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 JH/17 Rework error string handling in TLS interface so that the caller in
93 more cases is responsible for logging. This permits library-sourced
94 string to be attached to addresses during delivery, and collapses
95 pairs of long lines into single ones.
97 PP/06 Drop '.' from @INC in various Perl scripts.
99 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
101 PP/08 Reduce a number of compilation warnings under clang; building with
102 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
103 should be warning-free.
109 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
110 supports it and a size is available (ie. the sending peer gave us one).
112 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
113 years of being deprecated). The replacements are the ACLs
114 acl_smtp_mime and acl_not_smtp_mime.
116 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
117 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
118 or even an in-clear connection were permitted. Now, if the host lookup
119 was dnssec and dane was requested then the host is only used if the
120 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
121 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
122 if one fails this test.
123 This means that a poorly-configured remote DNS will make it incommunicado;
124 but it protects against a DNS-interception attack on it.
126 JH/04 Bug 1810: make continued-use of an open smtp transport connection
127 non-noisy when a race steals the message being considered.
129 JH/05 If main configuration option tls_certificate is unset, generate a
130 self-signed certificate for inbound TLS connections.
132 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
133 in rewrites and routers.
135 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
136 and logged a warning sing 4.83; now they are a configuration file error.
138 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
139 (lacking @domain). Apply the same qualification processing as RCPT.
141 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
143 JH/10 Support ${sha256:} applied to a string (as well as the previous
146 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
147 a cutthrough deliver is pending, as we always want to make a connection.
148 This also avoids re-routing the message when later placing the cutthrough
149 connection after a verify cache hit.
150 Do not update it with the verify result either.
152 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
153 when routing results in more than one destination address.
155 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
156 signing (which inhibits the cutthrough capability). Previously only
157 the presence of an option was tested; now an expansion evaluating as
158 empty is permissible (obviously it should depend only on data available
159 when the cutthrough connection is made).
161 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
162 the relevant preceding SMTP command did not note the pipelining mode.
164 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
165 Previously they were not counted.
167 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
168 as one having no matching records. Previously we deferred the message
169 that needed the lookup.
171 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
172 distinguished as "(=".
174 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
175 for missing MX records. Previously it only worked for missing A records.
177 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
179 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
180 after the data-go-ahead and data-ack. Patch from Jason Betts.
182 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
183 even for a "none" policy. Patch from Tony Meyer.
185 JH/22 Fix continued use of a connection for further deliveries. If a port was
186 specified by a router, it must also match for the delivery to be
189 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
190 When one of the recipients of a message was unsuitable for the connection
191 (has no matching addresses), we lost track of needing to mark it
192 deferred. As a result mail would be lost.
194 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
196 JH/25 Decoding ACL controls is now done using a binary search; the source code
197 takes up less space and should be simpler to maintain. Merge the ACL
198 condition decode tables also, with similar effect.
200 JH/26 Fix problem with one_time used on a redirect router which returned the
201 parent address unchanged. A retry would see the parent address marked as
202 delivered, so not attempt the (identical) child. As a result mail would
205 JH/27 Fix a possible security hole, wherein a process operating with the Exim
206 UID can gain a root shell. Credit to http://www.halfdog.net/ for
207 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
210 JH/28 Enable {spool,log} filesystem space and inode checks as default.
211 Main config options check_{log,spool}_{inodes,space} are now
212 100 inodes, 10MB unless set otherwise in the configuration.
214 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
215 Previously it only applied to the main-section connection policy
218 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
220 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
221 by me. Added RFC7919 DH primes as an alternative.
223 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
224 are not in the system include path.
226 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
227 GnuTLS, when a session startup failed (eg because the client disconnected)
228 Exim did stdio operations after fclose. This was exposed by a recent
229 change which nulled out the file handle after the fclose.
231 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
232 signed directly by the cert-signing cert, rather than an intermediate
233 OCSP-signing cert. This is the model used by LetsEncrypt.
235 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
237 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
238 an incoming connection.
240 HS/02 Bug 1802: Do not half-close the connection after sending a request
243 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
244 fallback to "prime256v1".
246 JH/34 SECURITY: Use proper copy of DATA command in error message.
247 Could leak key material. Remotely exploitable. CVE-2016-9963.
253 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
254 and 3.4.4 - once the server is enabled to respond to an OCSP request
255 it does even when not requested, resulting in a stapling non-aware
256 client dropping the TLS connection.
258 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
259 support variable-length bit vectors. No functional change.
261 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
262 The I= interface field on outgoing lines is now after the H= remote
263 host field, same as incoming lines. There is a separate
264 outgoing_interface log selector which allows you to disable the
267 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
268 If not running log_selector +smtp_connection the mainlog would be held
269 open indefinitely after a "too many connections" event, including to a
270 deleted file after a log rotate. Leave the per net connection logging
271 leaving it open for efficiency as that will be quickly detected by the
272 check on the next write.
274 HS/01 Bug 1671: Fix post transport crash.
275 Processing the wait-<transport> messages could crash the delivery
276 process if the message IDs didn't exist for some reason. When
277 using 'split_spool_directory=yes' the construction of the spool
278 file name failed already, exposing the same netto behaviour.
280 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
281 mime_regex ACL conditions.
283 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
284 to DSN fail messages (bounces): remote IP, remote greeting, remote response
285 to HELO, local diagnostic string.
287 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
288 log line to debug. Even when configured with a tls authenticator many
289 client connections are expected to not authenticate in this way, so
290 an authenticate fail is not an error.
292 HS/02 Add the Exim version string to the process info. This way exiwhat
293 gives some more detail about the running daemon.
295 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
296 matter for fast-change records such as DNSBLs.
298 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
299 retry record, even if constant. There may be multiple transports with
300 different interface settings and the retry behaviour needs to be kept
303 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
305 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
307 JH/10 Bug 840: fix log_defer_output option of pipe transport
309 JH/11 Bug 830: use same host for all RCPTS of a message, even under
310 hosts_randomize. This matters a lot when combined with mua_wrapper.
312 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
313 ${quote_pgsql:<string>} operator.
315 JH/13 Bug 1708: avoid misaligned access in cached lookup.
317 JH/14 Change header file name for freeradius-client. Relevant if compiling
318 with Radius support; from the Gentoo tree and checked under Fedora.
320 JH/15 Bug 1712: Introduce $prdr_requested flag variable
322 JH/16 Bug 1714: Permit an empty string as expansion result for transport
323 option transport_filter, meaning no filtering.
325 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
327 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
328 defaults to "*" (all hosts). The variable is now available when not built
329 with TLS, default unset, mainly to enable keeping the testsuite sane.
330 If a server certificate is not supplied (via tls_certificate) an error is
331 logged, and clients will find TLS connections fail on startup. Presumably
332 they will retry in-clear.
333 Packagers of Exim are strongly encouraged to create a server certificate
334 at installation time.
336 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
337 with the $config_file variable.
339 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
340 in transport context, after the attempt, and per-recipient. The latter type
341 is per host attempted. The event data is the error message, and the errno
342 information encodes the lookup type (A vs. MX) used for the (first) host,
343 and the trailing two digits of the smtp 4xx response.
345 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
346 to write to mainlog (or rejectlog, paniclog) in the window between file
347 creation and permissions/ownership being changed. Particularly affects
348 installations where exicyclog is run as root, rather than exim user;
349 result is that the running daemon panics and dies.
351 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
353 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
354 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
355 "pri" and "weight". Note that the previous implicit priority given by the
356 list order is no longer honoured.
358 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
361 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
362 by defining SUPPORT_SOCKS.
364 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
365 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
366 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
367 variables are renamed to proxy_{local,external}_{address,port}.
369 JH/27 Move Internationalisation support from Experimental to mainline, enabled
370 for a build by defining SUPPORT_I18N
372 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
373 of the query string, and make ${quote_redis:} do that quoting.
375 JH/29 Move Events support from Experimental to mainline, enabled by default
376 and removable for a build by defining DISABLE_EVENT.
378 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
380 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
381 cached by the daemon.
383 JH/32 Move Redis support from Experimental to mainline, enabled for a build
384 by defining LOOKUP_REDIS. The libhiredis library is required.
386 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
387 keys are given for lookup.
389 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
390 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
391 only supported when built with TLS support. The PolarSSL SHA routines
392 are still used when the TLS library is too old for convenient support.
394 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
395 openssl_options), for security. OpenSSL forces this from version 1.1.0
396 server-side so match that on older versions.
398 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
399 allocation for $value could be released as the expansion processing
400 concluded, but leaving the global pointer active for it.
402 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
403 and to use the domains and local_parts ACL conditions.
405 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
406 incorrectly not doubled on cutthrough transmission, hence seen as a
407 body-termination at the receiving system - resulting in truncated mails.
408 Commonly the sender saw a TCP-level error, and retransmitted the message
409 via the normal store-and-forward channel. This could result in duplicates
410 received - but deduplicating mailstores were liable to retain only the
411 initial truncated version.
413 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
415 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
417 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
418 we're in there, support oversigning also; bug 1309.
420 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
422 HS/04 Add support for keep_environment and add_environment options.
424 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
425 either intentional arithmetic overflow during PRNG, or testing config-
428 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
429 delivery resulted in actual delivery. Cancel cutthrough before DATA
432 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
433 rejects a recipient: pass the reject to the originator.
435 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
436 Many were false-positives and ignorable, but it's worth fixing the
439 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
440 for the new environment-manipulation done at startup. Move the routines
441 from being local to tls.c to being global via the os.c file.
443 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
444 an extract embedded as result-arg for a map, the first arg for extract
445 is unavailable so we cannot tell if this is a numbered or keyed
446 extraction. Accept either.
452 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
455 JH/02 The smtp transport option "multi_domain" is now expanded.
457 JH/03 The smtp transport now requests PRDR by default, if the server offers
460 JH/04 Certificate name checking on server certificates, when exim is a client,
461 is now done by default. The transport option tls_verify_cert_hostnames
462 can be used to disable this per-host. The build option
463 EXPERIMENTAL_CERTNAMES is withdrawn.
465 JH/05 The value of the tls_verify_certificates smtp transport and main options
466 default to the word "system" to access the system default CA bundle.
467 For GnuTLS, only version 3.0.20 or later.
469 JH/06 Verification of the server certificate for a TLS connection is now tried
470 (but not required) by default. The verification status is now logged by
471 default, for both outbound TLS and client-certificate supplying inbound
474 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
477 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
478 Status Notification (bounce) messages are now MIME format per RFC 3464.
479 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
480 under the control of the dsn_advertise_hosts option, and routers may
481 have a dsn_lasthop option.
483 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
484 default, modifiable by a malware= option. The list separator for
485 the options can now be changed in the usual way. Bug 68.
487 JH/10 The smtp_receive_timeout main option is now expanded before use.
489 JH/11 The incoming_interface log option now also enables logging of the
490 local interface on delivery outgoing connections.
492 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
493 if the interface and destination host and port all match.
495 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
498 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
499 Patch from Andrew Lewis.
501 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
502 now supports optional time-restrictions, weighting, and priority
503 modifiers per server. Patch originally by <rommer@active.by>.
505 JH/16 The spamd_address main option now supports a mixed list of local
506 and remote servers. Remote servers can be IPv6 addresses, and
507 specify a port-range.
509 JH/17 Bug 68: The spamd_address main option now supports an optional
510 timeout value per server.
512 JH/18 Bug 1581: Router and transport options headers_add/remove can
513 now have the list separator specified.
515 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
518 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
521 JH/21 Support for the A6 type of dns record is withdrawn.
523 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
524 rather than the verbs used.
526 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
527 from 255 to 1024 chars.
529 JH/24 Verification callouts now attempt to use TLS by default.
531 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
532 are generic router options now. The defaults didn't change.
534 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
535 Original patch from Alexander Shikoff, worked over by JH.
537 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
538 files if ZCAT_COMMAND is not executable.
540 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
542 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
544 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
545 Normally benign, it bites when the pair was led to by a CNAME;
546 modern usage is to not canonicalize the domain to a CNAME target
547 (and we were inconsistent anyway for A-only vs AAAA+A).
549 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
551 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
552 when evaluating $sender_host_dnssec.
