3599 command itself ended in error, the underlying message about the transport
3600 filter timeout was being overwritten with the pipe command error. Now the
3601 underlying error message should be appended to the second error message.
3603 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3605 PH/02 Reduce the amount of output that "make" produces by default. Full output
3606 can still be requested.
3608 PH/03 The warning log line about a condition test deferring for a "warn" verb
3609 was being output only once per connection, rather than after each
3610 occurrence (because it was using the same function as for successful
3611 "warn" verbs). This seems wrong, so I have changed it.
3613 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3614 it should not have, which might have caused a crash in the right
3615 circumstances, but probably never did.
3617 PH/04 Installed a modified version of Tony Finch's patch to make submission
3618 mode fix the return path as well as the Sender: header line, and to
3619 add a /name= option so that you can make the user's friendly name appear
3622 TF/03 Added the control = fakedefer ACL modifier.
3624 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3625 Mark Lowes for thorough testing.
3627 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3629 TK/03 Merged latest SRS patch from Miles Wilton.
3631 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3632 with the definition in sysexits.h (which is #included earlier).
3633 Fortunately, Exim does not actually use EX_OK. The code used to try to
3634 preserve the sysexits.h value, by assuming that macro definitions were
3635 scanned for macro replacements. I have been disabused of this notion,
3636 so now the code just undefines EX_OK before #including unistd.h.
3638 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3639 in the smtp transport. When a block could not be written in a single
3640 write() function, the timeout was being re-applied to each part-write.
3641 This seems wrong - if the receiver was accepting one byte at a time it
3642 would take for ever. The timeout is now adjusted when this happens. It
3643 doesn't have to be particularly precise.
3645 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3646 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3648 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3649 without a domain if the check to <postmaster@domain> fails.
3651 SC/01 Eximstats: added -xls and the ability to specify output files
3652 (patch written by Frank Heydlauf).
3654 SC/02 Eximstats: use FileHandles for outputting results.
3656 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3658 SC/04 Eximstats: fixed display of large numbers with -nvr option
3660 SC/05 Eximstats: fixed merging of reports with empty tables.
3662 SC/06 Eximstats: added the -include_original_destination flag
3664 SC/07 Eximstats: removed tabs and trailing whitespace.
3666 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3668 TK/06 MBOX spool code: Add real "From " MBOX separator line
3669 so the .eml file is really in mbox format (even though
3670 most programs do not really care). Patch from Alex Miller.
3672 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3673 The latter is generated from $received_to and is only set if the
3674 message has one envelope recipient. SA can use these headers,
3675 obviously out-of-the-box. Patch from Alex Miller.
3677 PH/08 The ${def test on a variable was returning false if the variable's
3678 value was "0", contrary to what the specification has always said!
3679 The result should be true unless the variable is empty.
3681 PH/09 The syntax error of a character other than { following "${if
3682 def:variable_name" (after optional whitespace) was not being diagnosed.
3683 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3684 accidental colon was present, for example, could give incorrect results.
3686 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3687 result is used (not including appendfile, where other changes are about
3690 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3691 This involved changing a lot of size variables from int to off_t. It
3692 should work with maildirs and everything.
3694 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3695 spamd dying while we are connected to it.
3697 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3700 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3701 function for BDB 4.3.
3703 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3705 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3706 the sources for PCRE have been re-arranged and more files are now
3709 PH/15 The code I had for printing potentially long long variables in PH/11
3710 above was not the best (it lost precision). The length of off_t variables
3711 is now inspected at build time, and an appropriate printing format (%ld
3712 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3713 to be "long long int" or "long int". This is needed for the internal
3714 formatting function string_vformat().
3716 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3717 the configuration file to be ":syslog", then the script "guesses" where
3718 the logs files are, rather than using the compiled in default. In our
3719 case the guess is not the same as the compiled default, so the script
3720 suddenly stopped working when I started to use syslog. The patch checks
3721 to see if log_file_path is "". If so, it attempts to read it from exim
3722 with no configuration file to get the compiled in version, before it
3723 falls back to the previous guessing code."
3725 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3726 implementing BATV in an Exim configuration. See NewStuff for the gory
3729 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3730 Makefile that are specific to HP-UX.
3732 PH/18 If the "use_postmaster" option was set for a recipient callout together
3733 with the "random" option, the postmaster address was used as the MAIL
3734 FROM address for the random test, but not for the subsequent recipient
3735 test. It is now used for both.
3737 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3738 patch removes a few documentation additions to RFC 3028, because the
3739 latest draft now contains them. It adds the new en;ascii-case comparator
3740 and a new error check for 8bit text in MIME parts. Comparator and
3741 require names are now matched exactly. I enabled the subaddress
3742 extension, but it is not well tested yet (read: it works for me)."
3744 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3745 rework some of the code of TK/09 above to avoid the hardwired use of
3746 "%lld" and "long long". Replaced the call to snprintf() with a call to
3749 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3750 records point to non-existent hosts", "retry timeout exceeded", and
3751 "retry time not reached for any host after a long failure period".
3753 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3754 experimental DomainKeys support:
3756 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3757 (2) On an error such as an illegally used "control", the wrong name for
3758 the control was given.
3760 These problems did NOT occur unless DomainKeys support was compiled.
3762 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3764 PH/24 Added ${if match_ip condition.
3766 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3767 databases so that it will be absolutely obvious if a crash occurs in the
3768 DB library. This is a regular occurrence (often caused by mis-matched
3771 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3772 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3773 picked up on a user's system that detects such things. There doesn't seem
3774 to be a gcc warning option for this - only an attribute that has to be
3775 put on the function's prototype. It seems that in Fedora Core 4 they have
3776 set this on a number of new functions. No doubt there will be more in due
3779 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3780 specify a transport. However, if an address that was verified by such a
3781 router was the subject of a callout, Exim crashed because it tried to
3782 read the rcpt_include_affixes from the non-existent transport. Now it
3783 just assumes that the setting of that option is false. This bug was
3784 introduced by 4.51/PH/31.
3786 PH/28 Changed -d+all to exclude +memory, because that information is very
3787 rarely of interest, but it makes the output a lot bigger. People tend to
3788 do -d+all out of habit.
3790 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3791 code in os-type was giving problems when libc.so lives in lib64, like on
3794 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3795 aren't the modern standard, and it seems that some systems' include files
3796 don't always have them. Exim was already checking for some of the newer
3797 ones like T_AAAA, and defining it itself. I've added checks for all the
3798 record types that Exim uses.
3800 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3801 not automatically generating a new one, as it is supposed to. This
3802 prevented TLS from working. If the file did exist, but contained invalid
3803 data, a new version was generated, as expected. It was only the case of a
3804 non-existent file that was broken.
3806 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3807 with a change in libdomainkeys > 0.64.
3809 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3810 from DNS. If the selector record carries the flag, it now has
3811 precedence over the domain-wide flag.
3813 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3815 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3816 the use of a port name that isn't defined in /etc/services) occurred, the
3817 message was deferred as in a normal delivery, and thus remained on the
3818 spool, instead of being failed because of the mua_wrapper setting. This
3819 is now fixed, and I tidied up some of the mua_wrapper messages at the
3822 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3823 the messages in a hash of arrays rather than using individual hashes.
3824 This is a bit cleaner and results in dramatic memory savings, albeit
3825 at a slight CPU cost.
3827 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3828 as requested by Marc Sherman.
3830 SC/10 Eximstats: added histograms for user specified patterns as requested
3833 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3835 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3836 fopen() in the content-scanning modules that did not already have it.
3842 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3843 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3845 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3847 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3849 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3850 to test Sieve filters that use "vacation".
3852 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3853 that changes the way the GnuTLS parameters are stored in the cache file.
3854 The new format can be generated externally. For backward compatibility,
3855 if the data in the cache doesn't make sense, Exim assumes it has read an
3856 old-format file, and it generates new data and writes a new file. This
3857 means that you can't go back to an older release without removing the
3860 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3861 work if there are any delivery delays because "one_time" forces the
3862 parent to be marked "delivered", so its unseen clone is never tried
3863 again. For this reason, Exim now forbids the simultaneous setting of
3866 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3867 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3868 entry if you want to know the details. The fix, however, neglected to
3869 consider the case when local delivery batching is involved. The test for
3870 "previously delivered" was not happening when checking to see if an
3871 address could be batched with a previous (undelivered) one; under
3872 certain circumstances this could lead to multiple deliveries to the same
3875 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3876 in its include files, and this causes problems building Exim.
3878 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3879 header_syntax) but Exim was just ignoring anything given after a slash.
3880 In particular, this caused confusion with an attempt to use "verify =
3881 reverse_host_lookup/defer_ok". An error is now given when options are
3882 supplied for verify items that do not have them. (Maybe reverse_host_
3883 lookup should have a defer_ok option, but that's a different point.)
3885 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3886 defined by RFC 821) to 2048, because there were problems with some AUTH
3887 commands, and RFC 1869 says the size should be increased for extended
3888 SMTP commands that take arguments.
3890 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3893 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3894 "unknown" error; now it says that the functionality isn't in the binary.
3896 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3897 an address' error message when a string expansion fails (syntax or
3898 whatever). Otherwise the password may appear in the log. Following change
3899 PH/42 below, there is no longer a chance of it appearing in a bounce
3902 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3904 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3905 of its addresses was ever tried. (Bugzilla bug #2.)
3907 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3908 the result incorrectly in the debug output. (It correctly added a newline
3909 to what was transported.)
3911 TF/01 Added $received_time.
3913 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3914 commented out examples of how to interface to a virus scanner and to
3915 SpamAssassin. Also added commented examples of av_scanner and
3916 spamd_address settings.
3918 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3919 and controls are allowed in which ACLs. There were a couple of minor
3920 errors. Some of the entries in the conditions table (which is a table of
3921 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3922 negation of where the condition IS allowed.
3924 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3926 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3927 header file does not have a version number, so I've had to invent a new
3928 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3929 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3930 radiusclient), but it was contributed by a Radius user.
3932 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3933 files or directories, for OpenSSL.
3935 PH/19 When an Exim process that is running as root has to create an Exim log
3936 file, it does so in a subprocess that runs as exim:exim so as to get the
3937 ownership right at creation (otherwise, other Exim processes might see
3938 the file with the wrong ownership). There was no test for failure of this
3939 fork() call, which would lead to the process getting stuck as it waited
3940 for a non-existent subprocess. Forks do occasionally fail when resources
3941 run out. I reviewed all the other calls to fork(); they all seem to check
3944 PH/20 When checking for unexpected SMTP input at connect time (before writing
3945 the banner), Exim was not dealing correctly with a non-positive return
3946 from the read() function. If the client had disconnected by this time,
3947 the result was a log entry for a synchronization error with an empty
3948 string after "input=" when read() returned zero. If read() returned -1
3949 (an event I could not check), uninitialized data bytes were printed.
3950 There were reports of junk text (parts of files, etc) appearing after
3953 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3955 PH/22 Added support for macro redefinition, and (re)definition in between
3956 driver and ACL definitions.
3958 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3959 forgetting to use the resulting value; it was using the unexpanded value.
3961 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3962 hadn't been configured. The fix is from Juergen Kreileder, who
3963 understands it better than I do:
3965 "Here's what I see happening with three configured cyrus_sasl
3966 authenticators configured (plain, login, cram-md5):
3968 On startup auth_cyrus_sasl_init() gets called for each of these.
3969 This means three calls to sasl_listmech() without a specified mech_list.
3970 => SASL tests which mechs of all available mechs actually work
3971 => three warnings about OTP not working
3972 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3974 With the patch, sasl_listmech() also gets called three times. But now
3975 SASL's mech_list option is set to the server_mech specified in the the
3976 authenticator. Or in other words, the answer from sasl_listmech()
3977 gets limited to just the mech you're testing for (which is different
3979 => the return list contains just 'plain' or 'login', 'cram-md5' or
3980 nothing depending on the value of ob->server_mech.
3982 I've just tested the patch: Authentication still works fine,
3983 unavailable mechs specified in the exim configuration are still
3984 caught, and the auth.log warnings about OTP are gone."
3986 PH/25 When debugging is enabled, the contents of the command line are added
3987 to the debugging output, even when log_selector=+arguments is not
3990 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3991 answer is "GNU", and only if the return is "GNU/something" is the answer
3994 PH/27 $acl_verify_message is now set immediately after the failure of a
3995 verification in an ACL, and so is available in subsequent modifiers. In
3996 particular, the message can be preserved by coding like this:
3998 warn !verify = sender
3999 set acl_m0 = $acl_verify_message
4001 Previously, $acl_verify_message was set only while expanding "message"
4002 and "log_message" when a very denied access.
4004 PH/28 Modified OS/os.c-Linux with
4006 -#ifndef OS_LOAD_AVERAGE
4007 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4009 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4010 nomenclature these days.)
4012 PH/29 Installed patch from the Sieve maintainer that adds the options
4013 sieve_useraddress and sieve_subaddress to the redirect router.
4015 PH/30 In these circumstances:
4016 . Two addresses routed to the same list of hosts;
4017 . First host does not offer TLS;
4018 . First host accepts first address;
4019 . First host gives temporary error to second address;
4020 . Second host offers TLS and a TLS session is established;
4021 . Second host accepts second address.
4022 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4023 and peerdn, if requested) that were in fact used only for the second
4026 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4027 attention to any local part prefix or suffix that was matched by the
4028 router that accepted the address. It now behaves in the same way as it
4029 does for delivery: the affixes are removed from the local part unless
4030 rcpt_include_affixes is set on the transport.