554 JH/31 Check the HELO verification lookup for DNSSEC, adding new
555 $sender_helo_dnssec variable.
557 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
559 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
561 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
563 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
564 documented as working, but never had. Support all but $spam_report.
566 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
567 added for tls authenticator.
569 HS/03 Add perl_taintmode main config option
575 TL/01 When running the test suite, the README says that variables such as
576 no_msglog_check are global and can be placed anywhere in a specific
577 test's script, however it was observed that placement needed to be near
578 the beginning for it to behave that way. Changed the runtest perl
579 script to read through the entire script once to detect and set these
580 variables, reset to the beginning of the script, and then run through
581 the script parsing/test process like normal.
583 TL/02 The BSD's have an arc4random API. One of the functions to induce
584 adding randomness was arc4random_stir(), but it has been removed in
585 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
586 function when detected.
588 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
589 cause callback expansion.
591 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
592 syntax errors in an expansion can be treated as a string instead of
593 logging or causing an error, due to the internal use of bool_lax
594 instead of bool when processing it.
596 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
597 server certificates when making smtp deliveries.
599 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
601 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
603 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
605 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
606 Merged patch from Sebastian Wiedenroth.
608 JH/05 Fix results-pipe from transport process. Several recipients, combined
609 with certificate use, exposed issues where response data items split
610 over buffer boundaries were not parsed properly. This eventually
611 resulted in duplicates being sent. This issue only became common enough
612 to notice due to the introduction of connection certificate information,
613 the item size being so much larger. Found and fixed by Wolfgang Breyha.
615 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
616 size buffer was used, resulting in syntax errors when an expansion
619 JH/07 Add support for directories of certificates when compiled with a GnuTLS
620 version 3.3.6 or later.
622 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
623 is EXPERIMENTAL_EVENT, the main-configuration and transport options
624 both become "event_action", the variables become $event_name, $event_data
625 and $event_defer_errno. There is a new variable $verify_mode, usable in
626 routers, transports and related events. The tls:cert event is now also
627 raised for inbound connections, if the main configuration event_action
630 TL/06 In test suite, disable OCSP for old versions of openssl which contained
631 early OCSP support, but no stapling (appears to be less than 1.0.0).
633 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
634 server certificate names available under the smtp transport option
635 "tls_verify_cert_hostname" now do not permit multi-component wildcard
638 JH/10 Time-related extraction expansions from certificates now use the main
639 option "timezone" setting for output formatting, and are consistent
640 between OpenSSL and GnuTLS compilations. Bug 1541.
642 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
643 encoded parameter in the incoming message. Bug 1558.
645 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
646 include certificate info, eximon was claiming there were spoolfile
649 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
651 JH/14 Log delivery-related information more consistently, using the sequence
652 "H=<name> [<ip>]" wherever possible.
654 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
655 are problematic for Debian distribution, omit them from the release
658 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
660 JH/16 Fix string representation of time values on 64bit time_t architectures.
663 JH/17 Fix a null-indirection in certextract expansions when a nondefault
664 output list separator was used.
669 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
670 checkers that were complaining about end of non-void function with no
673 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
674 This was a regression introduced in 4.83 by another bugfix.
676 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
678 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
679 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
685 TF/01 Correctly close the server side of TLS when forking for delivery.
687 When a message was received over SMTP with TLS, Exim failed to clear up
688 the incoming connection properly after forking off the child process to
689 deliver the message. In some situations the subsequent outgoing
690 delivery connection happened to have the same fd number as the incoming
691 connection previously had. Exim would try to use TLS and fail, logging
692 a "Bad file descriptor" error.
694 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
695 utilities have not been installed.
697 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
698 temporary space as the ACL may create new global variables.
700 TL/01 LDAP support uses per connection or global context settings, depending
701 upon the detected version of the libraries at build time.
703 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
704 to extract and use the src ip:port in logging and expansions as if it
705 were a direct connection from the outside internet. PPv2 support was
706 updated based on HAProxy spec change in May 2014.
708 JH/02 Add ${listextract {number}{list}{success}{fail}}.
710 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
711 Properly escape header and check for NULL return.
713 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
716 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
718 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
719 characters in header names, implemented as a verify condition.
720 Contributed by Michael Fischer v. Mollard.
722 TL/05 Rename SPF condition results err_perm and err_temp to standardized
723 results permerror and temperror. Previous values are deprecated but
724 still accepted. In a future release, err_perm and err_temp will be
725 completely removed, which will be a backward incompatibility if the
726 ACL tests for either of these two old results. Patch contributed by
727 user bes-internal on the mailing list.
729 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
731 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
732 selectors, in both main and reject logs.
734 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
737 JH/07 Add malware type "sock" for talking to simple daemon.
739 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
741 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
742 routers/transports under cutthrough routing.
744 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
745 numbers. Touch up "bool" conditional to keep the same definition.
747 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
749 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
751 JH/12 Expand items in router/transport headers_add or headers_remove lists
752 individually rather than the list as a whole. Bug 1452.
754 Required for reasonable handling of multiple headers_ options when
755 they may be empty; requires that headers_remove items with embedded
756 colons must have them doubled (or the list-separator changed).
758 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
759 view the policy declared in the DMARC record. Currently, $dmarc_status
760 is a combined value of both the record presence and the result of the
763 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
765 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
766 dnslookup router and the smtp transport (applying to the forward
769 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
770 of ldap servers used for a specific lookup. Patch provided by Heiko
773 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
774 New variable $lookup_dnssec_authenticated for observability.
776 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
777 Patch submitted by Lars Timman.
779 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
781 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
782 Requires trusted mode and valid format message id, aborts otherwise.
783 Patch contributed by Heiko Schlichting.
785 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
786 certextract with support for various fields. Bug 1358.
788 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
789 is requested by default, modifiable by smtp transport option
792 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
793 operate on certificate variables to give certificate fingerprints
794 Also new ${sha256:cert_variable}.
796 JH/23 The PRDR feature is moved from being Experimental into the mainline.
798 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
799 Christian Aistleitner.
801 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
803 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
804 file. Patch from Wolfgang Breyha.
806 JH/25 Expand the coverage of the delivery $host and $host_address to
807 client authenticators run in verify callout. Bug 1476.
809 JH/26 Port service names are now accepted for tls_on_connect_ports, to
810 align with daemon_smtp_ports. Bug 72.
812 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
813 support and error reporting did not work properly.
815 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
816 and is readable. Patch from Andrew Colin Kissa.
818 TL/14 Enhance documentation of ${run expansion and how it parses the
819 commandline after expansion, particularly in the case when an
820 unquoted variable expansion results in an empty value.
822 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
824 PP/02 Fix internal collision of T_APL on systems which support RFC3123
825 by renaming away from it. Addresses GH issue 15, reported by
828 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
830 TL/15 SECURITY: prevent double expansion in math comparison functions
831 (can expand unsanitized data). Not remotely exploitable.
838 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
840 PP/02 Make -n do something, by making it not do something.
841 When combined with -bP, the name of an option is not output.
843 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
846 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
847 $sender_host_name and config options to manage this, and basic check
850 PP/05 DSCP support for outbound connections and control modifier for inbound.
852 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
853 (Only plugin which currently uses this is kerberos4, which nobody should
854 be using, but we should make it available and other future plugins might
855 conceivably use it, even though it would break NAT; stuff *should* be
856 using channel bindings instead).
858 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
859 name; added for Sendmail compatibility; requires admin caller.
860 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
861 just ignore it); requires trusted caller.
862 Also parse but ignore: -Ac -Am -X<logfile>
865 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
867 TL/02 Add +smtp_confirmation as a default logging option.
869 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
870 Patch by Magnus Holmgren from 2007-02-20.
872 TL/04 Bugzilla 1281 - Spec typo.
873 Bugzilla 1283 - Spec typo.
874 Bugzilla 1290 - Spec grammar fixes.
876 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
878 TL/06 Add Experimental DMARC support using libopendmarc libraries.
880 TL/07 Fix an out of order global option causing a segfault. Reported to dev
881 mailing list by by Dmitry Isaikin.
883 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
885 JH/02 Support "G" suffix to numbers in ${if comparisons.
887 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
889 NM/01 Bugzilla 1197 - Spec typo
890 Bugzilla 1196 - Spec examples corrections
892 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
894 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
895 gnutls_enable_pkcs11, but renamed to more accurately indicate its
898 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
899 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
901 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
902 "acl {{name}{arg}...}", and optional args on acl condition
905 JH/05 Permit multiple router/transport headers_add/remove lines.
907 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
909 JH/07 Avoid using a waiting database for a single-message-only transport.
910 Performance patch from Paul Fisher. Bugzilla 1262.
912 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
915 JH/09 Add $headers_added variable, with content from use of ACL modifier
916 add_header (but not yet added to the message). Bugzilla 199.
918 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
919 Pulled from Bugzilla 817 by Wolfgang Breyha.
921 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
923 (nb: this is the same fix as in Exim 4.80.1)
925 JH/11 Add A= logging on delivery lines, and a client_set_id option on
928 JH/12 Add optional authenticated_sender logging to A= and a log_selector
931 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
933 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
934 advertise SMTP AUTH mechanism to us, instead of a generic
935 protocol violation error. Also, make Exim more robust to bad
936 data from the Dovecot auth socket.
938 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
940 When a queue runner is handling a message, Exim first routes the
941 recipient addresses, during which it prunes them based on the retry
942 hints database. After that it attempts to deliver the message to
943 any remaining recipients. It then updates the hints database using
946 So if a recipient address works intermittently, it can get repeatedly
947 deferred at routing time. The retry hints record remains fresh so the
948 address never reaches the final cutoff time.
950 This is a fairly common occurrence when a user is bumping up against
951 their storage quota. Exim had some logic in its local delivery code
952 to deal with this. However it did not apply to per-recipient defers
953 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
955 This change adds a proper retry rule check during routing so that the
956 final cutoff time is checked against the message's age. We only do
957 this check if there is an address retry record and there is not a
958 domain retry record; this implies that previous attempts to handle
959 the address had the retry_use_local_parts option turned on. We use
960 this as an approximation for the destination being like a local
961 delivery, as in LMTP.
963 I suspect this new check makes the old local delivery cutoff check
964 redundant, but I have not verified this so I left the code in place.
966 TF/02 Correct gecos expansion when From: is a prefix of the username.
968 Test 0254 submits a message to Exim with the header
972 When I ran the test suite under the user fanf2, Exim expanded
973 the header to contain my full name, whereas it should have added
974 a Resent-Sender: header. It erroneously treats any prefix of the
975 username as equal to the username.
977 This change corrects that bug.
979 GF/01 DCC debug and logging tidyup
980 Error conditions log to paniclog rather than rejectlog.
981 Debug lines prefixed by "DCC: " to remove any ambiguity.
983 TF/03 Avoid unnecessary rebuilds of lookup-related code.
985 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
986 Bug spotted by Jeremy Harris; was flawed since initial commit.
987 Would have resulted in OCSP responses post-SNI triggering an Exim
988 NULL dereference and crash.
990 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
992 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
993 Bug detection, analysis and fix by Samuel Thibault.
994 Bugzilla 1331, Debian bug #698092.
996 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
998 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
999 Server implementation by Todd Lyons, client by JH.
1000 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1001 config variable "prdr_enable" controls whether the server
1002 advertises the facility. If the client requests PRDR a new
1003 acl_data_smtp_prdr ACL is called once for each recipient, after
1004 the body content is received and before the acl_smtp_data ACL.
1005 The client is controlled by both of: a hosts_try_prdr option
1006 on the smtp transport, and the server advertisement.
1007 Default client logging of deliveries and rejections involving
1008 PRDR are flagged with the string "PRDR".
1010 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1011 fclose(). Diagnosis by Todd Lyons.
1013 PP/17 Update configure.default to handle IPv6 localhost better.
1014 Patch by Alain Williams (plus minor tweaks).
1017 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1018 This is now consistent with GnuTLS, and is now documented: the
1019 previous undocumented portable approach to treating the option as
1020 unset was to force an expansion failure. That still works, and
1021 an empty string is now equivalent.
1023 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1024 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1025 not performing validation itself.