4032 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4033 timeout during the DATA phase of an incoming message.
4035 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4036 applied a patch sanctioned by the Sieve maintainer.
4038 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4039 the uid or gid is negative. A case of a negative gid caused this to be
4040 noticed. The fix allows for either to be negative.
4042 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4043 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4046 PH/36 The change PH/12 above was broken. Fixed it.
4048 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4049 the grounds that routing the same address twice would always produce the
4050 same answer. This might have been true once, but it is certainly no
4051 longer true now. Routing a child address may depend on the previous
4052 routing that produced that child. Some complicated redirection strategies
4053 went wrong when messages had multiple recipients, and made Exim's
4054 behaviour dependent on the order in which the addresses were given.
4056 I have moved the duplicate checking until after the routing is complete.
4057 Exim scans the addresses that are assigned to local and remote
4058 transports, and removes any duplicates. This means that more work will be
4059 done, as duplicates will always all be routed, but duplicates are
4060 presumably rare, so I don't expect this is of any significance.
4062 For deliveries to pipes, files, and autoreplies, the duplicate checking
4063 still happens during the routing process, since they are not going to be
4066 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4067 It corrects a timeout issue with spamd. This is Ian's comment: "The
4068 background is that sometimes spamd either never reads data from a
4069 connection it has accepted, or it never writes response data. The exiscan
4070 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4071 blindly assumes that writes won't block so it may never time out."
4073 PH/39 Allow G after quota size as well as K and M.
4075 PH/40 The value set for $authenticated_id in an authenticator may not contain
4076 binary zeroes or newlines because the value is written to log lines and
4077 to spool files. There was no check on this. Now the value is run through
4078 the string_printing() function so that such characters are converted to
4079 printable escape sequences.
4081 PH/41 $message_linecount is a new variable that contains the total number of
4082 lines in the message. Compare $body_linecount, which is the count for the
4085 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4086 bounce and delay warning messages, except in certain special cases, which
4089 (a) An SMTP error message from a remote host;
4090 (b) A message specified in a :fail: redirection;
4091 (c) A message specified in a "fail" command in a system filter;
4092 (d) A message specified in a FAIL return from the queryprogram router;
4093 (e) A message specified by the cannot_route_message router option.
4095 In these cases only, Exim does include the error details in bounce and
4096 warning messages. There are also a few cases where bland messages such
4097 as "unrouteable address" or "local delivery error" are given.
4099 PH/43 $value is now also set for the "else" part of a ${run expansion.
4101 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4102 being worked on, but at least Exim now implements the latest version to
4105 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4106 process to complete was treated as a delivery failure, a timeout while
4107 writing the message to the pipe was logged, but erroneously treated as a
4108 successful delivery. Such timeouts include transport filter timeouts. For
4109 consistency with the overall process timeout, these timeouts are now
4110 treated as errors, giving rise to delivery failures by default. However,
4111 there is now a new Boolean option for the pipe transport called
4112 timeout_defer, which, if set TRUE, converts the failures into defers for
4113 both kinds of timeout. A transport filter timeout is now identified in
4116 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4117 systems where "make" and "gmake" are different, calling "gmake" at top
4118 level broke things. I've arranged for the value of $(MAKE) to be passed
4119 from the Makefile to this script so that it can call the same version of
4123 A note about Exim versions 4.44 and 4.50
4124 ----------------------------------------
4126 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4127 changes of various kinds. As a consequence, a big documentation update was
4128 needed. This delayed the release for rather longer than seemed good, especially
4129 in the light of a couple of (minor) security issues. Therefore, the changes
4130 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4131 release. So 4.44 and 4.50 are in effect two different branches that both start
4134 I have left the 4.50 change log unchanged; it contains all the changes since
4135 4.43. The change log for 4.44 is below; many of its items are identical to
4136 those for 4.50. This seems to be the most sensible way to preserve the
4137 historical information.
4143 1. Minor wording change to the doc/README.SIEVE file.
4145 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4146 computation of the current number of files was incorrect.
4148 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4149 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4152 4. Give more explanation in the error message when the command for a transport
4153 filter fails to execute.
4155 5. There are several places where Exim runs a non-Exim command in a
4156 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4157 was being done only for the command run by the queryprogram router. It is
4158 now done for all such subprocesses. The other cases are: ${run, transport
4159 filters, and the commands run by the lmtp and pipe transports.
4161 6. Added CONFIGURE_GROUP build-time option.
4163 7. Some older OS have a limit of 256 on the maximum number of file
4164 descriptors. Exim was using setrlimit() to set 1000 as a large value
4165 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4166 systems. I've change it so that if it can't get 1000, it tries for 256.
4168 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4169 was an oversight, and furthermore, ever since the addition of extra
4170 controls (e.g. 4.43/32), the checks on when to allow different forms of
4171 "control" were broken. There should now be diagnostics for all cases when a
4172 control that does not make sense is encountered.
4174 9. Added the /retain_sender option to "control=submission".
4176 10. $recipients is now available in the predata ACL (oversight).
4178 11. Tidy the search cache before the fork to do a delivery from a message
4179 received from the command line. Otherwise the child will trigger a lookup
4180 failure and thereby defer the delivery if it tries to use (for example) a
4181 cached ldap connection that the parent has called unbind on.
4183 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4184 of $address_data from the recipient verification was clobbered by the
4185 sender verification.
4187 13. The value of address_data from a sender verification is now available in
4188 $sender_address_data in subsequent conditions in the ACL statement.
4190 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4192 15. Added a new option "connect=<time>" to callout options, to set a different
4195 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4196 was its contents. (It was OK if the option was not defined at all.)
4198 17. A "Completed" log line is now written for messages that are removed from
4199 the spool by the -Mrm option.
4201 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4202 information about exactly what failed.
4204 19. Added -dd to debug only the daemon process.
4206 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4207 handles timeouts, both on the server side and network timeouts. Renamed the
4208 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4210 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4211 It is now set to "smtps".
4213 22. $host_address is now set to the target address during the checking of
4214 ignore_target_hosts.
4216 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4217 being passed; this would have caused $sender_host_name to have been used if
4218 matching the list had actually called for a host name (not very likely,
4219 since this list is usually IP addresses). A host name is now passed as
4222 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4223 code that specifies a non-restarting handler (typically sigaction() in
4224 modern systems) in an attempt to fix a rare and obscure crash bug.
4226 25. Narrowed the window for a race in the daemon that could cause it to ignore
4227 SIGCHLD signals. This is not a major problem, because they are used only to
4228 wake it up if nothing else does.
4230 26. A malformed maildirsize file could cause Exim to calculate negative values
4231 for the mailbox size or file count. Odd effects could occur as a result.
4232 The maildirsize information is now recalculated if the size or filecount
4235 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4236 support for a long time. Removed HAVE_SYS_VFS_H.
4238 28. Installed the latest version of exipick from John Jetmore.
4240 29. In an address list, if the pattern was not a regular expression, an empty
4241 subject address (from a bounce message) matched only if the pattern was an
4242 empty string. Non-empty patterns were not even tested. This was the wrong
4243 because it is perfectly reasonable to use an empty address as part of a
4244 database query. An empty address is now tested by patterns that are
4245 lookups. However, all the other forms of pattern expect the subject to
4246 contain a local part and a domain, and therefore, for them, an empty
4247 address still always fails if the pattern is not itself empty.
4249 30. Exim went into a mad DNS loop when attempting to do a callout where the
4250 host was specified on an smtp transport, and looking it up yielded more
4251 than one IP address.
4253 31. Re-factored the code for checking spool and log partition space into a
4254 function that finds that data and another that does the check. The former
4255 is then used to implement four new variables: $spool_space, $log_space,
4256 $spool_inodes, and $log_inodes.
4258 32. The RFC2047 encoding function was originally intended for short strings
4259 such as real names; it was not keeping to the 75-character limit for
4260 encoded words that the RFC imposes. It now respects the limit, and
4261 generates multiple encoded words if necessary. To be on the safe side, I
4262 have increased the buffer size for the ${rfc2047: expansion operator from
4265 33. It is now permitted to omit both strings after an "if" condition; if the
4266 condition is true, the result is "true". As before, when the second string
4267 is omitted, a false condition yields an empty string. This makes it less
4268 cumbersome to write custom ACL and router conditions.
4270 34. Failure to deliver a bounce message always caused it to be frozen, even if
4271 there was an errors_to setting on the router. The errors_to setting is now
4274 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4275 canonical form (fully expanded) before being placed in
4276 $sender_host_address.
4278 36. The table in the code that translates DNS record types into text (T_A to
4279 "A" for instance) was missing entries for NS and CNAME. It is just possible
4280 that this could have caused confusion if both these types were looked up
4281 for the same domain, because the text type is used as part of Exim's
4282 per-process caching. But the chance of anyone hitting this buglet seems
4285 37. The dnsdb lookup has been extended in a number of ways.
4287 (1) There is a new type, "zns", which walks up the domain tree until it
4288 finds some nameserver records. It should be used with care.
4290 (2) There is a new type, "mxh", which is like "mx" except that it returns
4291 just the host names, not the priorities.
4293 (3) It is now possible to give a list of domains (or IP addresses) to be
4294 looked up. The behaviour when one of the lookups defers can be
4295 controlled by a keyword.
4297 (4) It is now possible to specify the separator character for use when
4298 multiple records are returned.
4300 38. The dnslists ACL condition has been extended: it is now possible to supply
4301 a list of IP addresses and/or domains to be looked up in a particular DNS
4304 39. Added log_selector=+queue_time_overall.
4306 40. When running the queue in the test harness, wait just a tad after forking a
4307 delivery process, to get repeatability of debugging output.
4309 41. Include certificate and key file names in error message when GnuTLS fails
4310 to set them up, because the GnuTLS error message doesn't include the name
4311 of the failing file when there is a problem reading it.
4313 42. Allow both -bf and -bF in the same test run.
4315 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4317 44. The "Exiscan patch" is now merged into the mainline Exim source.
4319 45. Sometimes the final signoff response after QUIT could fail to get
4320 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4321 < 0 before doing a fflush(). This bug looks as though it goes back to the
4322 introduction of TLS in release 3.20, but "sometimes" must have been rare
4323 because the tests only now provoked it.
4325 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4326 (this can affect the format of dates).
4328 47. exim_tidydb, when checking for the continued existence of a message for
4329 which it has found a message-specific retry record, was not finding
4330 messages that were in split spool directories. Consequently, it was
4331 deleting retry records that should have stayed in existence.
4333 48. Steve fixed some bugs in eximstats.
4335 49. The SPA authentication driver was not abandoning authentication and moving
4336 on to the next authenticator when an expansion was forced to fail,
4337 contradicting the general specification for all authenticators. Instead it
4338 was generating a temporary error. It now behaves as specified.
4340 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4341 (the order specifies the preference for clients). The order is now AES256,
4342 AES128, 3DES, ARCFOUR128.
4344 51. Small patch to Sieve code - explicitly set From: when generating an
4347 52. Exim crashed if a remote delivery caused a very long error message to be
4348 recorded - for instance if somebody sent an entire SpamAssassin report back
4349 as a large number of 550 error lines. This bug was coincidentally fixed by
4350 increasing the size of one of Exim's internal buffers (big_buffer) that
4351 happened as part of the Exiscan merge. However, to be on the safe side, I
4352 have made the code more robust (and fixed the comments that describe what
4355 53. Now that there can be additional text after "Completed" in log lines (if
4356 the queue_time_overall log selector is set), a one-byte patch to exigrep
4357 was needed to allow it to recognize "Completed" as not the last thing in
4360 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4361 patch that reportedly fixes this has been added. I am not expert enough to
4362 create a test for it. This is what the patch creator wrote:
4364 "I found a little strange behaviour of ldap code when working with
4365 Windows 2003 AD Domain, where users was placed in more than one
4366 Organization Units. When I tried to give exim partial DN, the exit code
4367 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4368 But simultaneously result of request was absolutely normal ldap result,
4369 so I produce this patch..."
4371 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4372 REFERENCE, so I have modified the code to exclude the patch when that macro
4375 55. Some experimental protocols are using DNS PTR records for new purposes. The
4376 keys for these records are domain names, not reversed IP addresses. The
4377 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4378 leaves it alone. Component reversal etc. now happens only for IP addresses.
4381 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4383 57. Double the size of the debug message buffer (to 2048) so that more of very
4384 long debug lines gets shown.
4386 58. The exicyclog utility now does better if the number of log files to keep
4387 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4389 59. Two changes related to the smtp_active_hostname option:
4391 (1) $smtp_active_hostname is now available as a variable.
4392 (2) The default for smtp_banner uses $smtp_active_hostname instead
4393 of $primary_hostname.
4395 60. The host_aton() function is supposed to be passed a string that is known
4396 to be a valid IP address. However, in the case of IPv6 addresses, it was
4397 not checking this. This is a hostage to fortune. Exim now panics and dies
4398 if the condition is not met. A case was found where this could be provoked
4399 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4400 components; fortuitously, this particular loophole had already been fixed
4401 by change 4.50/55 above.
4403 If there are any other similar loopholes, the new check in host_aton()
4404 itself should stop them being exploited. The report I received stated that
4405 data on the command line could provoke the exploit when Exim was running as
4406 exim, but did not say which command line option was involved. All I could
4407 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4408 running as the user.
4411 61. There was a buffer overflow vulnerability in the SPA authentication code
4412 (which came originally from the Samba project). I have added a test to the
4413 spa_base64_to_bits() function which I hope fixes it.