1027 PP/20 Added force_command boolean option to pipe transport.
1028 Patch from Nick Koston, of cPanel Inc.
1030 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1033 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1035 PP/21 Fix eximon continuous updating with timestamped log-files.
1036 Broken in a format-string cleanup in 4.80, missed when I repaired the
1037 other false fix of the same issue.
1038 Report and fix from Heiko Schlichting.
1041 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1042 Report from Prashanth Katuri.
1044 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1045 It's SecureTransport, so affects any MacOS clients which use the
1046 system-integrated TLS libraries, including email clients.
1048 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1049 using a MIME ACL for non-SMTP local injection.
1050 Report and assistance in diagnosis by Warren Baker.
1052 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1054 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1056 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1057 last id that failed so it may be referenced in subsequent ACL's.
1059 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1062 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1063 ldap library initialization, allowing self-signed CA's to be
1064 used. Also properly sets require_cert option later in code by
1065 using NULL (global ldap config) instead of ldap handle (per
1066 session). Bug diagnosis and testing by alxgomz.
1068 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1069 the src/util/ subdirectory.
1071 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1072 renamed to Transport Post Delivery Action by Jeremy Harris, as
1075 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1076 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1077 redis_servers = needs to be configured which will be used by the redis
1078 lookup. Patch from Warren Baker, of The Packet Hub.
1080 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1082 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1083 hostname or reverse DNS when processing a host list. Used suggestions
1084 from multiple comments on this bug.
1086 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1088 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1089 Missed a few lines, added it to make the runtest require no keyboard
1092 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1093 contains upper case chars. Make router use caseful_local_part.
1095 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1096 support when GnuTLS has been built with p11-kit.
1102 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1104 This, or similar/improved, will also be change PP/11 of 4.82.
1110 PP/01 Handle short writes when writing local log-files.
1111 In practice, only affects FreeBSD (8 onwards).
1112 Bugzilla 1053, with thanks to Dmitry Isaikin.
1114 NM/01 Bugzilla 949 - Documentation tweak
1116 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1119 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1121 PP/02 Implemented gsasl authenticator.
1123 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1125 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1126 `pkg-config foo` for cflags/libs.
1128 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1129 with rest of GSASL and with heimdal_gssapi.
1131 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1132 `pkg-config foo` for cflags/libs for the TLS implementation.
1134 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1135 properties get this fed in as external SSF. A number of robustness
1136 and debugging improvements to the cyrus_sasl authenticator.
1138 PP/08 cyrus_sasl server now expands the server_realm option.
1140 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1141 Patch by Jeremy Harris.
1143 PP/10 Added dbmjz lookup type.
1145 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1147 PP/12 MAIL args handles TAB as well as SP, for better interop with
1148 non-compliant senders.
1149 Analysis and variant patch by Todd Lyons.
1151 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1152 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1153 Patch from Dirk Mueller <dmueller@suse.com>
1155 PP/13 tls_peerdn now print-escaped for spool files.
1156 Observed some $tls_peerdn in wild which contained \n, which resulted
1157 in spool file corruption.
1159 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1160 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1161 or write after TLS renegotiation, which otherwise led to messages
1164 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1165 as a tracking header (ie: a signed header comes before the signature).
1166 Patch from Wolfgang Breyha.
1168 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1169 comma-sep list; embedded commas doubled.
1171 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1173 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1175 Report and patch from Dmitry Banschikov.
1177 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1178 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1179 failures. We appear to now support TLS1.1+ with Exim.
1181 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1182 lets Exim select keys and certificates based upon TLS SNI from client.
1183 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1184 before an outbound SMTP session. New log_selector, +tls_sni.
1186 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1187 NULL dereference. Report and patch from Alun Jones.
1189 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1190 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1191 Not seeing resolver debug output on NetBSD, but suspect this is a
1192 resolver implementation change.
1194 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1195 Left warnings. Added "eximon gdb" invocation mode.
1197 PP/21 Defaulting "accept_8bitmime" to true, not false.
1199 PP/22 Added -bw for inetd wait mode support.
1201 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1202 locate the relevant includes and libraries. Made this the default.
1204 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1205 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1207 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1208 This may cause build issues on older platforms.
1210 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1211 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1212 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1213 Added SNI support via GnuTLS too.
1214 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1216 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1218 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1219 Applied second patch from Janne, implementing suggestion to default
1220 multiple-strings-in-record handling to match SPF spec.
1222 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1224 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1225 read-only, out of scope).
1226 Patch from Wolfgang Breyha, report from Stuart Northfield.
1228 PP/29 Fix three issues highlighted by clang analyser static analysis.
1229 Only crash-plausible issue would require the Cambridge-specific
1230 iplookup router and a misconfiguration.
1231 Report from Marcin Mirosław.
1233 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1235 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1236 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1237 As part of this, removing so much warning spew let me fix some minor
1238 real issues in debug logging.
1240 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1241 assignment on my part. Fixed.
1243 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1244 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1245 Janne Snabb (who went above and beyond: thank you).
1247 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1248 string otherwise requires a connection and a bunch more work and it's
1249 relatively easy to get wrong. Should also expose TLS library linkage
1252 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1253 64-bit ${eval} (JH/03).
1255 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1256 GNU libc to support some of the 64-bit stuff, should not lead to
1257 conflicts. Defined before os.h is pulled in, so if a given platform
1258 needs to override this, it can.
1260 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1261 protection layer was required, which is not implemented.
1262 Bugzilla 1254, patch from Wolfgang Breyha.
1264 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1265 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1266 tls_dhparam take prime identifiers. Also unbreak combination of
1267 OpenSSL+DH_params+TLSSNI.
1269 PP/39 Disable SSLv2 by default in OpenSSL support.
1275 PP/01 Solaris build fix for Oracle's LDAP libraries.
1276 Bugzilla 1109, patch from Stephen Usher.
1278 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1280 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1283 TF/02 Fix a couple more cases where we did not log the error message
1284 when unlink() failed. See also change 4.74-TF/03.
1286 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1287 lock up or crash if it happened to be inside a call to libc when it
1288 got a SIGUSR1 from exiwhat.
1290 The SIGUSR1 handler appends the current process status to the process
1291 log which is later printed by exiwhat. It used to use the general
1292 purpose logging code to do this, but several functions it calls are
1293 not safe for signals.
1295 The new output code in the SIGUSR1 handler is specific to the process
1296 log, and simple enough that it's easy to inspect for signal safety.
1297 Removing some special cases also simplifies the general logging code.
1298 Removing the spurious timestamps from the process log simplifies
1301 TF/04 Improved ratelimit ACL condition.
1303 The /noupdate option has been deprecated in favour of /readonly which
1304 has clearer semantics. The /leaky, /strict, and /readonly update modes
1305 are mutually exclusive. The update mode is no longer included in the
1306 database key; it just determines when the database is updated. (This
1307 means that when you upgrade Exim will forget old rate measurements.)
1309 Exim now checks that the per_* options are used with an update mode that
1310 makes sense for the current ACL. For example, when Exim is processing a
1311 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1312 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1313 must specify per_mail/readonly. If you omit the update mode it defaults to
1314 /leaky where that makes sense (as before) or /readonly where required.
1316 The /noupdate option is now undocumented but still supported for
1317 backwards compatibility. It is equivalent to /readonly except that in
1318 ACLs where /readonly is required you may specify /leaky/noupdate or
1319 /strict/noupdate which are treated the same as /readonly.
1321 A useful new feature is the /count= option. This is a generalization
1322 of the per_byte option, so that you can measure the throughput of other
1323 aggregate values. For example, the per_byte option is now equivalent
1324 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1326 The per_rcpt option has been generalized using the /count= mechanism
1327 (though it's more complicated than the per_byte equivalence). When it is
1328 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1329 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1330 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1331 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1332 Note that using per_rcpt with a non-readonly update mode in more than
1333 one ACL will cause the recipients to be double-counted. (The per_mail
1334 and per_byte options don't have this problem.)
1336 The handling of very low rates has changed slightly. If the computed rate
1337 is less than the event's count (usually one) then this event is the first
1338 after a long gap. In this case the rate is set to the same as this event's
1339 count, so that the first message of a spam run is counted properly.
1341 The major new feature is a mechanism for counting the rate of unique
1342 events. The new per_addr option counts the number of different
1343 recipients that someone has sent messages to in the last time period. It
1344 behaves like per_rcpt if all the recipient addresses are different, but
1345 duplicate recipient addresses do not increase the measured rate. Like
1346 the /count= option this is a general mechanism, so the per_addr option
1347 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1348 example, measure the rate that a client uses different sender addresses
1349 with the options per_mail/unique=$sender_address. There are further
1350 details in the main documentation.
1352 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1354 TF/06 Removed a few PCRE remnants.
1356 TF/07 Automatically extract Exim's version number from tags in the git
1357 repository when doing development or release builds.
1359 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1360 Bugzilla 879. Patch from Paul Fisher.
1362 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1363 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1366 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1368 PP/05 Variable $av_failed, true if the AV scanner deferred.
1369 Bugzilla 1078. Patch from John Horne.
1371 PP/06 Stop make process more reliably on build failure.
1372 Bugzilla 1087. Patch from Heiko Schlittermann.
1374 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1375 Bugzilla 1089. Patch from Heiko Schlittermann.
1377 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1378 Bugzilla 1131. Patch from Holger Weiß.
1380 PP/09 Handle IPv6 addresses with SPF.
1381 Bugzilla 860. Patch from Wolfgang Breyha.
1383 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1385 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1388 PP/11 match_* no longer expand right-hand-side by default.
1389 New compile-time build option, EXPAND_LISTMATCH_RHS.
1390 New expansion conditions, "inlist", "inlisti".
1392 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1394 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1396 PP/14 fix log_write() format string regression from TF/03.
1397 Bugzilla 1152. Patch from Dmitry Isaikin.
1403 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1405 PP/02 Harmonised TLS library version reporting; only show if debugging.
1406 Layout now matches that introduced for other libraries in 4.74 PP/03.
1408 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1410 PP/04 New "dns_use_edns0" global option.
1412 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1415 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1416 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1418 TK/01 Updated PolarSSL code to 0.14.2.
1419 Bugzilla 1097. Patch from Andreas Metzler.
1421 PP/07 Catch divide-by-zero in ${eval:...}.
1422 Fixes bugzilla 1102.
1424 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1427 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1428 format-string attack -- SECURITY: remote arbitrary code execution.
1430 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1431 time unintentionally subject to list matching rules, letting the header
1432 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1433 arbitrary string expansion). This allowed for information disclosure.
1435 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1436 INT_MIN/-1 -- value coerced to INT_MAX.
1442 NM/01 Workaround for PCRE version dependency in version reporting
1445 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1446 This fixes portability to compilers other than gcc, notably
1447 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1449 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1450 makefiles for portability to HP-UX and POSIX correctness.
1452 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1453 Also via indented variable definition in the Makefile.
1454 (Debugging by Oliver Heesakkers).
1456 PP/02 Restore caching of spamd results with expanded spamd_address.
1457 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1459 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1460 Improves build reliability. Fix from: Frank Elsner
1462 NM/02 Fix wide character breakage in the rfc2047 coding
1463 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1465 NM/03 Allow underscore in dnslist lookups
1466 Fixes bug 1026. Patch from Graeme Fowler
1468 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1469 Code patches from Adam Ciarcinski of NetBSD.
1471 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1474 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1475 is logged, to avoid truncation. Patch from John Horne.
1477 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1478 Patch from Jakob Hirsch.
1480 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1481 SQL string expansion failure details.
1482 Patch from Andrey Oktyabrski.
1484 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1485 Patch from Simon Arlott.
1487 PP/09 New lookups functionality failed to compile on old gcc which rejects
1488 extern declarations in function scope.
1489 Patch from Oliver Fleischmann
1491 PP/10 Use sig_atomic_t for flags set from signal handlers.
1492 Check getgroups() return and improve debugging.
1493 Fixed developed for diagnosis in bug 927 (which turned out to be
1496 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1497 Patch from Mark Zealey.
1499 PP/12 Bugzilla 1056: Improved spamd server selection.
1500 Patch from Mark Zealey.
1502 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1503 Based on patch from Heiko Schlittermann.