4416 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4417 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4419 63. The daemon start-up calls getloadavg() while still root for those OS that
4420 need the first call to be done as root, but it missed one case: when
4421 deliver_queue_load_max is set with deliver_drop_privilege. This is
4422 necessary for the benefit of the queue runner, because there is no re-exec
4423 when deliver_drop_privilege is set.
4425 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4426 This has been fixed.
4428 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4429 was in use, was not putting the data itself into the right store pool;
4430 consequently, it could be overwritten for a subsequent message in the same
4431 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4434 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4436 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4437 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4438 the function were treating the return as a boolean value, which happened to
4439 work because 0=false and not-0=true, but is not correct code.
4441 68. The host_aton() function was not handling scoped IPv6 addresses (those
4442 with, for example, "%eth0" on the end) correctly.
4444 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4445 negated items (that is, ~something) in unsigned ints. Some compilers
4446 apparently mutter when there is no cast.
4448 70. If an address verification called from an ACL failed, and did not produce a
4449 user-specific message (i.e. there was only a "system" message), nothing was
4450 put in $acl_verify_message. In this situation, it now puts the system
4453 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4454 session; change 4.31/43 added the unwanted input to the log line - except
4455 that it did not do this in the start of session case. It now does.
4457 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4458 This is wrong and can cause the other end to generate a synchronization
4459 error if it is another Exim or anything else that does the synchronization
4460 check. A QUIT command is no longer sent after a timeout.
4462 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4463 during host lookups.
4465 74. The defer_ok option of callout verification was not working if it was used
4466 when verifying addresses in header lines, that is, for this case:
4468 verify = header_sender/callout=defer_ok
4470 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4471 those file descriptors could be used for SMTP connections. If anything
4472 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4473 could be sent to the SMTP client, causing chaos. The daemon now opens
4474 stdin, stdout, and stderr to /dev/null when it puts itself into the
4477 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4478 log by default. The user can override this with suitable Perl magic.
4480 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4481 the log message when discard triggers, was not working for the DATA ACL or
4482 for the non-SMTP ACL.
4484 78. Error message wording change in sieve.c.
4486 79. If smtp_accept_max_per_host was set, the number of connections could be
4487 restricted to fewer than expected, because the daemon was trying to set up
4488 a new connection before checking whether the processes handling previous
4489 connections had finished. The check for completed processes is now done
4490 earlier. On busy systems, this bug wouldn't be noticed because something
4491 else would have woken the daemon, and it would have reaped the completed
4494 80. If a message was submitted locally by a user whose login name contained one
4495 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4496 It caused a spool format error. I have fixed the spool reading code. A
4497 related problem was that the "from" clause in the Received: line became
4498 illegal because of the space(s). It is now covered by ${quote_local_part.
4500 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4503 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4505 83. Patch from Sieve maintainer to fix unterminated string problem in
4506 "vacation" handling.
4508 84. Some minor changes to the Linux configuration files to help with other
4509 OS variants using glibc.
4511 85. One more patch for Sieve to update vacation handling to latest spec.
4514 ----------------------------------------------------
4515 See the note above about the 4.44 and 4.50 releases.
4516 ----------------------------------------------------
4522 1. Change 4.43/35 introduced a bug that caused file counts to be
4523 incorrectly computed when quota_filecount was set in an appendfile
4526 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4527 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4530 3. Give more explanation in the error message when the command for a transport
4531 filter fails to execute.
4533 4. There are several places where Exim runs a non-Exim command in a
4534 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4535 was being done only for the command run by the queryprogram router. It is
4536 now done for all such subprocesses. The other cases are: ${run, transport
4537 filters, and the commands run by the lmtp and pipe transports.
4539 5. Some older OS have a limit of 256 on the maximum number of file
4540 descriptors. Exim was using setrlimit() to set 1000 as a large value
4541 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4542 systems. I've change it so that if it can't get 1000, it tries for 256.
4544 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4545 was an oversight, and furthermore, ever since the addition of extra
4546 controls (e.g. 4.43/32), the checks on when to allow different forms of
4547 "control" were broken. There should now be diagnostics for all cases when a
4548 control that does not make sense is encountered.
4550 7. $recipients is now available in the predata ACL (oversight).
4552 8. Tidy the search cache before the fork to do a delivery from a message
4553 received from the command line. Otherwise the child will trigger a lookup
4554 failure and thereby defer the delivery if it tries to use (for example) a
4555 cached ldap connection that the parent has called unbind on.
4557 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4558 of $address_data from the recipient verification was clobbered by the
4559 sender verification.
4561 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4562 was its contents. (It was OK if the option was not defined at all.)
4564 11. A "Completed" log line is now written for messages that are removed from
4565 the spool by the -Mrm option.
4567 12. $host_address is now set to the target address during the checking of
4568 ignore_target_hosts.
4570 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4571 being passed; this would have caused $sender_host_name to have been used if
4572 matching the list had actually called for a host name (not very likely,
4573 since this list is usually IP addresses). A host name is now passed as
4576 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4577 code that specifies a non-restarting handler (typically sigaction() in
4578 modern systems) in an attempt to fix a rare and obscure crash bug.
4580 15. Narrowed the window for a race in the daemon that could cause it to ignore
4581 SIGCHLD signals. This is not a major problem, because they are used only to
4582 wake it up if nothing else does.
4584 16. A malformed maildirsize file could cause Exim to calculate negative values
4585 for the mailbox size or file count. Odd effects could occur as a result.
4586 The maildirsize information is now recalculated if the size or filecount
4589 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4590 support for a long time. Removed HAVE_SYS_VFS_H.
4592 18. Updated exipick to current release from John Jetmore.
4594 19. Allow an empty sender to be matched against a lookup in an address list.
4595 Previously the only cases considered were a regular expression, or an
4598 20. Exim went into a mad DNS lookup loop when doing a callout where the
4599 host was specified on the transport, if the DNS lookup yielded more than
4602 21. The RFC2047 encoding function was originally intended for short strings
4603 such as real names; it was not keeping to the 75-character limit for
4604 encoded words that the RFC imposes. It now respects the limit, and
4605 generates multiple encoded words if necessary. To be on the safe side, I
4606 have increased the buffer size for the ${rfc2047: expansion operator from
4609 22. Failure to deliver a bounce message always caused it to be frozen, even if
4610 there was an errors_to setting on the router. The errors_to setting is now
4613 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4614 canonical form (fully expanded) before being placed in
4615 $sender_host_address.
4617 24. Updated eximstats to version 1.33
4619 25. Include certificate and key file names in error message when GnuTLS fails
4620 to set them up, because the GnuTLS error message doesn't include the name
4621 of the failing file when there is a problem reading it.
4623 26. Expand error message when OpenSSL has problems setting up cert/key files.
4626 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4627 (this can affect the format of dates).
4629 28. exim_tidydb, when checking for the continued existence of a message for
4630 which it has found a message-specific retry record, was not finding
4631 messages that were in split spool directories. Consequently, it was
4632 deleting retry records that should have stayed in existence.
4634 29. eximstats updated to version 1.35
4635 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4636 1.35 - bugfix such that pie charts by volume are generated correctly
4638 30. The SPA authentication driver was not abandoning authentication and moving
4639 on to the next authenticator when an expansion was forced to fail,
4640 contradicting the general specification for all authenticators. Instead it
4641 was generating a temporary error. It now behaves as specified.
4643 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4644 (the order specifies the preference for clients). The order is now AES256,
4645 AES128, 3DES, ARCFOUR128.
4647 31. Small patch to Sieve code - explicitly set From: when generating an
4650 32. Exim crashed if a remote delivery caused a very long error message to be
4651 recorded - for instance if somebody sent an entire SpamAssassin report back
4652 as a large number of 550 error lines. This bug was coincidentally fixed by
4653 increasing the size of one of Exim's internal buffers (big_buffer) that
4654 happened as part of the Exiscan merge. However, to be on the safe side, I
4655 have made the code more robust (and fixed the comments that describe what
4658 33. Some experimental protocols are using DNS PTR records for new purposes. The
4659 keys for these records are domain names, not reversed IP addresses. The
4660 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4661 leaves it alone. Component reversal etc. now happens only for IP addresses.
4664 34. The host_aton() function is supposed to be passed a string that is known
4665 to be a valid IP address. However, in the case of IPv6 addresses, it was
4666 not checking this. This is a hostage to fortune. Exim now panics and dies
4667 if the condition is not met. A case was found where this could be provoked
4668 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4669 components; fortuitously, this particular loophole had already been fixed
4670 by change 4.50/55 or 4.44/33 above.
4672 If there are any other similar loopholes, the new check in host_aton()
4673 itself should stop them being exploited. The report I received stated that
4674 data on the command line could provoke the exploit when Exim was running as
4675 exim, but did not say which command line option was involved. All I could
4676 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4677 running as the user.
4680 35. There was a buffer overflow vulnerability in the SPA authentication code
4681 (which came originally from the Samba project). I have added a test to the
4682 spa_base64_to_bits() function which I hope fixes it.
4685 36. The daemon start-up calls getloadavg() while still root for those OS that
4686 need the first call to be done as root, but it missed one case: when
4687 deliver_queue_load_max is set with deliver_drop_privilege. This is
4688 necessary for the benefit of the queue runner, because there is no re-exec
4689 when deliver_drop_privilege is set.
4691 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4692 was in use, was not putting the data itself into the right store pool;
4693 consequently, it could be overwritten for a subsequent message in the same
4694 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4697 38. Sometimes the final signoff response after QUIT could fail to get
4698 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4699 < 0 before doing a fflush(). This bug looks as though it goes back to the
4700 introduction of TLS in release 3.20, but "sometimes" must have been rare
4701 because the tests only now provoked it.
4707 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4708 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4709 Now it returns an integer. A number of calls were still expecting a T/F
4710 return. Fortuitously, in all cases, the tests worked in OK situations,
4711 which is the norm. However, things would have gone wrong on any write
4712 failures on the smtp file descriptor. This function is used when sending
4713 messages over SMTP and also when doing verify callouts.
4715 2. When Exim is called to do synchronous delivery of a locally submitted
4716 message (the -odf or -odi options), it no longer closes stderr before doing
4719 3. Implemented the mua_wrapper option.
4721 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4723 5. Implemented the functions header_remove(), header_testname(),
4724 header_add_at_position(), and receive_remove_recipient(), and exported them
4727 6. If an ACL "warn" statement specified the addition of headers, Exim already
4728 inserted X-ACL-Warn: at the start if there was no header name. However, it
4729 was not making this test for the second and subsequent header lines if
4730 there were newlines in the string. This meant that an invalid header could
4731 be inserted if Exim was badly configured.
4733 7. Allow an ACL "warn" statement to add header lines at the start or after all
4734 the Received: headers, as well as at the end.
4736 8. Added the rcpt_4xx retry error code.
4738 9. Added postmaster_mailfrom=xxx to callout verification option.
4740 10. Added mailfrom=xxxx to the callout verification option, for verify=
4743 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4744 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4746 12. Inserted some casts to stop certain compilers warning when using pointer
4747 differences as field lengths or precisions in printf-type calls (mostly
4748 affecting debugging statements).
4750 13. Added optional readline() support for -be (dynamically loaded).
4752 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4753 same clock tick as a message's arrival, so that its received time was the
4754 same as the "first fail" time on the retry record, and that message
4755 remained on the queue past the ultimate address timeout, every queue runner
4756 would try a delivery (because it was past the ultimate address timeout) but
4757 after another failure, the ultimate address timeout, which should have then
4758 bounced the address, did not kick in. This was a "< instead of <=" error;
4759 in most cases the first failure would have been in the next clock tick
4760 after the received time, and all would be well.
4762 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4763 being recognized when the domain list was tested by the match_domain
4764 condition in an expansion string.
4766 16. Added the ${str2b64: operator.
4768 17. Exim was always calling setrlimit() to set a large limit for the number of
4769 processes, without checking whether the existing limit was already
4770 adequate. (It did check for the limit on file descriptors.) Furthermore,
4771 errors from getrlimit() and setrlimit() were being ignored. Now they are
4772 logged to the main and panic logs, but Exim does carry on, to try to do its
4773 job under whatever limits there are.
4775 18. Imported PCRE 5.0.
4777 19. Trivial typo in log message " temporarily refused connection" (the leading
4780 20. If the log selector return_path_on_delivery was set and an address was
4781 redirected to /dev/null, the delivery process crashed because it assumed
4782 that a return path would always be set for a "successful" delivery. In this
4783 case, the whole delivery is bypassed as an optimization, and therefore no
4786 21. Internal re-arrangement: the function for sending a challenge and reading
4787 a response while authentication was assuming a zero-terminated challenge
4788 string. It's now changed to take a pointer and a length, to allow for
4789 binary data in such strings.
4791 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4793 23. Exim was not respecting finduser_retries when seeking the login of the
4794 uid under which it was called; it was always trying 10 times. (The default
4795 setting of finduser_retries is zero.) Also, it was sleeping after the final
4796 failure, which is pointless.
4798 24. Implemented tls_on_connect_ports.
4800 25. Implemented acl_smtp_predata.
4802 26. If the domain in control=submission is set empty, Exim assumes that the
4803 authenticated id is a complete email address when it generates From: or
4804 Sender: header lines.
4806 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4807 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4808 chown and chgrp in /bin and hostname in /usr/bin.