1505 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1506 Patch from Uwe Doering, sign-off by Michael Haardt.
1508 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1509 variable declaration deep within a block. Bug and patch from
1512 PP/15 lookups-Makefile IRIX compatibility coercion.
1514 PP/16 Make DISABLE_DKIM build knob functional.
1516 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1517 Patch by Simon Arlott
1519 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1520 variable argument macros. Our copy now differs from upstream.
1526 TF/01 Failure to get a lock on a hints database can have serious
1527 consequences so log it to the panic log.
1529 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1530 controlled using the smtp_confirmation log selector.
1532 TF/03 Include the error message when we fail to unlink a spool file.
1534 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1535 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1536 for maintaining out-of-tree patches for some time.
1538 PP/01 Bugzilla 139: Documentation and portability issues.
1539 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1540 Handle per-OS dynamic-module compilation flags.
1542 PP/02 Let /dev/null have normal permissions.
1543 The 4.73 fixes were a little too stringent and complained about the
1544 permissions on /dev/null. Exempt it from some checks.
1545 Reported by Andreas M. Kirchwitz.
1547 PP/03 Report version information for many libraries, including
1548 Exim version information for dynamically loaded libraries. Created
1549 version.h, now support a version extension string for distributors
1550 who patch heavily. Dynamic module ABI change.
1552 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1553 privilege escalation vulnerability whereby the Exim run-time user
1554 can cause root to append content of the attacker's choosing to
1557 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1560 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1561 If dropping privileges for untrusted macros, we disabled normal logging
1562 on the basis that it would fail; for the Exim run-time user, this is not
1563 the case, and it resulted in successful deliveries going unlogged.
1564 Fixed. Reported by Andreas Metzler.
1570 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1571 only prepend for the Resent-* case. Fixes regression introduced in
1572 Exim 4.70 by NM/22 for Bugzilla 607.
1574 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1575 increasing numbers of administrators be bitten by this.
1577 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1579 PP/03 Bugzilla 994: added openssl_options main configuration option.
1581 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1583 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1585 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1587 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1588 without a peer certificate, leading to a segfault because of an
1589 assumption that peers always have certificates. Be a little more
1590 paranoid. Problem reported by Martin Tscholak.
1592 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1593 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1594 NB: ClamAV planning to remove STREAM in "middle of 2010".
1595 CL also introduces -bmalware, various -d+acl logging additions and
1596 more caution in buffer sizes.
1598 PP/09 Implemented reverse_ip expansion operator.
1600 PP/10 Bugzilla 937: provide a "debug" ACL control.
1602 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1604 PP/12 Bugzilla 973: Implement --version.
1606 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1608 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1610 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1612 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1613 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1614 ignore trailing whitespace.
1616 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1618 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1621 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1622 Notification from Dr Andrew Aitchison.
1624 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1625 ExtendedDetectionInfo response format.
1626 Notification from John Horne.
1628 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1631 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1632 XSL and documented dependency on system catalogs, with examples of how
1635 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1638 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1639 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1640 configuration file which is writeable by the Exim user or group.
1642 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1643 of configuration files to cover files specified with the -C option if
1644 they are going to be used with root privileges, not just the default
1647 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1648 option (effectively making it always true).
1650 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1651 files to be used while preserving root privileges.
1653 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1654 that rogue child processes cannot use them.
1656 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1657 run-time user, instead of root.
1659 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1660 Exim run-time user without dropping privileges.
1662 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1663 result string, instead of calling string_vformat() twice with the same
1666 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1667 for other users. Others should always drop root privileges if they use
1668 -C on the command line, even for a whitelisted configure file.
1670 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1672 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1678 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1679 $data_path, and $header_path variables; fixed documentation bugs and
1682 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1683 exipick to access non-standard spools, including the "frozen" queue
1686 NM/01 Bugzilla 965: Support mysql stored procedures.
1687 Patch from Alain Williams
1689 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1691 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1692 Patch from Andreas Metzler
1694 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1695 Patch from Kirill Miazine
1697 NM/05 Bugzilla 671: Added umask to procmail example.
1699 JJ/03 installed exipick 20100323.0, fixing doc bug
1701 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1702 directory. Notification and patch from Dan Rosenberg.
1704 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1706 TK/02 Improve log output when DKIM signing operation fails.
1708 MH/01 Treat the transport option dkim_domain as a colon separated
1709 list, not as a single string, and sign the message with each element,
1710 omitting multiple occurences of the same signer.
1712 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1713 Bugzilla 985, 986. Patch by Simon Arlott
1715 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1716 Patch by Simon Arlott
1718 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1719 MBX locking. Notification from Dan Rosenberg.
1725 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1727 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1729 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1731 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1733 NM/04 Bugzilla 915: Flag broken perl installation during build.
1739 TK/01 Added patch by Johannes Berg that expands the main option
1740 "spamd_address" if it starts with a dollar sign.
1742 TK/02 Write list of recipients to X-Envelope-Sender header when building
1743 the mbox-format spool file for content scanning (suggested by Jakob
1746 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1747 (http://www.dcc-servers.net/) support via dccifd. Activated by
1748 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1750 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1751 by Mark Daniel Reidel <mr@df.eu>.
1753 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1754 When building exim an external PCRE library is now needed -
1755 PCRE is a system library on the majority of modern systems.
1756 See entry on PCRE_LIBS in EDITME file.
1758 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1759 conversation. Added nologin parameter to request.
1760 Patch contributed by Kirill Miazine.
1762 TF/01 Do not log submission mode rewrites if they do not change the address.
1764 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1766 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1767 log files in place. Contributed by Roberto Lima.
1769 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1771 TF/03 Bugzilla 615: When checking the local_parts router precondition
1772 after a local_part_suffix or local_part_prefix option, Exim now
1773 does not use the address's named list lookup cache, since this
1774 contains cached lookups for the whole local part.
1776 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1777 Robert Millan. Documentation is in experimental-spec.txt.
1779 TF/04 Bugzilla 668: Fix parallel build (make -j).
1781 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1783 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1784 Patch provided by Jan Srzednicki.
1786 TF/05 Leading white space used to be stripped from $spam_report which
1787 wrecked the formatting. Now it is preserved.
1789 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1790 that they are available at delivery time.
1792 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1794 TF/08 TLS error reporting now respects the incoming_interface and
1795 incoming_port log selectors.
1797 TF/09 Produce a more useful error message if an SMTP transport's hosts
1798 setting expands to an empty string.
1800 NM/06 Bugzilla 744: EXPN did not work under TLS.
1801 Patch provided by Phil Pennock.
1803 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1804 Patch provided by Richard Godbee.
1806 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1807 acl_smtp_notquit, added index entry.
1809 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1810 Patch provided by Eugene Bujak.
1812 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1813 accept(). Patch provided by Maxim Dounin.
1815 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1816 Patch provided by Phil Pennock.
1818 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1820 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1821 Patch provided by Brad "anomie" Jorsch.
1823 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1824 Patch provided by Dean Brooks.
1826 TK/05 Add native DKIM support (does not depend on external libraries).
1828 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1829 Patch provided by Graeme Fowler.
1831 NM/16 Bugzilla 851: Documentation example syntax fix.
1833 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1835 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1838 NM/19 Bugzilla 745: TLS version reporting.
1839 Patch provided by Phil Pennock.
1841 NM/20 Bugzilla 167: bool: condition support.
1842 Patch provided by Phil Pennock.
1844 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1845 clients. Patch provided by Phil Pennock.
1847 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1848 Patch provided by Brad "anomie" Jorsch.
1850 NM/23 Bugzilla 687: Fix misparses in eximstats.
1851 Patch provided by Heiko Schlittermann.
1853 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1854 Patch provided by Heiko Schlittermann.
1856 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1857 plus update to original patch.
1859 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1861 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1862 Patch provided by David Brownlee.
1864 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1866 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1868 NM/30 Bugzilla 888: TLS documentation bugfixes.
1870 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1872 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1873 Unlike the original bugzilla I have changed all shell scripts in src tree.
1875 NM/33 Bugzilla 898: Transport filter timeout fix.
1876 Patch by Todd Rinaldo.
1878 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1879 Patch by Serge Demonchaux.
1881 NM/35 Bugzilla 39: Base64 decode bug fixes.
1882 Patch by Jakob Hirsch.
1884 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1886 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1888 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1890 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1896 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1897 ALT-N's libdkim that I have put here:
1898 http://duncanthrax.net/exim-experimental/
1900 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1901 were called 'true' and it seems that C99 defines that as a reserved
1902 keyword to be used with 'bool' variable types. That means you could
1903 not include C99-style headers which use bools without triggering
1904 build errors in sieve.c.
1906 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1907 as mailq or other aliases. Changed the --help handling significantly
1908 to do whats expected. exim_usage() emits usage/help information.
1910 SC/01 Added the -bylocaldomain option to eximstats.
1912 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1914 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1916 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1922 PH/01 Another patch from the Sieve maintainer.
1924 PH/02 When an IPv6 address is converted to a string for single-key lookup
1925 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1926 dots are used instead of colons so that keys in lsearch files need not
1927 contain colons. This was done some time before quoting was made available
1928 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1929 (notated using the quote facility) so as to distinguish them from IPv4
1930 keys. This meant that lookups for IP addresses in host lists did not work
1931 for iplsearch lookups.
1933 This has been fixed by arranging for IPv6 addresses to be expressed with
1934 colons if the lookup type is iplsearch. This is not incompatible, because
1935 previously such lookups could never work.
1937 The situation is now rather anomalous, since one *can* have colons in
1938 ordinary lsearch keys. However, making the change in all cases is
1939 incompatible and would probably break a number of configurations.
1941 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1944 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1945 conversion specification without a maximum field width, thereby enabling
1946 a rogue spamd server to cause a buffer overflow. While nobody in their
1947 right mind would setup Exim to query an untrusted spamd server, an
1948 attacker that gains access to a server running spamd could potentially
1949 exploit this vulnerability to run arbitrary code as the Exim user.
1951 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1952 $primary_hostname instead of what libspf2 thinks the hosts name is.
1954 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1955 a directory entry by the name of the lookup key. Previously, if a
1956 symlink pointed to a non-existing file or a file in a directory that
1957 Exim lacked permissions to read, a lookup for a key matching that
1958 symlink would fail. Now it is enough that a matching directory entry
1959 exists, symlink or not. (Bugzilla 503.)
1961 PH/03 The body_linecount and body_zerocount variables are now exported in the
1964 PH/04 Added the $dnslist_matched variable.
1966 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1967 This means they are set thereafter only if the connection becomes
1970 PH/06 Added the client_condition to authenticators so that some can be skipped
1971 by clients under certain conditions.
1973 PH/07 The error message for a badly-placed control=no_multiline_responses left
1974 "_responses" off the end of the name.
1976 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1978 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1979 (without spaces) instead of just copying the configuration text.
1981 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1983 PH/11 Added $max_received_linelength.
1985 PH/12 Added +ignore_defer and +include_defer to host lists.
1987 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1988 way in which PCRE > 7.0 is built.
1990 PH/14 Implemented queue_only_load_latch.
1992 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1993 MAIL command. The effect was to mangle the value on 64-bit systems.
1995 PH/16 Another patch from the Sieve maintainer.
1997 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1999 PH/18 If a system quota error occurred while trying to create the file for
2000 a maildir delivery, the message "Mailbox is full" was not appended to the
2001 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2002 only to a quota excession during the actual writing of the file.
2004 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2005 characters?) which causes problems in log lines. The DN values are now
2006 passed through string_printing() before being added to log lines.
2008 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2009 and InterBase are left for another time.)
2011 PH/21 Added message_body_newlines option.
2013 PH/22 Guard against possible overflow in moan_check_errorcopy().
2015 PH/23 POSIX allows open() to be a macro; guard against that.
2017 PH/24 If the recipient of an error message contained an @ in the local part
2018 (suitably quoted, of course), incorrect values were put in $domain and
2019 $local_part during the evaluation of errors_copy.
2025 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2026 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2029 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2030 issue a MAIL command.