4810 28. Exim was keeping the "process log" file open after each use, just as it
4811 does for the main log. This opens the possibility of it remaining open for
4812 long periods when the USR1 signal hits a daemon. Occasional processlog
4813 errors were reported, that could have been caused by this. Anyway, it seems
4814 much more sensible not to leave this file open at all, so that is what now
4817 29. The long-running daemon process does not normally write to the log once it
4818 has entered its main loop, and it closes the log before doing so. This is
4819 so that log files can straightforwardly be renamed and moved. However,
4820 there are a couple of unusual error situations where the daemon does write
4821 log entries, and I had neglected to close the log afterwards.
4823 30. The text of an SMTP error response that was received during a remote
4824 delivery was being truncated at 512 bytes. This is too short for some of
4825 the long messages that one sometimes sees. I've increased the limit to
4828 31. It is now possible to make retry rules that apply only when a message has a
4829 specific sender, in particular, an empty sender.
4831 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4832 it possible to be selective about when SMTP synchronization is enforced.
4834 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4836 32. Implemented hosts_connection_nolog.
4838 33. Added an ACL for QUIT.
4840 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4843 35. Added mailbox_size and mailbox_filecount to appendfile.
4845 36. Added control = no_multiline_responses to ACLs.
4847 37. There was a bug in the logic of the code that waits for the clock to tick
4848 in the case where the clock went backwards by a substantial amount such
4849 that the microsecond fraction of "now" was more than the microsecond
4850 fraction of "then" (but the whole seconds number was less).
4852 38. Added support for the libradius Radius client library this is found on
4853 FreeBSD (previously only the radiusclient library was supported).
4859 1. When certain lookups returned multiple values in the form name=value, the
4860 quoting of the values was not always being done properly. Specifically:
4861 (a) If the value started with a double quote, but contained no whitespace,
4863 (b) If the value contained whitespace other than a space character (i.e.
4864 tabs or newlines or carriage returns) it was not quoted.
4865 This fix has been applied to the mysql and pgsql lookups by writing a
4866 separate quoting function and calling it from the lookup code. The fix
4867 should probably also be applied to nisplus, ibase and oracle lookups, but
4868 since I cannot test any of those, I have not disturbed their existing code.
4870 2. A hit in the callout cache for a specific address caused a log line with no
4871 reason for rejecting RCPT. Now it says "Previous (cached) callout
4872 verification failure".
4874 3. There was an off-by-one bug in the queryprogram router. An over-long
4875 return line was truncated at 256 instead of 255 characters, thereby
4876 overflowing its buffer with the terminating zero. As well as fixing this, I
4877 have increased the buffer size to 1024 (and made a note to document this).
4879 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4880 when Exim is waiting for an SMTP response from a remote server, Exim
4881 restarts its select() call on the socket, thereby resetting its timeout.
4882 This is not a problem when such interrupts are rare. Somebody set up a cron
4883 job to run exiwhat every 2 minutes, which is less than the normal select()
4884 timeout (5 or 10 minutes). This meant that the select() timeout never
4885 kicked in because it was always reset. I have fixed this by comparing the
4886 time when an interrupt arrives with the time at the start of the first call
4887 to select(). If more time than the timeout has elapsed, the interrupt is
4888 treated as a timeout.
4890 5. Some internal re-factoring in preparation for the addition of Sieve
4891 extensions (by MH). In particular, the "personal" test is moved to a
4892 separate function, and given an option for scanning Cc: and Bcc: (which is
4893 not set for Exim filters).
4895 6. When Exim created an email address using the login of the caller as the
4896 local part (e.g. when creating a From: or Sender: header line), it was not
4897 quoting the local part when it contained special characters such as @.
4899 7. Installed new OpenBSD configuration files.
4901 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4902 try to make them clearer.
4904 9. Callout options, other than the timeout value, were being ignored when
4905 verifying sender addresses in header lines. For example, when using
4907 verify = header_sender/callout=no_cache
4909 the cache was (incorrectly) being used.
4911 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4912 only the Cygwin environment.
4914 11. When return_path_on_delivery was set as a log selector, if different remote
4915 addresses in the same message used different return paths and parallel
4916 remote delivery occurred, the wrong values would sometimes be logged.
4917 (Whenever a remote delivery process finished, the return path value from
4918 the most recently started remote delivery process was logged.)
4920 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4921 lines when AUTH and/or TLS are in use. This is the "received protocol"
4922 field. Exim used to use "asmtp" for authenticated SMTP, without any
4923 indication (in the protocol name) for TLS use. Now it follows the RFC and
4924 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4925 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4926 names appear in log lines as well as in Received: header lines.
4928 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4929 extensions, and comparison tests, and to fix some bugs.
4931 14. Changes to the "personal" filter test:
4933 (1) The test was buggy in that it was just doing the equivalent of
4934 "contains" tests on header lines. For example, if a user's address was
4935 anne@some.where, the "personal" test would incorrectly be true for
4937 To: susanne@some.where
4939 This test is now done by extracting each address from the header in turn,
4940 and checking the entire address. Other tests that are part of "personal"
4941 are now done using regular expressions (for example, to check local parts
4942 of addresses in From: header lines).
4944 (2) The list of non-personal local parts in From: addresses has been
4945 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4946 taken from the Sieve specification recommendations.
4948 (3) If the message contains any header line starting with "List-" it is
4949 treated as non-personal.
4951 (4) The test for "circular" in the Subject: header line has been removed
4952 because it now seems ill-conceived.
4954 15. Minor typos in src/EDITME comments corrected.
4956 16. Installed latest exipick from John Jetmore.
4958 17. If headers_add on a router specified a text string that was too long for
4959 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4960 of string_sprintf() is now avoided.
4962 18. $message_body_size was not set (it was always zero) when running the DATA
4963 ACL and the local_scan() function.
4965 19. For the "mail" command in an Exim filter, no default was being set for
4966 the once_repeat time, causing a random time value to be used if "once" was
4967 specified. (If the value happened to be <= 0, no repeat happened.) The
4968 default is now 0s, meaning "never repeat". The "vacation" command was OK
4969 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4970 (I found it when inspecting the code).
4972 20. There is now an overall timeout for performing a callout verification. It
4973 defaults to 4 times the callout timeout, which applies to individual SMTP
4974 commands during the callout. The overall timeout applies when there is more
4975 than one host that can be tried. The timeout is checked before trying the
4976 next host. This prevents very long delays if there are a large number of
4977 hosts and all are timing out (e.g. when the network connections are timing
4978 out). The value of the overall timeout can be changed by specifying an
4979 additional sub-option for "callout", called "maxwait". For example:
4981 verify = sender/callout=5s,maxwait=20s
4983 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4984 to the end before writing, but this should make it even safer).
4986 22. Exim was forgetting that it had advertised PIPELINING for the second and
4987 subsequent messages on an SMTP connection. It was also not resetting its
4988 memory on STARTTLS and an internal HELO.
4990 23. When Exim logs an SMTP synchronization error within a session, it now
4991 records whether PIPELINING has been advertised or not.
4993 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4994 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4995 rather than long int.
4997 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4999 26. Added the never_mail option to autoreply.
5005 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5006 crash if the getsockname() call failed; this can happen if a connection is
5007 closed very soon after it is established. The problem was simply in the
5008 order in which certain operations were done, causing Exim to try to write
5009 to the SMTP stream before it had set up the file descriptor. The bug has
5010 been fixed by making things happen in the correct order.
5016 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5017 before the connection was closed, thus losing the rejection response.
5019 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5020 some early Solaris releases, but causes trouble in current releases where
5021 socklen_t is defined.
5023 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5026 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5029 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5030 that releasing the top of it at the end releases what was used for sub-
5031 expansions (unless the block got too big). However, discard this block if
5032 the first thing is a variable or header, so that we can use its block when
5033 it is dynamic (useful for very large $message_headers, for example).
5035 6. Lookups now cache *every* query, not just the most recent. A new, separate
5036 store pool is used for this. It can be recovered when all lookup caches are
5037 flushed. Lookups now release memory at the end of their result strings.
5038 This has involved some general refactoring of the lookup sources.
5040 7. Some code has been added to the store_xxx() functions to reduce the amount
5041 of flapping under certain conditions.
5043 8. log_incoming_interface used to affect only the <= reception log lines. Now
5044 it causes the local interface and port to be added to several more SMTP log
5045 lines, for example "SMTP connection from", and rejection lines.
5047 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5049 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5051 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5052 could be overwritten at the end of the current message (or the start of a
5053 new message if it was set in a HELO ACL). The value is now preserved for
5054 the duration of the SMTP connection.
5056 12. If a transport had a headers_rewrite setting, and a matching header line
5057 contained an unqualified address, that address was qualified, even if it
5058 did not match any rewriting rules. The underlying bug was that the values
5059 of the flags that permit the existence of unqualified sender and recipient
5060 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5061 non-local messages, and by -bnq for local messages) were not being
5062 preserved with the message after it was received.
5064 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5065 "next input=" as part of the text comprising the host identity instead of
5066 the correct text. The code was using the same buffer for two different
5067 strings. However, depending on which order the printing function evaluated
5068 its arguments, the bug did not always show up. Under Linux, for example, my
5069 test suite worked just fine.
5071 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5072 doesn't work with some older versions of Perl. It has been changed to "my",
5073 which in any case is probably the better facility to use.
5075 15. A really picky compiler found some instances of statements for creating
5076 error messages that either had too many or two few arguments for the format
5079 16. The size of the buffer for calls to the DNS resolver has been increased
5080 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5081 for addresses that have a lot of PTR records. This alleviates a problem; it
5082 does not fully solve it.
5084 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5085 buffer now truncates the list and logs the incident, which is the same
5086 action as happens when Exim is looking up a host name and its aliases.
5087 Previously in this situation something unpredictable would happen;
5088 sometimes it was "internal error: store_reset failed".
5090 18. If a server dropped the connection unexpectedly when an Exim client was
5091 using GnuTLS and trying to read a response, the client delivery process
5092 crashed while trying to generate an error log message.
5094 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5095 string, for example:
5097 warn message = H1: something\nH2: something
5099 the text was added as a single header line from Exim's point of view
5100 though it ended up OK in the delivered message. However, searching for the
5101 second and subsequent header lines using $h_h2: did not work. This has been
5102 fixed. Similarly, if a system filter added multiple headers in this way,
5103 the routers could not see them.
5105 20. Expanded the error message when iplsearch is called with an invalid key to
5106 suggest using net-iplsearch in a host list.
5108 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5109 ACLs are no longer actually imposed (and a message to that effect is
5112 22. If a "gecos" field in a passwd entry contained escaped characters, in
5113 particular, if it contained a \" sequence, Exim got it wrong when building
5114 a From: or a Sender: header from that name. A second bug also caused
5115 incorrect handling when an unquoted " was present following a character
5116 that needed quoting.
5118 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5119 was not being matched caselessly.
5121 24. Arranged for all hyphens in the exim.8 source to be escaped with
5124 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5125 properly. Recipient callout cache records were still being keyed to include
5126 the sender, even when use_sender was set false. This led to far more
5127 callouts that were necessary. The sender is no longer included in the key
5128 when use_sender is false.
5130 26. Added "control = submission" modifier to ACLs.
5132 27. Added the ${base62d: operator to decode base 62 numbers.
5134 28. dnsdb lookups can now access SRV records.
5136 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5137 the configuration file.
5139 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5140 (-v) mode. This makes the output for a verbose queue run more intelligible.
5142 31. Added a use_postmaster feature to recipient callouts.
5144 32. Added the $body_zerocount variable, containing the number of binary zero
5145 bytes in the message body.
5147 33. The time of last modification of the "new" subdirectory is now used as the
5148 "mailbox time last read" when there is a quota error for a maildir
5151 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5153 35. Added +ignore_unknown as a special item in host lists.
5155 36. Code for decoding IPv6 addresses in host lists is now included, even if
5156 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5157 address was recognized as an IP address, but was then not correctly decoded
5158 into binary, causing unexpected and incorrect effects when compared with
5165 1. Very minor rewording of debugging text in manualroute to say "list of
5166 hosts" instead of "hostlist".
5168 2. If verify=header_syntax was set, and a header line with an unqualified
5169 address (no domain) and a large number of spaces between the end of the
5170 name and the colon was received, the reception process suffered a buffer
5171 overflow, and (when I tested it) crashed. This was caused by some obsolete
5172 code that should have been removed. The fix is to remove it!
5174 3. When running in the test harness, delay a bit after writing a bounce
5175 message to get a bit more predictability in the log output.
5177 4. Added a call to search_tidyup() just before forking a reception process. In
5178 theory, someone could use a lookup in the expansion of smtp_accept_max_
5179 per_host which, without the tidyup, could leave open a database connection.
5181 5. Added the variables $recipient_data and $sender_data which get set from a
5182 lookup success in an ACL "recipients" or "senders" condition, or a router
5183 "senders" option, similar to $domain_data and $local_part_data.
5185 6. Moved the writing of debug_print from before to after the "senders" test
5188 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5189 problems for message scanning, either using a data ACL, or using
5190 local_scan() because the Received: header was not generated till after they
5191 were called (in order to set the time as the time of reception completion).
5192 I have revised the way this works. The header is now generated after the
5193 body is received, but before the ACL or local_scan() are called. After they
5194 are run, the timestamp in the header is updated.
5200 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5201 before starting a queue runner without re-exec. This happened only when
5202 deliver_drop_privilege was set or when the Exim user was set to root. The
5203 effect of the bug was that timeouts during subsequent deliveries caused
5204 crashes instead of being properly handled. The handler is now left at its
5205 default (and expected) setting.