2032 PH/02 In an ACL statement such as
2034 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2036 if a client was not listed at all, or was listed with a value other than
2037 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2038 the condition was not true (as it should be), so access was not denied.
2039 The bug was that the ! inversion was incorrectly passed on to the second
2040 item. This has been fixed.
2042 PH/03 Added additional dnslists conditions == and =& which are different from
2043 = and & when the dns lookup returns more than one IP address.
2045 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2046 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2048 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2049 FSYNC, which compiles an option called disable_fsync that allows for
2050 bypassing fsync(). The documentation is heavily laced with warnings.
2052 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2054 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2055 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2056 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2057 including adding "make clean"; (3) Added -fPIC when compiling the test
2058 dynamically loaded module, to get rid of a warning.
2060 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2061 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2062 = 0s. The bug is otherwise harmless.
2064 PH/07 There was a bug in the dovecot authenticator such that the value of
2065 $auth1 could be overwritten, and so not correctly preserved, after a
2066 successful authentication. This usually meant that the value preserved by
2067 the server_setid option was incorrect.
2069 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2071 PH/09 Installed PCRE release 7.0.
2073 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2074 run for batched SMTP input. It is now run at the start of every message
2075 in the batch. While fixing this I discovered that the process information
2076 (output by running exiwhat) was not always getting set for -bs and -bS
2077 input. This is fixed, and it now also says "batched" for BSMTP.
2079 PH/11 Added control=no_pipelining.
2081 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2082 patch, slightly modified), and move the expansion of helo_data till after
2083 the connection is made in the smtp transport (so it can use these
2086 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2088 PH/14 Added log_selector = +pid.
2090 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2092 PH/16 Add ${if forany and ${if forall.
2094 PH/17 Added dsn_from option to vary the From: line in DSNs.
2096 PH/18 Flush SMTP output before performing a callout, unless control =
2097 no_callout_flush is set.
2099 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2100 was true (the default) a successful delivery failed to delete the retry
2101 item, thus causing premature timeout of the address. The bug is now
2104 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2106 PH/21 Long custom messages for fakedefer and fakereject are now split up
2107 into multiline responses in the same way that messages for "deny" and
2108 other ACL rejections are.
2110 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2111 with slight modification.
2113 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2114 draft, changing the syntax and factoring some duplicate code".
2116 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2117 for deliveries of the second and subsequent messages over the same SMTP
2120 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2121 ${reduce, with only minor "tidies".
2123 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2125 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2126 expansion side effects.
2128 PH/27 When a message times out after an over-quota error from an Exim-imposed
2129 quota, the bounce message says "mailbox is full". This message was not
2130 being given when it was a system quota that was exceeded. It now should
2133 MH/03 Made $recipients available in local_scan(). local_scan() already has
2134 better access to the recipient list through recipients_list[], but
2135 $recipients can be useful in postmaster-provided expansion strings.
2137 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2138 in the case of a MAIL command with additional options following the
2139 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2140 were accidentally chopped off.
2142 PH/29 SMTP synchronization checks are implemented when a command is read -
2143 there is a check that no more input is waiting when there shouldn't be
2144 any. However, for some commands, a delay in an ACL can mean that it is
2145 some time before the response is written. In this time, more input might
2146 arrive, invalidly. So now there are extra checks after an ACL has run for
2147 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2148 pipelining has not been advertised.
2150 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2152 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2153 returned in the SMTP response when smtp_return_error_details was false.
2154 This has been fixed.
2156 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2157 instead of the C I/O that was originally supplied, because problems were
2158 reported on Solaris.
2160 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2161 Exim which did not show up earlier: it was assuming that a call to
2162 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2163 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2164 was a macro that became an assignment, so it seemed to work. This has
2165 changed to a proper function call with a void return, hence the compile
2166 error. Exim's code has been fixed.
2168 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2171 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2173 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2175 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2176 because it turns out to be too big or there is a timeout) there is no
2177 "Completed" line in the log. When some messages of this type were
2178 selected by exigrep, they were listed as "not completed". Others were
2179 picked up by some special patterns. I have improved the selection
2180 criteria to be more general.
2182 PH/38 The host_find_failed option in the manualroute router can now be set
2183 to "ignore", to completely ignore a host whose IP address cannot be
2184 found. If all hosts are ignored, the behaviour is controlled by the new
2185 host_all_ignored option.
2187 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2188 homing or because of multiple MX records with /mx) generated more than
2189 one IP address, and the following item turned out to be the local host,
2190 all the secondary addresses of the first item were incorrectly removed
2191 from the list, along with the local host and any following hosts (which
2192 is what is supposed to happen).
2194 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2195 whoever called Exim into the -H file. In the case of the daemon it was
2196 behaving confusingly. When first started, it used values for whoever
2197 started the daemon, but after a SIGHUP it used the Exim user (because it
2198 calls itself on a restart). I have changed the code so that it now always
2201 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2202 message are rejected with the same error (e.g. no authentication or bad
2203 sender address), and a DATA command is nevertheless sent (as can happen
2204 with PIPELINING or a stupid MUA), the error message that was given to the
2205 RCPT commands is included in the rejection of the DATA command. This is
2206 intended to be helpful for MUAs that show only the final error to their
2209 PH/42 Another patch from the Sieve maintainer.
2211 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2212 Eximstats - Fixed some broken HTML links and added missing column headers
2214 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2215 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2217 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2219 PH/43 Yet another patch from the Sieve maintainer.
2221 PH/44 I found a way to check for a TCP/IP connection going away before sending
2222 the response to the final '.' that terminates a message, but only in the
2223 case where the client has not sent further data following the '.'
2224 (unfortunately, this is allowed). However, in many cases there won't be
2225 any further data because there won't be any more messages to send. A call
2226 to select() can be used: if it shows that the input is "ready", there is
2227 either input waiting, or the socket has been closed. An attempt to read
2228 the next input character can distinguish the two cases. Previously, Exim
2229 would have sent an OK response which the client would never have see.
2230 This could lead to message repetition. This fix should cure that, at
2231 least in a lot of common cases.
2233 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2234 advertised in response to EHLO.
2240 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2241 fixed by 4.65/MH/01 (is this a record?) are fixed:
2243 (i) An empty string was always treated as zero by the numeric comparison
2244 operators. This behaviour has been restored.
2246 (ii) It is documented that the numeric comparison operators always treat
2247 their arguments as decimal numbers. This was broken in that numbers
2248 starting with 0 were being interpreted as octal.
2250 While fixing these problems I realized that there was another issue that
2251 hadn't been noticed. Values of message_size_limit (both the global option
2252 and the transport option) were treated as octal if they started with 0.
2253 The documentation was vague. These values are now always treated as
2254 decimal, and I will make that clear in the documentation.
2260 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2261 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2264 MH/01 Don't check that the operands of numeric comparison operators are
2265 integers when their expansion is in "skipping" mode (fixes bug
2266 introduced by 4.64-PH/07).
2268 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2269 child addresses, Exim now panics and dies. Previously, because the count
2270 is held in a short int, deliveries were likely to be lost. As such a
2271 large number of recipients for a single message is ridiculous
2272 (performance will be very, very poor), I have chosen to impose a limit
2273 rather than extend the field.
2279 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2280 leftover -K file (the existence of which was triggered by #402).
2281 While we were at it, introduced process PID as part of the -K
2282 filename. This should rule out race conditions when creating
2285 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2286 processing considerably. Previous code took too long for large mails,
2287 triggering a timeout which in turn triggers #401.
2289 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2290 in the DK code in transports.c. sendfile() is not really portable,
2291 hence the _LINUX specificness.
2293 TF/01 In the add_headers option to the mail command in an Exim filter,
2294 there was a bug that Exim would claim a syntax error in any
2295 header after the first one which had an odd number of characters
2298 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2299 callout verification, Exim cached a "reject" for the entire domain. This
2300 is correct for most verifications, but it is not correct for a recipient
2301 verification with use_sender or use_postmaster set, because in that case
2302 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2303 case of MAIL FROM:<> rejection from other early rejections (e.g.
2304 rejection of HELO). When verifying a recipient using a non-null MAIL
2305 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2306 Whatever the result of the callout, the value of the domain cache is
2307 left unchanged (for any other kind of callout, getting as far as trying
2308 RCPT means that the domain itself is ok).
2310 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2313 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2314 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2315 session, the daemon ignores ECONNECT errors and logs others; it now
2316 ignores EPIPE as well.
2318 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2319 (quoted-printable decoding).
2321 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2322 later the small subsequent patch to fix an introduced bug.
2324 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2326 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2328 PH/08 An error is now given if message_size_limit is specified negative.
2330 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2331 to be given (somewhat) arbitrary names.
2333 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2336 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2337 miscellaneous code fixes
2339 PH/10 Added the log_reject_target ACL modifier to specify where to log
2342 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2343 hostname. This is wrong, because it relates to the incoming message (and
2344 probably the interface on which it is arriving) and not to the outgoing
2345 callout (which could be using a different interface). This has been
2346 changed to use the value of the helo_data option from the smtp transport
2347 instead - this is what is used when a message is actually being sent. If
2348 there is no remote transport (possible with a router that sets up host
2349 addresses), $smtp_active_hostname is used.
2351 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2352 tweaks were necessary in order to get it to work (see also 21 below):
2353 (a) The code assumed that strncpy() returns a negative number on buffer
2354 overflow, which isn't the case. Replaced with Exim's string_format()
2356 (b) There were several signed/unsigned issues. I just did the minimum
2357 hacking in of casts. There is scope for a larger refactoring.
2358 (c) The code used strcasecmp() which is not a standard C function.
2359 Replaced with Exim's strcmpic() function.
2360 (d) The code set only $1; it now sets $auth1 as well.
2361 (e) A simple test gave the error "authentication client didn't specify
2362 service in request". It would seem that Dovecot has changed its
2363 interface. Fortunately there's a specification; I followed it and
2364 changed what the client sends and it appears to be working now.
2366 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2369 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2370 address A is aliased to B and C, where B exists and C does not. Without
2371 -v the output is "A verified" because verification stops after a
2372 successful redirection if more than one address is generated. However,
2373 with -v the child addresses are also verified. Exim was outputting "A
2374 failed to verify" and then showing the successful verification for C,
2375 with its parentage. It now outputs "B failed to verify", showing B's
2376 parentage before showing the successful verification of C.
2378 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2379 look up a TXT record in a specific list after matching in a combined
2382 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2383 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2384 they consult the DNS. I had assumed they would set it the way they
2385 wanted; and indeed my experiments on Linux seem to show that in some
2386 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2387 To be on the safe side, however, I have now made the interface to
2388 host_find_byname() similar to host_find_bydns(), with an argument
2389 containing the DNS resolver options. The host_find_byname() function now
2390 sets these options at its start, just as host_find_bydns() does. The smtp
2391 transport options dns_qualify_single and dns_search_parents are passed to
2392 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2393 of host_find_byname() use the default settings of RES_DEFNAMES
2394 (qualify_single) but not RES_DNSRCH (search_parents).
2396 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2397 spool_read_header() do less string testing, by means of a preliminary
2398 switch on the second character of optional "-foo" lines. (This is
2399 overdue, caused by the large number of possibilities that now exist.
2400 Originally there were few.) While I was there, I also converted the
2401 str(n)cmp tests so they don't re-test the leading "-" and the first
2402 character, in the hope this might squeeze out yet more improvement.
2404 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2405 flag allowing group syntax was set by the header_syntax check but not
2406 turned off, possible causing trouble later; (2) The flag was not being
2407 set at all for the header_verify test, causing "group"-style headers to
2408 be rejected. I have now set it in this case, and also caused header_
2409 verify to ignore an empty address taken from a group. While doing this, I
2410 came across some other cases where the code for allowing group syntax
2411 while scanning a header line wasn't quite right (mostly, not resetting
2412 the flag correctly in the right place). These bugs could have caused
2413 trouble for malformed header lines. I hope it is now all correct.
2415 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2416 with the "reply" argument non-NULL. The code, however (which originally
2417 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2418 but it didn't always do it. This confused somebody who was copying the
2419 code for some other use. I have removed all the tests.
2421 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2422 feature that was used to support insecure browsers during the U.S. crypto
2423 embargo. It requires special client support, and Exim is probably the
2424 only MTA that supported it -- and would never use it because real RSA is
2425 always available. This code has been removed, because it had the bad
2426 effect of slowing Exim down by computing (never used) parameters for the
2427 RSA_EXPORT functionality.
2429 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2430 authenticator to fail if there's a tab character in the incoming data
2431 (there should never be unless someone is messing about, as it's supposed
2432 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2433 if the connection is using TLS or if the remote IP is the same as the
2434 local IP, and the "valid-client-cert option" if a client certificate has
2437 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2438 authenticators. This can be used for authorization after authentication
2439 succeeds. (In the case of plaintext, it servers for both authentication
2442 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2443 if any retry times were supplied.
2445 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2446 connection, where there is no client IP address to check. In this
2447 situation, the verify now always succeeds.
2449 PH/25 Applied John Jetmore's -Mset patch.
2451 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2453 PH/27 In a string expansion for a processed (not raw) header when multiple
2454 headers of the same name were present, leading whitespace was being
2455 removed from all of them, but trailing whitespace was being removed only
2456 from the last one. Now trailing whitespace is removed from each header
2457 before concatenation. Completely empty headers in a concatenation (as
2458 before) are ignored.
2460 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2461 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2463 PH/29 [Removed. This was a change that I later backed out, and forgot to
2464 correct the ChangeLog entry (that I had efficiently created) before
2465 committing the later change.]
2467 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2468 address errors (4xx response to RCPT) over the same connection as other
2469 messages routed to the same hosts. Such deliveries are always "forced",
2470 so retry times are not inspected. This resulted in far too many retries
2471 for the affected addresses. The effect occurred only when there were more
2472 hosts than the hosts_max_try setting in the smtp transport when it had
2473 the 4xx errors. Those hosts that it had tried were not added to the list
2474 of hosts for which the message was waiting, so if all were tried, there
2475 was no problem. Two fixes have been applied:
2477 (i) If there are any address or message errors in an SMTP delivery, none
2478 of the hosts (tried or untried) are now added to the list of hosts
2479 for which the message is waiting, so the message should not be a
2480 candidate for sending over the same connection that was used for a
2481 successful delivery of some other message. This seems entirely
2482 reasonable: after all the message is NOT "waiting for some host".
2483 This is so "obvious" that I'm not sure why it wasn't done
2484 previously. Hope I haven't missed anything, but it can't do any
2485 harm, as the worst effect is to miss an optimization.
2487 (ii) If, despite (i), such a delivery is accidentally attempted, the
2488 routing retry time is respected, so at least it doesn't keep
2489 hammering the server.
2491 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2492 in ${readsocket because some servers need this prod.
2494 PH/32 Added some extra debug output when updating a wait-xxx database.
2496 PH/33 The hint "could be header name not terminated by colon", which has been
2497 given for certain expansion errors for a long time, was not being given
2498 for the ${if def:h_colon_omitted{... case.
2500 PH/34 The spec says: "With one important exception, whenever a domain list is
2501 being scanned, $domain contains the subject domain." There was at least
2502 one case where this was not true.
2504 PH/35 The error "getsockname() failed: connection reset by peer" was being
2505 written to the panic log as well as the main log, but it isn't really
2506 panic-worthy as it just means the connection died rather early on. I have
2507 removed the panic log writing for the ECONNRESET error when getsockname()
2510 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2511 runs only) independently of the message's sender address. This meant
2512 that, if the 4xx error was in fact related to the sender, a different
2513 message to the same recipient with a different sender could confuse
2514 things. In particular, this can happen when sending to a greylisting
2515 server, but other circumstances could also provoke similar problems.
2516 I have changed the default so that the retry time for these errors is now
2517 based a combination of the sender and recipient addresses. This change
2518 can be overridden by setting address_retry_include_sender=false in the
2521 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2522 remote server are returned as part of bounce messages. This was not
2523 happening for LMTP over a pipe (the lmtp transport), but now it is the
2524 same for both kinds of LMTP.
2526 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2527 in header lines that were in fact CNAMEs. This is no longer the case.
2529 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2530 and queue runs started by the daemon processed all messages. This has
2531 been fixed so that -R and -S can now usefully be given with -q<time>.
2533 PH/40 Import PCRE release 6.7 (fixes some bugs).
2535 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2537 PH/42 Give an error if -q is specified more than once.
2539 PH/43 Renamed the variables $interface_address and $interface_port as
2540 $received_ip_address and $received_port, to make it clear that these
2541 values apply to message reception, and not to the outgoing interface when
2542 a message is delivered. (The old names remain recognized, of course.)
2544 PH/44 There was no timeout on the connect() call when using a Unix domain
2545 socket in the ${readsocket expansion. There now is.
2547 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2548 be meaningful with "accept".
2550 SC/01 Eximstats V1.43
2551 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2553 SC/02 Eximstats V1.44
2554 Use a glob alias rather than an array ref in the generated
2555 parser. This improves both readability and performance.
2557 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2558 Collect SpamAssassin and rejection statistics.
2559 Don't display local sender or destination tables unless
2560 there is data to show.
2561 Added average volumes into the top table text output.
2563 SC/04 Eximstats V1.46
2564 Collect data on the number of addresses (recipients)
2565 as well as the number of messages.
2567 SC/05 Eximstats V1.47
2568 Added 'Message too big' to the list of mail rejection
2569 reasons (thanks to Marco Gaiarin).
2571 SC/06 Eximstats V1.48
2572 Mainlog lines which have GMT offsets and are too short to
2573 have a flag are now skipped.
2575 SC/07 Eximstats V1.49 (Alain Williams)
2576 Added the -emptyok flag.
2578 SC/08 Eximstats V1.50
2579 Fixes for obtaining the IP address from reject messages.
2581 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2582 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2583 whitespace changes from 4.64-PH/27
2585 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2588 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2589 are found, allow negative numbers in numeric criteria)
2591 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2593 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2594 to match changes made in 4.64-PH/43
2596 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2598 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2599 "domain" is actually an IP address, and force a failure. This locks out
2600 those revolvers/nameservers that support "A-for-A" lookups, in
2601 contravention of the specifications.
2603 PH/48 When a host name was looked up from an IP address, and the subsequent
2604 forward lookup of the name timed out, the host name was left in
2605 $sender_host_name, contrary to the specification.
2607 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2608 restricted to single-key lookups, Exim was not diagnosing an error if
2609 * or *@ was used with a query-style lookup.
2611 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2613 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2614 long ago, but noone interested enough thought of it. Let's just say that
2615 the "1.1" means that there are some new functions that weren't there at
2616 some point in the past.
2618 PH/51 Error processing for expansion failure of helo_data from an smtp
2619 transport during callout processing was broken.
2621 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2622 tested/used via the -bh/-bhc/-bs options.
2624 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2625 bug, fixed in subsequent PCRE releases).
2627 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2628 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2630 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2636 SC/01 Use a glob alias rather than an array ref in eximstats generated
2637 parser. This improves both readability and performance.
2639 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2640 Don't display local sender or destination tables in eximstats unless
2641 there is data to show.
2642 Added average volumes into the eximstats top table text output.
2644 SC/03 Collect data on the number of addresses (recipients) as well
2645 as the number of messages in eximstats.
2647 TF/01 Correct an error in the documentation for the redirect router. Exim
2648 does (usually) call initgroups() when daemonizing.
2650 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2651 with consistent privilege compared to when running as a daemon.
2653 TF/03 Note in the spec that $authenticated_id is not set for local
2654 submissions from trusted users.
2656 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2657 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2659 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2660 by adding some example configuration directives to the default
2661 configuration file. A little bit of work is required to uncomment the
2662 directives and define how usernames and passwords are checked, but
2663 there is now a framework to start from.
2665 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2666 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2667 without this. I don't know how relevant this is to other LDAP libraries.
2669 PH/02 Add the verb name to the "unknown ACL verb" error.
2671 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2673 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2675 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2676 directory not expanded when it should be if an expanded home directory
2677 was set for the address (which is overridden by the transport).
2679 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2682 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2683 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2684 because it is too late at that time, and has no effect.
2686 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2687 security issue with \' (bugzilla #107). I could not use the
2688 PQescapeStringConn() function, because it needs a PGconn value as one of
2691 PH/08 When testing addresses using -bt, indicate those final addresses that
2692 are duplicates that would not cause an additional delivery. At least one
2693 person was confused, thinking that -bt output corresponded to deliveries.
2694 (Suppressing duplicates isn't a good idea as you lose the information
2695 about possibly different redirections that led to the duplicates.)
2697 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2698 systems where poll() doesn't work, in particular OS X.
2700 PH/10 Added more information to debugging output for retry time not reached.
2702 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2703 operations in malware.c.
2705 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2708 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2709 syslog_duplication set false, log lines that would normally be written
2710 both the the main log and to the reject log were not written to syslog at
2713 PH/14 In the default configuration, change the use of "message" in ACL warn
2714 statements to "add_header".
2716 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2717 not followed by a command (e.g. "seen endif").
2719 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2720 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2723 PH/17 Added extra conditions to the default value of delay_warning_condition
2727 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2728 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2729 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2732 The Auto-Submitted: and various List- headers are standardised, whereas I
2733 don't think Precedence: ever was.
2735 PH/18 Refactored debugging code in route_finduser() to show more information,
2736 in particular, the error code if getpwnam() issues one.
2738 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2739 This is apparently needed in addition to the PH/07 change above to avoid
2740 any possible encoding problems.
2742 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2743 but not after initializing Perl.
2745 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2746 output them only if debugging. By default they are written stderr,
2747 apparently, which is not desirable.
2749 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2752 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2755 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2757 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2758 authenticated or an ident call has been made. Suppress the default
2759 values for $authenticated_id and $authenticated_sender (but permit -oMai
2760 and -oMas) when testing with -bh.
2762 PH/24 Re-jigged the order of the tests in the default configuration so that the
2763 tests for valid domains and recipients precede the DNS black list and CSA
2764 tests, on the grounds that those ones are more expensive.
2766 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2767 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2768 This bug exists in every version of Exim that I still have, right back to
2771 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2772 However, an attempt to turn on case-sensitivity in a regex key by
2773 including (?-i) didn't work because the subject string was already
2774 lowercased, and the effects were non-intuitive. It turns out that a
2775 one-line patch can be used to allow (?-i) to work as expected.
2781 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2782 other effects) broke the use of negated acl sub-conditions.
2784 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2787 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2788 "Deny" causes Exim to reject the incoming connection with a 554 error.
2789 Unfortunately, if there is a major crisis, such as a disk failure,
2790 tcp-wrappers gives "deny", whereas what one would like would be some
2791 kind of temporary error. A kludge has been added to help with this.
2792 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2793 554 error is used if errno is still zero or contains ENOENT (which occurs
2794 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2797 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2799 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2800 errors. Otherwise a message that provokes a temporary error (when other
2801 messages do not) can cause a whole host to time out.
2803 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2804 addresses were routed directly to files or pipes from a redirect router.
2805 File deliveries just didn't batch; pipe deliveries might have suffered
2808 PH/06 A failure to get a lock for a hints database would erroneously always say
2809 "Failed to get write lock", even when it was really a read lock.
2811 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2812 of 0600. This has been changed to use the value of the lockfile_mode
2813 option (which defaults to 0600).
2815 PH/08 Applied small patch from the Sieve maintainer.
2817 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2818 folder from quota calculations, a direct delivery into this folder messed
2819 up the contents of the maildirsize file. This was because the regex was
2820 used only to exclude .Trash (or whatever) when the size of the mailbox
2821 was calculated. There was no check that a delivery was happening into an
2822 excluded directory. This bug has been fixed by ignoring all quota
2823 processing for deliveries into excluded directories.
2825 PH/10 Added the maildirfolder_create_regex option to appendfile.
2831 PH/01 The code for finding all the local interface addresses on a FreeBSD
2832 system running IPv6 was broken. This may well have applied to all BSD
2833 systems, as well as to others that have similar system calls. The broken
2834 code found IPv4 interfaces correctly, but gave incorrect values for the
2835 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2836 that it would not match correctly against @[] and not recognize the IPv6
2839 PH/02 The ipliteral router was not recognizing addresses of the form user@
2840 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2842 PH/03 Added disable_ipv6.