5207 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5208 message, again when deliver_drop_privilege is set or Exim is run as root.
5209 The bug described in (1) was not present in this case, but the tidying up
5210 of the other signals was missing. I have made the two cases consistent.
5212 3. The ignore_target_hosts setting on a manualroute router was being ignored
5213 for hosts that were looked up using the /MX notation.
5215 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5218 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5219 operated on the sender address. After changing the $sender_address to <>
5220 for the sender address verify, Exim was re-instated it as the original
5221 (before rewriting) address, but remembering that it had rewritten it, so it
5222 wasn't rewriting it again. This bug also had the effect of breaking the
5223 sender address verification caching when the sender address was rewritten.
5225 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5226 This has been changed so that if the ip literal address matches
5227 ignore_target_hosts, the router declines.
5229 7. Added expansion conditions match_domain, match_address, and match_local_
5230 part (NOT match_host).
5232 8. The placeholder for the Received: header didn't have a length field set.
5234 9. Added code to Exim itself and to exim_lock to test for a specific race
5235 condition that could lead to file corruption when using MBX delivery. The
5236 issue is with the lockfile that is created in /tmp. If this file is removed
5237 after a process has opened it but before that process has acquired a lock,
5238 there is the potential for a second process to recreate the file and also
5239 acquire a lock. This could lead to two Exim processes writing to the file
5240 at the same time. The added code performs the same test as UW imapd; it
5241 checks after acquiring the lock that its file descriptor still refers to
5242 the same named file.
5244 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5245 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5248 11. Added the smtp_active_hostname option. If used, this will typically be made
5249 to depend on the incoming interface address. Because $interface_address is
5250 not set up until the daemon has forked a reception process, error responses
5251 that can happen earlier (such as "too many connections") no longer contain
5254 12. If an expansion in a condition on a "warn" statement fails because a lookup
5255 defers, the "warn" statement is abandoned, and the next ACL statement is
5256 processed. Previously this caused the whole ACL to be aborted.
5258 13. Added the iplsearch lookup type.
5260 14. Added ident_timeout as a log selector.
5262 15. Added tls_certificate_verified as a log selector.
5264 16. Added a global option tls_require_ciphers (compare the smtp transport
5265 option of the same name). This controls incoming TLS connections.
5267 17. I finally figured out how to make tls_require_ciphers do a similar thing
5268 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5269 before starting the TLS session.
5271 18. Tabs are now shown as \t in -bP output.
5273 19. If the log selector return_path_on_delivery was set, Exim crashed when
5274 bouncing a message because it had too many Received: header lines.
5276 20. If two routers both had headers_remove settings, and the first one included
5277 a superfluous trailing colon, the final name in the first list and the
5278 first name in the second list were incorrectly joined into one item (with a
5279 colon in the middle).
5285 1. Added -C and -D options to the exinext utility, mainly to make it easier
5286 to include in the automated testing, but these could be helpful when
5287 multiple configurations are in use.
5289 2. The exinext utility was not formatting the output nicely when there was
5290 an alternate port involved in the retry record key, nor when there was a
5291 message id as well (for retries that were specific to a specific message
5292 and a specific host). It was also confused by IPv6 addresses, because of
5293 the additional colons they contain. I have fixed the IPv4 problem, and
5294 patched it up to do a reasonable job for IPv6.
5296 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5297 delivery, the log line now contains "pipelined" if PIPELINING was used.
5299 4. An SMTP transport process used to panic and die if the bind() call to set
5300 an explicit outgoing interface failed. This has been changed; it is now
5301 treated in the same way as a connect() failure.
5303 5. A reference to $sender_host_name in the part of a conditional expansion
5304 that was being skipped was still causing a DNS lookup. This no longer
5307 6. The def: expansion condition was not recognizing references to header lines
5308 that used bh_ and bheader_.
5310 7. Added the _cache feature to named lists.
5312 8. The code for checking quota_filecount in the appendfile transport was
5313 allowing one more file than it should have been.
5315 9. For compatibility with Sendmail, the command line option
5323 and sets the incoming protocol and host name (for trusted callers). The
5324 host name and its colon can be omitted when only the protocol is to be set.
5325 Note the Exim already has two private options, -pd and -ps, that refer to
5326 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5327 "s", but I don't think that's a major issue.
5329 10. A number of refactoring changes to the code, none of which should affect
5332 (a) The number of logging options was getting close to filling up the
5333 32-bit word that was used as a bit map. I have split them into two classes:
5334 those that are passed in the argument to log_write(), and those that are
5335 only ever tested independently outside of that function. These are now in
5336 separate 32-bit words, so there is plenty of room for expansion again.
5337 There is no change in the user interface or the logging behaviour.
5339 (b) When building, for example, log lines, the code previously used a
5340 macro that called string_cat() twice, in order to add two strings. This is
5341 not really sufficiently general. Furthermore, there was one instance where
5342 it was actually wrong because one of the argument was used twice, and in
5343 one call a function was used. (As it happened, calling the function twice
5344 did not affect the overall behaviour.) The macro has been replaced by a
5345 function that can join an arbitrary number of extra strings onto a growing
5348 (c) The code for expansion conditions now uses a table and a binary chop
5349 instead of a serial search (which was left over from when there were very
5350 few conditions). Also, it now recognizes conditions like "pam" even when
5351 the relevant support is not compiled in: a suitably worded error message is
5352 given if an attempt is made to use such a condition.
5354 11. Added ${time_interval:xxxxx}.
5356 12. A bug was causing one of the ddress fields not to be passed back correctly
5357 from remote delivery subprocesses. The field in question was not being
5358 subsequently used, so this caused to problems in practice.
5360 13. Added new log selectors queue_time and deliver_time.
5362 14. Might have fixed a bug in maildirsizefile handling that threw up
5363 "unexpected character" debug warnings, and recalculated the data
5364 unnecessarily. In any case, I expanded the warning message to give more
5367 15. Added the message "Restricted characters in address" to the statements in
5368 the default ACL that block characters like @ and % in local parts.
5370 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5371 Three changes have been made:
5373 (a) There was a serious bug; a negative response to MAIL caused the whole
5374 recipient domain to be cached as invalid, thereby blocking all messages
5375 to all local parts at the same domain, from all senders. This bug has
5376 been fixed. The domain is no longer cached after a negative response to
5377 MAIL if the sender used is not empty.
5379 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5382 (c) A new callout option, "use_sender" has been added for people who want
5383 the modified behaviour.
5389 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5392 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5393 indeed breaks things for older releases.
5395 3. Added additional logging to the case where there is a problem reading data
5396 from a filter that is running in a subprocess using a pipe, in order to
5397 try to track down a specific problem.
5399 4. Testing facility fudge: when running in the test harness and attempting
5400 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5401 getting "No route to host". Convert this to a timeout.
5403 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5406 6. Some OS don't have socklen_t but use size_t instead. This affects the
5407 fifth argument of getsockopt() amongst other things. This is now
5408 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5409 can be set for individual OS. I have set it for SunOS5, OSF1, and
5410 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5411 some earlier ones do not.
5413 7. Change 4.30/15 was not doing the test caselessly.
5415 8. The standard form for an IPv6 address literal was being rejected by address
5416 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5417 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5418 this, as well as the form without the "IPv6" on the front (but only when
5419 address literals are enabled, of course).
5421 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5423 10. Exim crashed if a message with an empty sender address specified by -f
5424 encountered a router with an errors_to setting. This could be provoked only
5425 by a command such as
5429 where an empty string was supplied; "<>" did not hit this bug.
5431 11. Installed PCRE release 4.5.
5433 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5434 remained set. It is now erased.
5436 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5437 times from message ids (which are base 36 rather than the normal 62).
5439 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5440 were being counted as actual protocol errors, and logged if the log
5441 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5442 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5443 and DATA following a set of rejected RCPTs do not count as protocol errors.
5444 In other words, Exim assumes they were pipelined, though this may not
5445 actually be the case. Of course, in all cases the client gets an
5446 appropriate error code.
5448 15. If a lookup fails in an ACL condition, a message about the failure may
5449 be available; it is used if testing the ACL cannot continue, because most
5450 such messages specify what the cause of the deferral is. However, some
5451 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5452 that caused an old message to be retained and used if a later statement
5453 caused a defer, replacing the real cause of the deferral.
5455 16. If an IP address had so many PTR records that the DNS lookup buffer
5456 was not large enough to hold them, Exim could crash while trying to process
5457 the truncated data. It now detects and logs this case.
5459 17. Further to 4.21/58, another change has been made: if (and only if) the
5460 first line of a message (the first header line) ends with CRLF, a bare LF
5461 in a subsequent header line has a space inserted after it, so as not to
5462 terminate the header.
5464 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5465 unnecessarily, used atoi() instead of strtol(), and didn't check the
5466 termination when getting file sizes from file names by regex.
5468 19. Completely re-implemented the support for maildirsize files, in the light
5469 of a number of problems with the previous contributed implementation
5470 (4.30/29). In particular:
5472 . If the quota is zero, the maildirsize file is maintained, but no quota is
5475 . If the maildir directory does not exist, it is created before any attempt
5476 to write a maildirsize file.
5478 . The quota value in the file is just a cache; if the quota is changed in
5479 the transport, the new value overrides.
5481 . A regular expression is available for excluding directories from the
5484 20. The autoreply transport checks the characters in options that define the
5485 message's headers; it allows continued headers, but it was checking with
5486 isspace() after an embedded newline instead of explicitly looking for a
5489 21. If all the "regular" hosts to which an address was routed had passed their
5490 expiry times, and had not reached their retry times, the address was
5491 bounced, even if fallback hosts were defined. Now Exim should go on to try
5494 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5495 equivalent code in the SMTP transport. Some hosts send humungous responses
5496 to HELO/EHLO, more than 1024 it seems.
5498 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5499 gives compiler warnings in some environments. I've now done it "properly",
5502 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5503 (because of problems with the built-in one) was declared to return uschar *
5504 instead of char *, causing compiler failure.
5506 25. Fixed a file descriptor leak when processing alias/forward files.
5508 26. Fixed a minor format string issue in dbfn.c.
5510 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5512 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5513 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5516 29. When a sender address is verified, it is cached, to save repeating the test
5517 when there is more than one recipient in a message. However, when the
5518 verification involves a callout, it is possible for different callout
5519 options to be set for different recipients. It is too complicated to keep
5520 track of this in the cache, so now Exim always runs a verification when a
5521 callout is required, relying on the callout cache for the optimization.
5522 The overhead is duplication of the address routing, but this should not be
5525 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5526 to be verified with callout=postmaster, and the main callout worked but the
5527 postmaster check failed, the verification correctly failed. However, if a
5528 subsequent RCPT command asked for sender verification *without* the
5529 postmaster check, incorrect caching caused this verification also to fail,
5532 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5533 it was not caching the DNS options (qualify_single, search_parents) that
5534 were used when the lookup failed. A subsequent lookup with different
5535 options therefore always gave the same answer, though there were cases
5536 where it should not have. (Example: a "domains = !$mx_any" option on a
5537 dnslookup router: the "domains" option is always processed without any
5538 widening, but the router might have qualify_single set.) Now Exim uses the
5539 cached value only when the same options are set.
5541 32. Added John Jetmore's "exipick" utility to the distribution.
5543 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5544 than a timeout (e.g. a certificate is required, and is not provided), an
5545 Exim server now closes the connection immediately. Previously it waited for
5546 the client to close - but if the client is SSL, it seems that they each
5547 wait for each other, leading to a delay before one of them times out.
5549 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5550 maintained 0.8.x compatibility because I don't think many are using it, and
5551 it is clearly obsolete.
5553 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5556 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5557 client certificate was expired. A simple patch fixes this, though I don't
5558 understand the full logic of why the verify callback is called multiple
5561 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5562 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5563 which causes problems with some clients (such as the Certicom SSL Plus
5564 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5565 disables the coutermeasure allowing Eudora to connect."
5567 38. Exim was not checking that a write() to a log file succeeded. This could
5568 lead to Bad Things if a log got too big, in particular if it hit a file
5569 size limit. Exim now panics and dies if it cannot write to a log file, just
5570 as it does if it cannot open a log file.
5572 39. Modified OS/Makefile-Linux so that it now contains
5574 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5576 The two -D definitions ensure that Exim is compiled with large file
5577 support, which makes it possible to handle log files that are bigger than
5580 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5581 instance) a domain was checked against a named list that involved a lookup,
5582 causing $domain_data to be set, then another domain was checked against the
5583 same list, then the first domain was re-checked, the value of $domain_data
5584 after the final check could be wrong. In particular, if the second check
5585 failed, it could be set empty. This bug probably also applied to
5588 41. The strip_trailing_dot option was not being applied to the address given
5589 with the -f command-line option.
5591 42. The code for reading a message's header from the spool was incrementing
5592 $received_count, but never initializing it. This meant that the value was
5593 incorrect (doubled) while delivering a message in the same process in which
5594 it was received. In the most common configuration of Exim, this never
5595 happens - a fresh exec is done - but it can happen when
5596 deliver_drop_privilege is set.
5598 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5599 it now includes up to 150 characters of the unexpected data in the log
5602 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5603 and building data strings. The size of both of these buffers was 10 000
5604 bytes - far larger than anybody would *ever* want, thought I. Needless to
5605 say, somebody hit the limit. I have increased the maximum line length to
5606 20 000 and the maximum data length of concatenated lines to 100 000. I have
5607 also fixed two bugs, because there was no checking on these buffers. Tsk,
5608 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5609 buffer is too small.
5611 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5612 lsearch lookups. Now it does.
5614 46. When parsing a route_list item in a manualroute router, a fixed-length
5615 buffer was used for the list of hosts. I made this 1024 bytes long,
5616 thinking that nobody would ever have a list of hosts that long. Wrong.
5617 Somebody had a whole pile of complicated expansion conditions, and the
5618 string was silently truncated, leading to an expansion error. It turns out
5619 that it is easier to change to an unlimited length (owing to other changes
5620 that have happened since this code was originally written) than to build
5621 structure for giving a limitation error. The length of the item that
5622 expands into the list of hosts is now unlimited.
5624 47. The lsearch lookup could not handle data where the length of text line was
5625 more than 4095 characters. Such lines were truncated, leading to shortened
5626 data being returned. It should now handle lines of any length.
5628 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5629 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5632 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5633 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5634 no longer happens in comment lines. A list of replacements is now placed
5635 at the head of all of the source files, except those whose only change is
5636 to replace PERL_COMMAND in the very first #! line.
5638 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5639 messages on the queue, with a bottom-up merge sort, using code contributed
5640 by Michael Haardt. This should make operations like -bp somewhat faster on
5641 large queues. It won't affect queue runners, except when queue_run_in_order
5644 51. Installed eximstats 1.31 in the distribution.
5646 52. Added support for SRV lookups to the dnslookup router.
5648 53. If an ACL referred to $message_body or $message_body_end, the value was not
5649 reset for any messages that followed in the same SMTP session.
5651 54. The store-handling optimization for building very long strings was not
5652 differentiating between the different store pools. I don't think this
5653 actually made any difference in practice, but I've tidied it.
5655 55. While running the routers to verify a sender address, $sender_address
5656 was still set to the sender address. This is wrong, because when routing to
5657 send a bounce to the sender, it would be empty. Therefore, I have changed
5658 it so that, while verifying a sender address, $sender_address is set to <>.
5659 (There is no change to what happens when verifying a recipient address.)
5661 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5662 target A or AAAA records (if not already returned) without resetting the
5663 qualify_single or search_parents options of the DNS resolver. These are
5664 inappropriate in this case because the targets of MX and SRV records must
5665 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5666 target that, when qualified, matched something in the local domain. These
5667 two options are now turned off when doing these lookups.
5669 57. It seems that at least some releases of Reiserfs (which does not have the
5670 concept of a fixed number of inodes) returns zero and not -1 for the
5671 number of available inodes. This interacted badly with check_spool_inodes,
5672 which assumed that -1 was the "no such thing" setting. What I have done is
5673 to check that the total number of inodes is greater than zero before doing
5674 the test of how many are available.
5676 58. When a "warn" ACL statement has a log_message modifier, the message is
5677 remembered, and not repeated. This is to avoid a lot of repetition when a
5678 message has many recipients that cause the same warning to be written.
5679 However, Exim was preserving the list of already written lines for an
5680 entire SMTP session, which doesn't seem right. The memory is now reset if a
5681 new message is started.
5683 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5684 debugging output unless log_rewrite was also set.
5686 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5687 of (int)(handle) when we know that handle contains (void *)(-1).
5689 61. The Exim daemon panic-logs an error return when it closes the incoming
5690 connection. However "connection reset by peer" seems to be common, and
5691 isn't really an error worthy of noting specially, so that particular error
5694 62. When Exim is trying to find all the local interfaces, it used to panic and
5695 die if the ioctl to get the interface flags failed. However, it seems that
5696 on at least one OS (Solaris 9) it is possible to have an interface that is
5697 included in the list of interfaces, but for which you get a failure error
5698 for this call. This happens when the interface is not "plumbed" into a
5699 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5700 failure of the "get flags" call assumes that the interface is down.
5702 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5703 makes life easier for people who are doing arithmetic on fields extracted
5704 from dates, where you often get leading zeros that should not be
5705 interpreted as octal.
5707 64. Added qualify_domain to the redirect router, to override the global
5710 65. If a pathologically long header line contained very many addresses (the
5711 report of this problem mentioned 10 000) and each of them was rewritten,
5712 Exim could use up a very large amount of memory. (It kept on making new
5713 copies of the header line as it rewrote, and never released the old ones.)
5714 At the expense of a bit more processing, the header rewriting function has
5715 been changed so that it no longer eats memory in this way.
5717 66. The generation of the Received: header has been moved from the time that a
5718 message starts to be received, to the time that it finishes. The timestamp
5719 in the Received: header should now be very close to that of the <= log
5720 line. There are two side-effects of this change:
5722 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5723 logged header lines no longer include the local Received: line, because
5724 it has not yet been created. The same applies to a copy of the message
5725 that is returned to a non-SMTP sender when a message is rejected.
5727 (b) When a filter file is tested using -bf, no additional Received: header
5728 is added to the test message. After some thought, I decided that this
5731 This change does not affect the value of $received_for. It is still set
5732 after address rewriting, but before local_scan() is called.
5734 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5736 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5737 gave an unhelpful panic error message, and a defer error. I have managed to
5738 change this behaviour so that it now rejects any supplied certificate,
5739 which seems right, as the list of acceptable certificates is empty.
5741 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5742 gave an unhelpful defer error. I have not managed to make this reject any
5743 supplied certificates, but the error message it gives is "no certificate
5744 supplied", which is not helpful.
5746 70. exigrep's output now also includes lines that are not associated with any
5747 message, but which match the given pattern. Implemented by a patch from
5748 Martin Sluka, which also tidied up the Perl a bit.
5750 71. Recipient callout verification, like sender verification, was using <> in
5751 the MAIL FROM command. This isn't really the right thing, since the actual
5752 sender may affect whether the remote host accepts the recipient or not. I
5753 have changed it to use the actual sender in the callout; this means that
5754 the cache record is now keyed on a recipient/sender pair, not just the
5755 recipient address. There doesn't seem to be a real danger of callout loops,
5756 since a callout by the remote host to check the sender would use <>.
5757 [SEE ABOVE: changed after hitting problems.]
5759 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5760 temporary errors. However, in the case of such a code being given after
5761 the end of a data transmission (i.e. after ".") Exim was failing to write
5762 a retry record for the message. (Yes, there was some broken host that was
5763 actually sending 8xx at this point.)
5765 73. An unknown lookup type in a host list could cause Exim to panic-die when
5766 the list was checked. (An example that provoked this was putting <; in the
5767 middle of a list instead of at the start.) If this happened during a DATA
5768 ACL check, a -D file could be left lying around. This kind of configuration
5769 error no longer causes Exim to die; instead it causes a defer error. The
5770 incident is still logged to the main and panic logs.
5772 74. Buglet left over from Exim 3 conversion. The message "too many messages
5773 in one connection" was written to the rejectlog but not the mainlog, except
5774 when address rewriting (yes!) was being logged.
5776 75. Added write_rejectlog option.
5778 76. When a system filter was run not as root (that is, when system_filter_user
5779 was set), the values of the $n variables were not being returned to the
5780 main process; thus, they were not subsequently available in the $sn
5783 77. Added +return_path_on_delivery log selector.
5785 78. A connection timeout was being treated differently from recipients deferred
5786 when testing hosts_max_try with a message that was older than the host's
5787 retry timeout. (The host should not be counted, thus allowing all hosts to
5788 be tried at least once before bouncing.) This may have been the cause of an
5789 occasionally reported bug whereby a message would remain on the queue
5790 longer than the retry timeout, but would be bounced if a delivery was
5791 forced. I say "may" because I never totally pinned down the problem;
5792 setting up timeout/retry tests is difficult. See also the next item.
5794 79. The ultimate address timeout was not being applied to errors that involved
5795 a combination of host plus message (for example, a timeout on a MAIL
5796 command). When an address resolved to a number of possible hosts, and they
5797 were not all tried for each delivery (e.g. because of hosts_max_try), a
5798 message could remain on the queue longer than the retry timeout.
5800 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5803 81. Fixed an obscure SMTP outgoing bug which required at least the following
5804 conditions: (a) there was another message waiting for the same server;
5805 (b) the server returned 5xx to all RCPT commands in the first message so
5806 that the message was not completed; (c) the server dropped the connection
5807 or gave a negative response to the RSET that Exim sends to abort the
5808 transaction. The observed case was a dropped connection after DATA that had
5809 been sent in pipelining mode. That is, the server had advertised PIPELINING
5810 but was not implementing it correctly. The effect of the bug was incorrect
5811 behaviour, such as trying another host, and this could lead to a crash.
5817 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5818 and daemon.c were passed as pointers to ints; they should have been
5819 pointers to socklen_t variables (which are typically unsigned ints).
5821 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5824 3. Fixed a really odd bug that affected only the testing scheme; patching a
5825 certain fixed string in the binary changed the value of another string that
5826 happened to be identical to the end of the original first string.
5828 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5829 name", it returns that address as the IP address. On some operating
5830 systems (e.g. Solaris), it also passes back the IP address string as the
5831 "host name". However, on others (e.g. Linux), it passes back an empty
5832 string. Exim wasn't checking for this, and was changing the host name to an
5833 empty string, assuming it had been canonicalized.
5835 5. Although rare, it is permitted to have more than one PTR record for a given
5836 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5837 all the names associated with an address, because they do in Solaris.
5838 However, it seems that they do not in Linux for data that comes from the
5839 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5840 I found this out when I moved to a new Linux workstation and tried to run
5841 the Exim test suite.
5843 To get round this problem I have changed the code so that it now does its
5844 own call to the DNS to look up PTR records when searching for a host name.
5845 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5846 addresses that are only in /etc/hosts are still found.
5848 This behaviour is, however, controlled by an option called host_lookup_
5849 order, which defaults to "bydns:byaddr". If people want to use the other
5850 order, or indeed, just use one or the other means of lookup, they can
5851 specify it in this variable.
5853 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5854 some operating systems, this comes back from gethostbyaddr() as an empty
5855 string, and this is what Exim used to test for. However, it seems that in
5856 other systems, "." is yielded. Exim now tests for this case too.
5858 7. The values of check_spool_space and check_log_space are now held internally
5859 as a number of kilobytes instead of an absolute number of bytes. If a
5860 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5861 kilobyte. This means that much larger values can be stored.
5863 8. Exim monitor: an attempt to get the action menu when not actually pointing
5864 at a message produces an empty menu entitled "No message selected". This
5865 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5866 no entries in it ("Shell widget menu has zero width and/or height"). So I
5867 have added a single, blank menu entry in this case.
5869 9. Added ${quote_local_part.
5871 10. MIME decoding is now applied to the contents of Subject: header lines when
5874 11. Now that a reference to $sender_host_address automatically causes a reverse
5875 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5876 host lookup before query-style lookups in lists that might use this
5877 variable. This has therefore been abolished, and the "net-" prefix is no
5878 longer necessary for query-style lookups.
5880 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5881 have been a typo for LFLAGS, so it has been changed.
5883 13. The install script calls Exim with "-C /dev/null" in order to find the
5884 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5885 to be output. However, since Exim outputs its version number before the
5886 error, it didn't break the script. It just looked ugly. I fixed this by
5887 always allowing "-C /dev/null" if the caller is root.
5889 14. Ignore overlarge ACL variable number when reading spool file - insurance
5890 against a later release with more variables having written the file.
5892 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5893 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5894 this, as well as the form without the "IPv6" on the front.
5896 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5897 OS/Makefile-Darwin file.
5899 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5900 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5902 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5903 variable. However, some OS use ints for inodes. I've added cast to long int
5904 to get rid of the compiler warning.
5906 19. I had forgotten to lock out "/../" in configuration file names when
5907 ALT_CONFIG_PREFIX was set.
5909 20. Routers used for verification do not need to specify transports. However,
5910 if such a router generated a host list, and callout was configured, Exim
5911 crashed, because it could not find a port number from the (non-existent)
5912 transport. It now assumes port 25 in this circumstance.
5914 21. Added the -t option to exigrep.
5916 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5917 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5918 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5919 them is set, LOOKUP_LSEARCH is forced.
5921 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5923 24. Added sender and host information to the "rejected by local_scan()" log
5924 line; previously there was no indication of these.
5926 25. Added .include_if_exists.
5928 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5929 Linux. It turns out that not all file systems support this. Apparently some
5930 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5931 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5932 not supported on the file descriptor, is now ignored when Exim is trying to
5933 sync a directory. This applies only to Linux.
5935 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5937 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5938 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5941 29. Added support for maildirsize files from supplied patch (modified a bit).
5943 30. The use of :fail: followed by an empty string could lead Exim to respond to
5944 sender verification failures with (e.g.):
5946 550 Verification failed for <xxx>
5947 550 Sender verify failed
5949 where the first response line was missing the '-' that indicates it is not
5950 the final line of the response.
5952 31. The loop for finding the name of the user that called Exim had a hardwired
5953 limit of 10; it now uses the value of finduser_retries, which is used for
5954 all other user lookups.
5956 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5959 33. Exim was neglecting to zero errno before one call of strtol() when
5960 expanding a string and expecting an integer value. On some systems this
5961 resulted in spurious "integer overflow" errors. Also, it was casting the
5962 result into an int without checking.
5964 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5965 did not work. The code looks as if it has *never* worked, though it appears
5966 to have been documented since at least release 1.62. I have made it work.