2844 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2845 decoded form, because it is most often used to construct To: headers
2846 lines in autoreplies, and the decoded form may well be syntactically
2847 invalid. However, $reply_address has leading white space removed, and all
2848 newlines turned into spaces so that the autoreply transport does not
2851 PH/05 If group was specified without a user on a router, and no group or user
2852 was specified on a transport, the group from the router was ignored.
2854 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2855 for visible compile-time settings that can be used to change these
2856 numbers, for those that want even more. Backwards compatibility with old
2857 spool files has been maintained. However, going back to a previous Exim
2858 release will lost any variables that are in spool files.
2860 PH/07 Two small changes when running in the test harness: increase delay when
2861 passing a TCP/IP connection to a new process, in case the original
2862 process has to generate a bounce, and remove special handling of
2863 127.0.0.2 (sic), which is no longer necessary.
2865 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2866 be the same on different OS.
2868 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2871 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2872 whether --show-vars was specified or not
2874 JJ/02 exipick: Added support for new ACL variable spool format introduced
2877 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2878 syntactically invalid From: or Reply-to: line, and a filter used this to
2879 generate an autoreply, and therefore failed to obtain an address for the
2880 autoreply, Exim could try to deliver to a non-existent relative file
2881 name, causing unrelated and misleading errors. What now happens is that
2882 it logs this as a hard delivery error, but does not attempt to create a
2885 PH/11 The exinext utility has a -C option for testing purposes, but although
2886 the given file was scanned by exinext itself; it wasn't being passed on
2887 when Exim was called.
2889 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2890 an end-of-file indication when reading a command response.
2892 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2893 compiled. In many other places in Exim, IPv6 addresses are always
2894 recognized, so I have changed this. It also means that IPv4 domain
2895 literals of the form [IPV4:n.n.n.n] are now always recognized.
2897 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2898 used if the router is not running as root, for example, when verifying at
2899 ACL time, or when using -bh. The debugging output from this situation was
2900 non-existent - all you got was a failure to exec. I have made two
2903 (a) Failures to set uid/gid, the current directory, or a process leader
2904 in a subprocess such as that created by queryprogram now generate
2905 suitable debugging output when -d is set.
2907 (b) The queryprogram router detects when it is not running as root,
2908 outputs suitable debugging information if -d is set, and then runs
2909 the subprocess without attempting to change uid/gid.
2911 PH/15 Minor change to Makefile for building test_host (undocumented testing
2914 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2915 additional section of a DNS packet that returns MX or SRV records.
2916 Instead, it always explicitly searches for A/AAAA records. This avoids
2917 major problems that occur when a DNS server includes only records of one
2918 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2919 fixed another bug: if SRV records were looked up and the corresponding
2920 address records were *not* found in the additional section, the port
2921 values from the SRV records were lost.
2923 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2924 using the correct key (the original address) when searching the retry
2925 rules in order to find which one to use for generating the retry hint.
2927 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2928 adding the default one. Similarly, if it contains a Reply-To: header, the
2929 errors_reply_to option, if set, is not used.
2931 PH/19 When calculating a retry time, Exim used to measure the "time since
2932 failure" by looking at the "first failed" field in the retry record. Now
2933 it does not use this if it is later than than the arrival time of the
2934 message. Instead it uses the arrival time. This makes for better
2935 behaviour in cases where some deliveries succeed, thus re-setting the
2936 "first failed" field. An example is a quota failure for a huge message
2937 when small messages continue to be delivered. Without this change, the
2938 "time since failure" will always be short, possible causing more frequent
2939 delivery attempts for the huge message than are intended.
2940 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2942 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2943 $1, $2, $3) because the numerical variables can be reset during some
2944 expansion items (e.g. "match"), thereby losing the authentication data.
2946 PH/21 Make -bV show the size of off_t variables so that the test suite can
2947 decide whether to run tests for quotas > 2G.
2949 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2950 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2951 filecount value is greater than 2G or if a quota value is greater than 2G
2952 on a system where the size of off_t is not greater than 4, a panic error
2955 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2956 never match. The debug and -bh output now contains an explicit error
2957 message indicating a malformed IPv4 address or mask.
2959 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2960 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2961 PH/23 above applies.
2963 PH/25 Do not write to syslog when running in the test harness. The only
2964 occasion when this arises is a failure to open the main or panic logs
2965 (for which there is an explicit test).
2967 PH/26 Added the /no_tell option to "control=freeze".
2969 PH/27 If a host name lookup failed very early in a connection, for example, if
2970 the IP address matched host_lookup and the reverse lookup yielded a name
2971 that did not have a forward lookup, an error message of the form "no IP
2972 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2973 could be logged. Now it outputs the IP address instead of "NULL".
2975 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2976 allows the sender and the authenticated sender to be set when
2977 submitting a message from within Exim. Since child_open_exim() is
2978 documented for local_scan(), the new function should be too.
2980 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2981 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2982 results in an empty string is now treated as unset.
2984 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2986 PH/31 Added sender_verify_fail logging option.
2988 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2989 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2990 tidied the source and removed it altogether.
2992 PH/33 When a queue run was abandoned because the load average was too high, a
2993 log line was always written; now it is written only if the queue_run log
2994 selector is set. In addition, the log line for abandonment now contains
2995 information about the queue run such as the pid. This is always present
2996 in "start" and "stop" lines but was omitted from the "abandon" line.
2998 PH/34 Omit spaces between a header name and the colon in the error message that
2999 is given when verify = headers_syntax fails (if there are lots of them,
3000 the message gets confusing).
3002 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3003 names, as there are now some PTR records that contain slashes. This check
3004 is only to protect against broken name servers that fall over on strange
3005 characters, so the fact that it applies to all lookups doesn't matter.
3007 PH/36 Now that the new test suite is complete, we can remove some of the
3008 special code in Exim that was needed for the old test suite. For example,
3009 sorting DNS records because real resolvers return them in an arbitrary
3010 order. The new test suite's fake resolver always returns records in the
3013 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3014 bounces) except when queue_only is set, to avoid logging races between
3015 the different processes.
3017 PH/38 Panic-die if .include specifies a non-absolute path.
3019 PH/39 A tweak to the "H" retry rule from its user.
3021 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3022 a label. They prevented compilation on older perls.
3024 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3025 a warning to be raised on newish perls.
3027 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3028 on queue. Changes to match documented behaviour of showing count of
3029 messages matching specified criteria.
3031 PH/40 Changed the default ident timeout from 30s to 5s.
3033 PH/41 Added support for the use of login_cap features, on those BSD systems
3034 that have them, for controlling the resources used by pipe deliveries.
3036 PH/42 The content-scanning code uses fopen() to create files in which to put
3037 message data. Previously it was not paying any attention to the mode of
3038 the files. Exim runs with umask(0) because the rest of the code creates
3039 files with open(), and sets the required mode explicitly. Thus, these
3040 files were ending up world-writeable. This was not a big issue, because,
3041 being within the spool directory, they were not world-accessible. I have
3042 created a function called modefopen, which takes an additional mode
3043 argument. It sets umask(777), creates the file, chmods it to the required
3044 mode, then resets the umask. All the relevant calls to fopen() in the
3045 content scanning code have been changed to use this function.
3047 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3048 to 24 hours. This avoids potential overflow problems when processing G
3049 and H retry rules. I suspect nobody ever tinkers with this value.
3051 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3053 PH/45 When the plaintext authenticator is running as a client, the server's
3054 challenges are checked to ensure they are valid base64 strings. By
3055 default, the authentication attempt is cancelled if an invalid string is
3056 received. Setting client_ignore_invalid_base64 true ignores these errors.
3057 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3058 they are received. Thus, the responses can be made to depend on the
3059 challenges. If an invalid string is ignored, an empty string is placed in
3062 PH/46 Messages that are created by the autoreply transport now contains a
3063 References: header, in accordance with RFCs 2822 and 3834.
3065 PH/47 Added authenticated_sender_force to the smtp transport.
3067 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3069 PH/49 Installed latest patch from the Sieve maintainer.
3071 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3072 was also set, the appendfile transport was unnecessarily scanning a
3073 directory of message files (e.g. for maildir delivery) to find the count
3074 of files (along with the size), even though it did not need this
3075 information. It now does the scan only if it needs to find either the
3076 size of the count of files.
3078 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3080 PH/52 Two bugs concerned with error handling when the smtp transport is
3083 (i) Exim was not creating retry information for temporary errors given
3084 for individual recipients after the DATA command when the smtp transport
3085 was used in LMTP mode. This meant that they could be retried too
3086 frequently, and not timed out correctly.
3088 (ii) Exim was setting the flag that allows error details to be returned
3089 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3090 recipients that were returned after the DATA command.
3092 PH/53 This is related to PH/52, but is more general: for any failing address,
3093 when detailed error information was permitted to be returned to the
3094 sender, but the error was temporary, then after the final timeout, only
3095 "retry timeout exceeded" was returned. Now it returns the full error as
3096 well as "retry timeout exceeded".
3098 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3099 do this, and (what is worse) MTAs that accept it.
3101 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3102 will now be deprecated.
3104 PH/56 New os.c-cygwin from the Cygwin maintainer.
3106 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3107 formats (previously only available in exim formats via -bpr, -bpru,
3108 and -bpra. Now also available in native and exiqgrep formats)
3110 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3111 with very large, slow to parse queues
3113 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3115 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3117 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3118 responses to authentication challenges, though it was showing the
3119 challenges; (ii) I've removed the CR characters from the debug output for
3122 PH/58 Allow for the insertion of a newline as well as a space when a string
3123 is turned into more than one encoded-word during RFC 2047 encoding. The
3124 Sieve code now uses this.
3126 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3127 data_4xx, lost_connection, tls_required.
3129 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3130 message was being sent as an SMTP response.
3132 PH/61 Add -l and -k options to exicyclog.
3134 PH/62 When verifying, if an address was redirected to one new address, so that
3135 verification continued, and the new address failed or deferred after
3136 having set something in $address_data, the value of $address_data was not
3137 passed back to the ACL. This was different to the case when no
3138 redirection occurred. The value is now passed back in both cases.
3140 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3141 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3142 use login_cap.h, so on its own it isn't the distinguishing feature. The
3143 new name refers directly to the setclassresources() function.
3145 PH/65 Added configuration files for NetBSD3.
3147 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3149 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3150 is preferred over IPv4.
3152 PH/68 The bounce_return_message and bounce_return_body options were not being
3153 honoured for bounces generated during the reception of non-SMTP messages.
3154 In particular, this applied to messages rejected by the ACL. This bug has
3155 been fixed. However, if bounce_return_message is true and bounce_return_
3156 body is false, the headers that are returned for a non-SMTP message
3157 include only those that have been read before the error was detected.
3158 (In the case of an ACL rejection, they have all been read.)
3160 PH/69 The HTML version of the specification is now built in a directory called
3161 spec_html instead of spec.html, because the latter looks like a path with
3162 a MIME-type, and this confuses some software.
3164 PH/70 Catch two compiler warnings in sieve.c.
3166 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3167 function verify_get_ident() calls ip_connect() to connect a socket, but
3168 if the "connect()" function timed out, ip_connect() used to close the
3169 socket. However, verify_get_ident() also closes the socket later, and in
3170 between Exim writes to the log, which may get opened at this point. When
3171 the socket was closed in ip_connect(), the log could get the same file
3172 descriptor number as the socket. This naturally causes chaos. The fix is
3173 not to close the socket in ip_connect(); the socket should be closed by
3174 the function that creates it. There was only one place in the code where
3175 this was missing, in the iplookup router, which I don't think anybody now
3176 uses, but I've fixed it anyway.
3178 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3179 well as to direct DNS lookups. Otherwise the handling of names in host
3180 lists is inconsistent and therefore confusing.
3186 PH/01 Two changes to the default runtime configuration:
3188 (1) Move the checks for relay_from_hosts and authenticated clients from
3189 after to before the (commented out) DNS black list checks.
3191 (2) Add control=submission to the relay_from_hosts and authenticated
3192 clients checks, on the grounds that messages accepted by these
3193 statements are most likely to be submissions.
3195 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3197 (1) Generate an error if the third argument for the ${prvs expansion is
3200 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3203 (3) Reset the variables that are obtained from the first argument of
3204 ${prvscheck and used in the second argument before leaving the code,
3205 because their memory is reclaimed, so using them afterwards may do
3208 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3209 one (it's much easier than Tom thought :-).
3211 (5) Because of (4), we can now allow for the use of $prvscheck_result
3212 inside the third argument.
3214 PH/03 For some reason, the default setting of PATH when running a command from
3215 a pipe transport was just "/usr/bin". I have changed it to
3218 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3219 anything to be listed in the output from -bV.
3221 PH/05 When a filter generated an autoreply, the entire To: header line was
3222 quoted in the delivery log line, like this:
3224 => >A.N.Other <ano@some.domain> <original@ddress> ...
3226 This has been changed so that it extracts the operative address. There
3227 may be more than one such address. If so, they are comma-separated, like
3230 => >ano@some.domain,ona@other.domain <original@ddress> ...
3232 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3233 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3234 not being looked up in the rDNS to get a host name, Exim was showing the
3235 IP address twice in Received: lines, even though the IP addresses were
3236 identical. For example:
3238 Received: from [1.2.3.4] (helo=[1.2.3.4])
3240 However, if the real host name was known, it was omitting the HELO data
3241 if it matched the actual IP address. This has been tidied up so that it
3242 doesn't show the same IP address twice.
3244 PH/07 When both +timestamp and +memory debugging was on, the value given by
3245 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3246 called by the debug printing, thereby overwriting the timestamp buffer.
3247 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3249 PH/08 When the original message was included in an autoreply transport, it
3250 always said "this is a copy of the message, including all the headers",
3251 even if body_only or headers_only was set. It now gives an appropriate
3254 PH/09 Applied a patch from the Sieve maintainer which:
3256 o fixes some comments
3257 o adds the (disabled) notify extension core
3258 o adds some debug output for the result of if/elsif tests
3259 o points to the current vacation draft in the documentation
3260 and documents the missing references header update
3264 o fixes a bug in processing the envelope test (when testing
3265 multiple envelope elements, the last element determined the
3268 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3269 Electronic Mail") by including:
3271 Auto-submitted: auto-generated
3273 in the messages that it generates (bounce messages and others, such as
3274 warnings). In the case of bounce messages for non-SMTP messages, there was
3275 also a typo: it was using "Auto_submitted" (underscore instead of
3276 hyphen). Since every message generated by Exim is necessarily in response
3277 to another message, thes have all been changed to:
3279 Auto-Submitted: auto-replied
3281 in accordance with these statements in the RFC:
3283 The auto-replied keyword:
3285 - SHOULD be used on messages sent in direct response to another
3286 message by an automatic process,
3288 - MUST NOT be used on manually-generated messages,
3290 - MAY be used on Delivery Status Notifications (DSNs) and Message
3291 Disposition Notifications (MDNs),
3293 - MUST NOT be used on messages generated by automatic or periodic
3294 processes, except for messages which are automatic responses to
3297 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3298 to the default Received: header definition.
3300 PH/12 Added log selector acl_warn_skipped (default on).
3302 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3303 variables because (a) they are in the wrong storage pool and (b) even if
3304 they were copied, it wouldn't work properly because of the caching.
3306 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3307 checking when decoding. Apparently there are clients that generate
3308 overlong encoded strings. Why am I not surprised?
3310 PH/15 If the first argument of "${if match_address" was not empty, but did not
3311 contain an "@" character, Exim crashed. Now it writes a panic log message
3312 and treats the condition as false.
3314 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3316 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3317 extension "envelope-auth". The code is finished and in agreement with
3318 other implementations, but there is no documentation so far and in fact,
3319 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3320 not changing the active code.
3322 Print executed "if" and "elsif" statements when debugging is used. This
3323 helps a great deal to understand what a filter does.
3325 Document more things not specified clearly in RFC3028. I had all this
3326 sorted out, when out of a sudden new issues came to my mind. Oops."
3328 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3331 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3332 canonical form (as documented). However, after a host name lookup from
3333 the IP address, check_host() was doing a simple string comparison with
3334 addresses acquired from the DNS when checking that the found name did
3335 have the original IP as one of its addresses. Since any found IPv6
3336 addresses are likely to be in abbreviated form, the comparison could
3337 fail. Luckily, there already exists a function for doing the comparison
3338 by converting both addresses to binary, so now that is used instead of
3339 the text comparison.
3341 PH/20 There was another similar case to PH/19, when a complete host name was
3342 given in a host list; looking up its IP address could give an abbreviated
3343 form, whereas the current host's name might or might not be abbreviated.
3344 The same fix has been applied.
3350 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3351 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3354 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3355 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3357 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3359 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3360 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3361 Also, for "Auto-Submitted", treat anything other than "no" as
3362 non-personal, in accordance with RFC 3834. (Previously it treated
3363 anything starting "auto-" as non-personal.)
3365 TF/01 The control=submission/name=... option had a problem with syntax
3366 errors if the name included a slash character. The /name= option
3367 now slurps the rest of the string, so it can include any characters
3368 but it must come last in the list of options (after /sender_retain
3371 PH/05 Some modifications to the interface to the fake nameserver for the new
3379 TK/01 Added the "success_on_redirect" address verification option. See
3380 NewStuff for rationale and an example.
3382 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3384 PH/02 Patch to exigrep to allow it to work on syslog lines.
3386 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3387 fread() to skip over the body file's header line, because in Cygwin the
3388 header line is locked and is inaccessible.
3390 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3391 co-exist for some time) to make it clear that it is the Exim ID that is
3392 referenced, not the Message-ID: header line.
3394 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3395 string_format() function, because snprintf() does not exist on all
3398 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3399 ${stat: expansion item.
3401 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3402 protocol synchronization error", to keep the pedants happy.
3404 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3405 well as for IRIX systems, when gcc is being used. See the host.c source
3408 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3410 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3413 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3414 transport and to the smtp transport in LMTP mode.
3416 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3418 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3419 run a filter in a subprocess. This could lead to confusion in subsequent
3420 lookups in the parent process. There should also be a search_tidyup() at
3421 the end of the subprocess.
3423 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3424 only if the host matched helo_try_verify_hosts, which caused the
3425 verification to occur when the EHLO/HELO command was issued. The ACL just
3426 tested the remembered result. Now, if a previous verification attempt has
3427 not happened, "verify = helo" does it there and then.
3429 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3431 TK/03 Fix log output including CR from clamd.
3433 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3434 exist provoked a memory error which could cause a segfault.
3436 PH/15 Installed PCRE 6.2
3438 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3440 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3441 of the problem. Specifically, suggested +O2 rather than +O1 for the
3444 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3446 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3447 record, the port was not correctly logged when the outgoing_port log
3448 selector was set (it logged the transort's default port).
3450 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3451 fallback_hosts, and "hosts" in the smtp transport.
3453 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3454 host errors such as "Connection refused".
3456 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3457 authentication with radiusclient 0.4.9:
3459 - Error returned from rc_read_config was caught wrongly
3460 - Username/password not passed on to radius server due to wrong length.
3462 The presumption is that some radiusclient API changes for 4.51/PH/17
3463 were not taken care of correctly. The code is still untested by me (my
3464 Linux distribution still has 0.3.2 of radiusclient), but it was
3465 contributed by a Radius user.
3467 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3468 expanding the "port" option of the smtp transport.
3470 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3471 while reading a MIME header. Thanks to Tom Hughes for a patch.
3473 PH/24 Include config.h inside local_scan.h so that configuration settings are
3476 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3477 that in an ACL for RCPT (for example), you can examine exactly what was
3480 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3481 commands, but it was not correctly comparing the address with the actual
3482 client host address. Thus, it would show the EHLO address in Received:
3483 header lines when this was not necessary.
3485 PH/27 Added the % operator to ${eval:}.
3487 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3488 it should be ignoring failures (because with -C, for example, it has lost
3489 privilege). It wasn't ignoring creation failures other than "already
3492 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3495 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3496 because an input line was too long, either on its own, or by virtue of
3497 too many continuations, the temporary file was not being removed, and the
3498 return code was incorrect.
3500 PH/31 Missing "BOOL" in function definition in filtertest.c.
3502 PH/32 Applied Sieve patches from the maintainer.
3504 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3506 PH/33 Added "verify = not_blind".
3508 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3509 Local/Makefile (with some defaults set). These are used in built scripts
3510 such as exicyclog, but they have never been used in the exim_install
3511 script (though there are many overriding facilities there). I have
3512 arranged that the exim_install script now takes note of these two
3515 PH/35 Installed configuration files for Dragonfly.
3517 PH/36 When a locally submitted message by a trusted user did not contain a
3518 From: header, and the sender address was obtained from -f or from an SMTP
3519 MAIL command, and the trusted user did not use -F to supply a sender
3520 name, $originator_name was incorrectly used when constructing a From:
3521 header. Furthermore, $originator_name was used for submission mode
3522 messages from external hosts without From: headers in a similar way,
3523 which is clearly wrong.
3525 PH/37 Added control=suppress_local_fixups.
3527 PH/38 When log_selector = +received_sender was set, and the addition of the
3528 sender made the log line's construction buffer exactly full, or one byte
3529 less than full, an overflow happened when the terminating "\n" was
3532 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3533 when the result of a list match is failure because a DNS lookup failed.
3535 PH/40 RM_COMMAND is now used in the building process.
3537 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3538 the "build-* directories that it finds.
3540 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3541 address in a domain literal was a prefix of an interface address.
3543 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3544 when verifying a sender address, unless rewrite_headers is false.
3546 PH/44 Wrote a long comment about why errors_to addresses are verified as
3547 recipients, not senders.
3549 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3550 the ratelimit ACL was added.
3552 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3554 PH/46 Added extra information about PostgreSQL errors to the error string.
3556 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3557 avoiding the need to install special zones in a real server. This is
3558 backwards compatible; if it can't find the fake resolver, it drops back.
3559 Thus, both old and new test suites can be run.
3561 TF/02 Added util/ratelimit.pl
3563 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3564 clock is set back in time.
3566 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3567 Candler <B.Candler@pobox.com>.
3569 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3570 OK for addresses that are the result of redirections.
3572 PH/48 A number of further additions for the benefit of the new test suite,
3573 including a fake gethostbyname() that interfaces to the fake DNS resolver
3576 TF/06 The fix for widen_domains has also been applied to qualify_single and
3577 search_parents which are the other dnslookup options that can cause
3580 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3583 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3585 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3586 service" if it encountered an MX record with an empty target hostname.
3587 The message is now "an MX or SRV record indicated no SMTP service".
3589 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3590 if the DNS of the sending site is misconfigured. This is quite a
3591 common situation. This change restores the behaviour of treating a
3592 helo verification defer as a failure.
3594 PH/51 If self=fail was set on a router, the bounce message did not include the
3595 actual error message.
3601 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3603 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3604 command itself ended in error, the underlying message about the transport
3605 filter timeout was being overwritten with the pipe command error. Now the
3606 underlying error message should be appended to the second error message.
3608 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3610 PH/02 Reduce the amount of output that "make" produces by default. Full output
3611 can still be requested.
3613 PH/03 The warning log line about a condition test deferring for a "warn" verb
3614 was being output only once per connection, rather than after each
3615 occurrence (because it was using the same function as for successful
3616 "warn" verbs). This seems wrong, so I have changed it.
3618 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3619 it should not have, which might have caused a crash in the right
3620 circumstances, but probably never did.
3622 PH/04 Installed a modified version of Tony Finch's patch to make submission
3623 mode fix the return path as well as the Sender: header line, and to
3624 add a /name= option so that you can make the user's friendly name appear
3627 TF/03 Added the control = fakedefer ACL modifier.
3629 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3630 Mark Lowes for thorough testing.
3632 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3634 TK/03 Merged latest SRS patch from Miles Wilton.
3636 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3637 with the definition in sysexits.h (which is #included earlier).
3638 Fortunately, Exim does not actually use EX_OK. The code used to try to