5968 35. The "timeout_DNS" error in retry rules, also documented since at least
5969 1.62, also never worked. As it isn't clear exactly what this means, and
5970 clearly it isn't a major issue, I have abolished the feature by treating it
5971 as "timeout", and writing a warning to the main and panic logs.
5973 36. The display of retry rules for -brt wasn't always showing the error code
5976 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5977 timeout_connect_A, timeout_connect_MX.
5979 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5980 to the empty sender.
5982 39. The daemon was not analysing the content of -oX till after it had closed
5983 stderr and disconnected from the controlling terminal. This meant that any
5984 syntax errors were only noted on the panic log, and the return code from
5985 the command was 0. By re-arranging the code a little, I've made the
5986 decoding happen first, so such errors now appear on stderr, and the return
5987 code is 1. However, the actual setting up of the sockets still happens in
5988 the disconnected process, so errors there are still only recorded on the
5991 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5992 connections (as happens on some IP stacks) was logged at start up time as
5993 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5994 it from "IPv6 and IPv4", which means that two separate sockets are being
5997 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5998 says whether AF_INET or AF_INET6 was passed as an argument.
6000 42. Exiwhat output was messed up when time zones were included in log
6003 43. Exiwhat now gives more information about the daemon's listening ports,
6004 and whether -tls-on-connect was used.
6006 44. The "port" option of the smtp transport is now expanded.
6008 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6009 silently ignored. Now an error message is written to the main and panic
6012 46. There's a new ACL modifier called "logwrite" which writes to a log file
6013 as soon as it is encountered.
6015 47. Added $local_user_uid and $local_user_gid at routing time.
6017 48. Exim crashed when trying to verify a sender address that was being
6020 49. Exim was recognizing only a space character after ".include". It now also
6021 recognizes a tab character.
6023 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6024 extracting the relevant information from the specification. The man page no
6025 longer contains scrambled data for the -d option, and I've added a section
6026 at the front about calling Exim under different names.
6028 51. Added "extra_headers" argument to the "mail" command in filter files.
6030 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6033 53. Installed eximstats 1.29.
6035 54. Added transport_filter_timeout as a generic transport option.
6037 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6038 Cc: header lines. This was required by RFC 822, but it not required by RFC
6041 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6042 incoming messages that did not have them. Now it does so only if the
6043 message originates locally, that is, if there is no associated remote host
6044 address. When Resent- header lines are present, this applies to the Resent-
6045 lines rather than the non-Resent- lines.
6047 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6048 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6050 58. Messages for configuration errors now include the name of the main
6051 configuration file - useful now that there may be more than one file in a
6052 list (.included file names were always shown).
6054 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6055 for those rare installations that do not start the daemon as root or run it
6056 setuid root. I've cut out the call to initgroups() if the daemon is not
6059 60. The Exim user and group can now be bound into the binary as text strings
6060 that are looked up at the start of Exim's processing.
6062 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6064 62. Added $mailstore_basename variable.
6066 63. Installed patch to sieve.c from Michael Haardt.
6068 64. When Exim failed to open the panic log after failing to open the main log,
6069 the original message it was trying to log was written to stderr and debug
6070 output, but if they were not available (the usual case in production), it
6071 was lost. Now it is written to syslog before the two lines that record the
6072 failures to open the logs.
6074 65. Users' Exim filters run in subprocesses under the user's uid. It is
6075 possible for a "deliver" command or an alias in a "personal" command to
6076 provoke an address rewrite. If logging of address rewriting is configured,
6077 this fails because the process is not running as root or exim. There may be
6078 a better way of dealing with this, but for the moment (because 4.30 needs
6079 to be released), I have disabled address rewrite logging when running a
6080 filter in a non-root, non-exim process.
6086 1. The buildconfig auxiliary program wasn't quoting the value set for
6087 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6088 not defined. This bug was masked in 4.22 by the effect that was fixed in
6091 2. Some messages that were rejected after a message id was allocated were
6092 shown as "incomplete" by exigrep. It no longer does this for messages that
6093 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6095 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6096 have allow_domain_literals set, the ID did not get logged in the <= line.
6097 Domain literals are now always recognized in Message-ID: header lines.
6099 4. The first argument for a ${extract expansion item is the key name or field
6100 number. Leading and trailing spaces in this item were not being ignored,
6101 causing some misleading effects.
6103 5. When deliver_drop_privilege was set, single queue runner processes started
6104 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6105 same command in the process it spins off) were not dropping privilege.
6107 6. When the daemon running as "exim" started a queue runner, it always
6108 re-executed Exim in the spun-off process. This is a waste of effort when
6109 deliver_drop_privilege is set. The new process now just calls the
6110 queue-runner function directly.
6116 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6119 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6120 ignored. Though the use of -oP was forcing the writing of a pid file, it
6121 was always written to the default place.
6123 3. If the message "no IP address found for host xxxx" is generated during
6124 incoming verification, it is now followed by identification of the incoming
6125 connection (so you can more easily find what provoked it).
6127 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6129 5. Added some features to "harden" Exim a bit more against certain attacks:
6131 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6132 be put in Local/Makefile. This is like the never_users runtime option,
6133 but it cannot be overridden. The default setting is "root".
6135 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6136 prefix string with which any file named in a -C command line option
6139 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6140 is retained for -C and -D only if the caller of Exim is root. Without
6141 it, the exim user may also use -C and -D and retain privilege.
6143 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6144 command line option is disabled.
6146 6. Macro names set by the -D option must start with an upper case letter, just
6147 like macro names defined in the configuration file.
6149 7. Added "dereference=" facility to LDAP.
6151 8. Two instances of the typo "uknown" in the source files are fixed.
6153 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6154 the Configure-Makefile script screwed up while processing it.
6156 10. Incorporated PCRE 4.4.
6158 11. The SMTP synchronization check was not operating right at the start of an
6159 SMTP session. For example, it could not catch a HELO sent before the client
6160 waited for the greeting. There is now a check for outstanding input at the
6161 point when the greeting is written. Because of the duplex, asynchronous
6162 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6163 way, but not yet received, when the check is performed.
6165 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6166 on TCP/IP sockets, because this apparently causes some broken clients to
6169 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6170 unchanged) from the Cygwin maintainer.
6172 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6173 was supported instead of testing for mbx. Effectively a typo.
6175 15. The spa authenticator server code was not checking that the input it
6176 received was valid base64.
6178 16. The debug output line for the "set" modifier in ACLs was not showing the
6179 name of the variable that was being set.
6181 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6183 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6184 lookup on its own. Something else was needed to trigger the lookup. For
6185 example, a match in host_lookup or the need for a host name in a host list.
6186 Now, if $sender_host_name is referenced and the host name has not yet been
6187 looked up, a lookup is performed. If the lookup fails, the variable remains
6188 empty, and $host_lookup_failed is set to "1".
6190 19. Added "eqi" as a case-independent comparison operator.
6192 20. The saslauthd authentication condition could segfault if neither service
6193 nor realm was specified.
6195 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6196 error message that was logged was misleading, and incoming SMTP
6197 connections were dropped. The message is now more accurate, and temporary
6198 errors are given to SMTP connections.
6200 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6201 (except RSET) with a 421 error, until QUIT is received. However, it was
6202 failing to send a response to QUIT.
6204 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6205 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6206 ACL is now run after the helo_try_verify_hosts code.
6208 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6209 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6210 case-independent, so other case variants are also recognized). Apparently
6211 some systems use these upper case variants.
6213 25. If more than two messages were waiting for the same host, and a transport
6214 filter was specified for the transport, Exim sent two messages over the
6215 same TCP/IP connection, and then failed with "socket operation on non-
6216 socket" when it tried to send the third.
6218 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6220 27. The extern definition of crypt16() in expand.c was not being excluded when
6221 the OS had its own crypt16() function.
6223 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6224 as a preferred synonym for return_size_limit, both as an option and as an
6227 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6229 30. Changed the default configuration ACL to relax the local part checking rule
6230 for addresses that are not in any local domains. For these addresses,
6231 slashes and pipe symbols are allowed within local parts, but the sequence
6232 /../ is explicitly forbidden.
6234 31. SPA server authentication was not clearing the challenge buffer before
6237 32. log_message in a "warn" ACL statement was writing to the reject log as
6238 well as to the main log, which contradicts the documentation and doesn't
6239 seem right (because no rejection is happening). So I have stopped it.
6241 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6242 However, I am unable to do any testing of this.
6244 34. Fixed an infelicity in the appendfile transport. When checking directories
6245 for a mailbox, to see if any needed to be created, it was accidentally
6246 using path names with one or more superfluous leading slashes; tracing
6247 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6249 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6250 contents are added to the log line that is written for every discarded
6251 recipient. (Previously a log_message setting was ignored.)
6253 36. The ${quote: operator now quotes the string if it is empty.
6255 37. The install script runs exim in order to find its version number. If for
6256 some reason other than non-existence or emptiness, which it checks, it
6257 could not run './exim', it was installing it with an empty version number,
6258 i.e. as "exim-". This error state is now caught, and the installation is
6261 38. An argument was missing from the function that creates an error message
6262 when Exim fails to connect to the socket for saslauthd authentication.
6263 This could cause Exim to crash, or give a corrupted message.
6265 39. Added isip, isip4, and isip6 to ${if conditions.
6267 40. The ACL variables $acl_xx are now saved with the message, and can be
6268 accessed later in routers, transports, and filters.
6270 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6271 strings in the file are not string-expanded.
6273 42. If a MAIL command specified a SIZE value that was too large to fit into an
6274 int variable, the check against message_size_limit failed. Such values are
6275 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6276 day this will have to be increased, but I don't think I want to be around
6277 when emails are that large.
6284 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6285 iconv() is not standard in FreeBSD.
6287 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6288 IPv6 enabled. The observed symptom was a segmentation fault on return from
6289 the function os_common_find_running_interfaces() in src/os.c.
6291 3. In the check_special_case() function in daemon.c I had used "errno" as an
6292 argument name, which causes warnings on some systems. This was basically a
6293 typo, since it was named "eno" in the comments!
6295 4. The code that waits for the clock to tick (at a resolution of some fraction
6296 of a second) so as to ensure message-id uniqueness was always waiting for
6297 at least one whole tick, when it could have waited for less. [This is
6298 almost certainly not relevant at current processor speeds, where it is
6299 unlikely to ever wait at all. But we try to future-proof.]
6301 5. The function that sleeps for a time interval that includes fractions of a
6302 second contained a race. It did not block SIGALRM between setting the
6303 timer, and suspending (a couple of lines later). If the interval was short
6304 and the sigsuspend() was delayed until after it had expired, the suspension
6305 never ended. On busy systems this could lead to processes getting stuck for
6308 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6309 process, before it forks any delivery processes. The open lookup caching
6310 mechanism meant that the open file or database connection was passed into
6311 the delivery process. The problem was that delivery processes always tidy
6312 up cached lookup data. This could cause a problem for the next delivery
6313 process started by the queue runner, because the external queue runner
6314 process does not know about the closure. So the next delivery process
6315 still has data in the lookup cache. In the case of a file lookup, there was
6316 no problem because closing a file descriptor in a subprocess doesn't affect
6317 the parent. However, if the lookup was caching a connection to a database,
6318 the connection was closed, and the second delivery process was likely to
6319 see errors such as "PGSQL: query failed: server closed the connection
6320 unexpectedly". The problem has been fixed by closing all cached lookups
6321 in a queue runner before running a delivery process.
6323 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6324 seem to have the "const" qualifier which it has on other OS. I've
6327 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6328 *of the same type* (client or server) with the same public name that an
6329 error should be diagnosed.
6331 9. When Exim looked up a host name for an IP address, but failed to find the
6332 original IP address when looking up the host name (a safety check), it
6333 output the message "<ip address> does not match any IP for NULL", which was
6334 confusing, to say the least. The bug was that the host name should have
6335 appeared instead of "NULL".
6337 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6338 user that is built into the binary, and the -C or -D options is used, root
6339 privilege is dropped before the configuration file is read. In addition,
6340 logging is switched to stderr instead of the normal log files. If the
6341 configuration then re-defines the Exim user, the unprivileged environment
6342 is probably not what is expected, so Exim logs a panic warning message (but
6345 However, if deliver_drop_privilege is set, the unprivileged state may well
6346 be exactly what is intended, so the warning has been cut out in that case,
6347 and Exim is allowed to try to write to its normal log files.
6353 1. smtp_return_error_details was not giving details for temporary sender
6354 or receiver verification errors.
6356 2. Diagnose a configuration error if two authenticators have the same public
6359 3. Exim used not to create the message log file for a message until the first
6360 delivery attempt. This could be confusing when incoming messages were held
6361 for policy or load reasons. The message log file is now created at the time
6362 the message is received, and an initial "Received" line is written to it.
6364 4. The automatically generated man page for command line options had a minor
6365 bug that caused no ill effects; however, a more serious problem was that
6366 the procedure for building the man page automatically didn't always
6367 operate. Consequently, release 4.20 contains an out-of-date version. This
6368 shouldn't happen again.
6370 5. When building Exim with embedded Perl support, the script that builds the
6371 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6372 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6374 6. The freeze_tell option was not being used for messages that were frozen on
6375 arrival, either by an ACL or by local_scan().
6377 7. Added the smtp_incomplete_transaction log selector.
6379 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6380 was accepting AUTH without a new EHLO.
6382 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6383 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6386 10. Logging of TCP/IP connections (when configured) now happens in the main
6387 daemon process instead of the child process, so that the TCP/IP connection
6388 count is more accurate (but it can never be perfect).
6390 11. The use of "drop" in a nested ACL was not being handled correctly in the
6391 outer ACL. Now, if condition failure induced by the nested "drop" causes
6392 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6393 or "require"), the connection is dropped.
6395 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6396 that yield "discard" can now be used with an "accept" or a "discard" verb,
6397 but an error is generated for any others (because I can't see a useful way
6398 to define what should happen).
6400 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6401 are now processed in the same way as lines in the Exim configuration file.
6402 In particular, continuation lines are supported.
6404 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6406 15. Added -ti meaning -t -i.
6408 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6409 domains, and warn by logging if others are found.
6411 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6412 returned from GIFCONF when Exim is trying to find the list of interfaces on
6413 a host. The code in os.c has been modified to copy each ifreq to an aligned
6414 structure in all cases.
6416 Also, in some cases, the returned ifreq's were being copied to a 'struct
6417 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6418 means the last couple of bytes of an IPv6 address could be chopped if the
6419 ifreq contained only a normal sockaddr (14 bytes storage).
6421 18. Named domain lists were not supported in the hosts_treat_as_local option.
6422 An entry such as +xxxx was not recognized, and was treated as a literal
6425 19. Ensure that header lines added by a DATA ACL are included in the reject log
6426 if the ACL subsequently rejects the message.
6428 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6429 MD5 (which is deprecated).
6431 21. When testing a filter file using -bf, Exim was writing a message when it
6432 took the sender from a "From " line in the message, but it was not doing so
6433 when it took $return_path from a Return-Path: header line. It now does.
6435 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6436 with a valid header line field name (a series of printing characters
6437 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6439 23. Changed "disc" in the source to "disk" to conform to the documentation and
6440 the book and for uniformity.
6442 24. Ignore Sendmail's -Ooption=value command line item.
6444 25. When execve() failed while trying to run a command in a pipe transport,
6445 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6446 could be confused with a return value of 69 from the command itself. This
6447 has been changed to 127, the value the shell returns if it is asked to run
6448 a non-existent command. The wording for the related log line suggests a
6449 non-existent command as the problem.
6451 26. If received_header_text expands to an empty string, do not add a Received:
6452 header line to the message. (Well, it adds a token one on the spool, but
6453 marks it "old" so that it doesn't get used or transmitted.)
6455 27. Installed eximstats 1.28 (addition of -nt option).
6457 28. There was no check for failure on the call to getsockname() in the daemon
6458 code. This can fail if there is a shortage of resources on the system, with
6459 ENOMEM, for example. A temporary error is now given on failure.
6461 29. Contrary to the C standard, it seems that in some environments, the
6462 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6463 program. Exim now does this explicitly; it affects the formatting of
6464 timestamps using strftime().
6466 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6467 complaints. I've now initialized all the variables, to avoid this.
6469 32. Header lines added by a system filter were not being "seen" during
6470 transport-time rewrites.
6472 33. The info_callback() function passed to OpenSSL is set up with type void
6473 (*)(SSL *, int, int), as described somewhere. However, when calling the
6474 function (actually a macro) that sets it up, the type void(*)() is
6475 expected. I've put in a cast to prevent warnings from picky compilers.
6477 34. If a DNS black list lookup found a CNAME record, but there were no A
6478 records associated with the domain it pointed at, Exim crashed.
6480 35. If a DNS black list lookup returned more than one A record, Exim ignored
6481 all but the first. It now scans all returned addresses if a particular IP
6482 value is being sought. In this situation, the contents of the
6483 $dnslist_value variable are a list of all the addresses, separated by a
6486 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6487 to accept a host name and all its aliases if the forward lookup for any of
6488 them yielded the IP address of the incoming connection. Now it accepts only
6489 those names whose forward lookup yields the correct IP address. Any other
6490 names are discarded. This closes a loophole whereby a rogue DNS
6491 administrator could create reverse DNS records to break through a
6492 wildcarded host restriction in an ACL.
6494 37. If a user filter or a system filter that ran in a subprocess used any of
6495 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6496 the wrong values were passed to the pipe command ($thisaddress had the
6497 value of $0, $0 had the value of $1, etc). This bug was introduced by
6498 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6500 38. Improved the line breaking for long SMTP error messages from ACLs.
6501 Previously, if there was no break point between 40 and 75 characters, Exim
6502 left the rest of the message alone. Two changes have been made: (a) I've
6503 reduced the minimum length to 35 characters; (b) if it can't find a break
6504 point between 35 and 75 characters, it looks ahead and uses the first one
6505 that it finds. This may give the occasional overlong line, but at least the
6506 remaining text gets split now.
6508 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6509 file descriptors might be low, and that setting 1000 would always raise it.
6510 It turns out that in some environments, the limit is already over 1000 and
6511 that lowering it causes trouble. So now Exim takes care not to decrease it.
6513 40. When delivering a message, the value of $return_path is set to $sender_
6514 address at the start of routing (routers may change the value). By an
6515 oversight, this default was not being set up when an address was tested by
6516 -bt or -bv, which affected the outcome if any router or filter referred to
6519 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6520 log only when "message" or "log_message" are set. However, if one of the
6521 conditions was an address verification, or a call to a nested ACL, the
6522 messages generated by the underlying test were being passed through. This
6523 no longer happens. The underlying message is available in $acl_verify_
6524 message for both "message" and "log_message" expansions, so it can be
6525 passed through if needed.
6527 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6528 new expansion $bh_ to give the encoded byte string without charset
6529 translation. Translation happens only if iconv() is available; HAVE_ICONV
6530 indicates this at build time. HEADERS_CHARSET gives the charset to
6531 translate to; headers_charset can change it in the configuration, and
6532 "headers charset" can change it in an individual filter file.
6534 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6535 that creates RFC 2047 encoded "words" labels them as that charset instead
6536 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6537 expansion operator; (ii) when Exim creates a From: line for a local
6538 message; (iii) when a header line is rewritten to include a "phrase" part.
6540 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6541 buggy, causing it to skip the first lines of messages whose message ID
6542 ended in 'D'. This would not have bitten before Exim release 4.14, because
6543 message IDs were unlikely to end in 'D' before then. The effect was to have
6544 incorrect size information for certain domains.
6546 45. #include "config.h" was missing at the start of the crypt16.c module. This
6547 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6550 46. If there was a timeout during a "random" callout check, Exim treated it as
6551 a failure of the random address, and carried on sending RSET and the real
6552 address. If the delay was just some slowness somewhere, the response to the
6553 original RCPT would be taken as a response to RSET and so on, causing
6554 mayhem of various kinds.
6556 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6557 when I implemented it. It didn't allow for the fact that some option values
6558 may legitimately be negative (e.g. size_addition), and it didn't even do
6559 the right test for positive values.
6561 48. Domain names in DNS records are case-independent. Exim always looks them up
6562 in lower case. Some resolvers return domain names in exactly the case they
6563 appear in the zone file, that is, they may contain uppercase letters. Not
6564 all resolvers do this - some return always lower case. Exim was treating a
6565 change of case by a resolver as a change of domain, similar to a widening
6566 of a domain abbreviation. This triggered its re-routing code and so it was
6567 trying to route what was effectively the same domain again. This normally
6568 caused routing to fail (because the router wouldn't handle the domain
6569 twice). Now Exim checks for this case specially, and just changes the
6570 casing of the domain that it ultimately uses when it transmits the message
6573 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6576 50. If a filter generated a file delivery with a non-absolute name (possible if
6577 no home directory exists for the router), the forbid_file option was not
6580 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6581 the existing equality matching.
6583 52. Exim was using ints instead of ino_t variables in some places where it was
6584 dealing with inode numbers.
6586 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6587 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6588 called TMPDIR, and if it finds it is different, it changes its value.
6590 54. The smtp_printf() function is now made available to local_scan() so
6591 additional output lines can be written before returning. There is also an
6592 smtp_fflush() function to enable the detection of a dropped connection.
6593 The variables smtp_input and smtp_batched_input are exported to
6596 55. Changed the default runtime configuration: the message "Unknown user"
6597 has been removed from the ACL, and instead placed on the localuser router,
6598 using the cannot_route_message feature. This means that any verification
6599 failures that generate their own messages won't get overridden. Similarly,
6600 the "Unrouteable address" message that was in the ACL for unverifiable
6601 relay addresses has also been removed.
6603 56. Added hosts_avoid_esmtp to the smtp transport.
6605 57. The exicyclog script was not checking for the esoteric option
6606 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6607 will work only if exicyclog is run under the appropriate euid.
6609 58. Following a discussion on the list, the rules by which Exim recognises line
6610 endings on incoming messages have been changed. The -dropcr and drop_cr
6611 options are now no-ops, retained only for backwards compatibility. The
6612 following line terminators are recognized: LF CRLF CR. However, special
6613 processing applies to CR:
6615 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6616 nor a local message in the state where . is a terminator.
6618 (ii) If a bare CR is encountered in a header line, an extra space is added
6619 after the line terminator so as not to end the header. The reasoning
6620 behind this is that bare CRs in header lines are most likely either
6621 to be mistakes, or people trying to play silly games.
6623 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6624 was being incorrectly given as 18 bytes larger than it should have been.
6625 This is a VOB (very old bug).
6627 60. This may never have affected anything current, but just in case it has:
6628 When the local host is found other than at the start of a list of hosts,
6629 the local host, those with the same MX, and any that follow, are discarded.
6630 When the list in question was part of a longer list of hosts, the following
6631 hosts (not currently being processed) were also being discarded. This no
6632 longer happens. I'm not sure if this situation could ever has previously
6635 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6638 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6639 header. This is something that is recommended in a new Internet Draft, and
6640 is something that is documented as being done by Sendmail. There are two
6641 possible values. For messages generated by the autoreply transport, Exim
6644 Auto-Submitted: auto-replied
6646 whereas for all other generated messages (e.g. bounces) it adds
6648 Auto-Submitted: auto-generated
6650 63. The "personal" condition in filters now includes a test for the
6651 Auto-Submitted: header. If it contains the string "auto-" the message it
6652 not considered personal.
6654 64. Added rcpt_include_affixes as a generic transport option.
6656 65. Added queue_only_override (default true).
6658 66. Added the syslog_duplication option.
6660 67. If what should have been the first header line of a message consisted of
6661 a space followed by a colon, Exim was mis-interpreting it as a header line.
6662 It isn't of course - it is syntactically invalid and should therefore be
6663 treated as the start of the message body. The misbehaviour could have
6664 caused a number of strange effects, including loss of data in subsequent
6665 header lines, and spool format errors.
6667 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6668 client host had authenticated. This control can now be exercised by an ACL
6669 for more flexibility.
6671 69. By default, callouts do not happen when testing with -bh. There is now a
6672 variant, -bhc, which does actually run the callout code, including
6673 consulting and updating the callout cache.
6675 70. Added support for saslauthd authentication, courtesy of Alexander
6678 71. If statvfs() failed on the spool or log directories while checking their
6679 size for availability, Exim confusingly gave the error "space shortage".
6680 Furthermore, in debugging mode it crashed with a floating point exception.
6681 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6682 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6683 serious problem, Exim now writes to the main and panic logs when this
6684 happens, with details of the failure. It then refuses to accept the
6685 incoming message, giving the message "spool directory problem" or "log
6686 directory problem" with a 421 code for SMTP messages.
6688 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6689 0, 1, and 2 exist, because some OS complain for execs without them (see
6690 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6691 descriptors. However, the code omitted to check that the open succeeded,
6692 causing mysterious errors if for some reason the permissions on /dev/null
6693 got screwed. Now Exim writes a message to the main and panic logs, and
6694 bombs out if it can't open /dev/null.
6696 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6697 interact so that it is all more flexible. It is supposed to remain
6698 backwards compatible. Also added extra_local_interfaces.
6700 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6701 to bomb out with an assertion failure - to the client this appears as a
6702 connection drop. This problem occurs in the part of the code that was taken
6703 from the Samba project. Fortunately, the assertion is in a very simple
6704 function, so I have fixed this by reproducing the function inline in the
6705 one place where it is called, and arranging for authentication to fail
6706 instead of killing the process with assert().
6708 75. The SPA client code was not working when the server requested OEM rather
6709 than Unicode encoding.
6711 76. Added code to make require_files with a specific uid setting more usable in
6712 the case where statting the file as root fails - usually a non-root-mounted
6713 NFS file system. When this happens and the failure is EACCES, Exim now
6714 forks a subprocess and does the per-uid checking as the relevant uid.
6716 77. Added process_log_path.
6718 78. If log_file_path was not explicitly set, a setting of check_log_space or
6719 check_log_inodes was ignored.
6721 79. If a space check for the spool or log partitions fails, the incident is now
6722 logged. Of course, in the latter case the data may get lost...
6724 80. Added the %p formatting code to string_format() so that it can be used to
6725 print addresses in debug_print(). Adjusted all the address printing in the
6726 debugging in store.c to use %p rather than %d.
6728 81. There was a concern that a line of code in smtp_in.c could overflow a
6729 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6730 initially expressed, the concern was not well-founded, because trailing
6731 spaces are removed early. However, if the trailing spaces were followed by
6732 a NULL, they did not get removed, so the overflow was possible. Two fixes
6735 (a) I re-wrote the offending code in a cleaner fashion.
6736 (b) If an incoming SMTP command contains a NULL character, it is rejected
6739 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6740 runs initgroups(), so that if the Exim user is in any additional groups,
6741 they will be used during message reception.
6747 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob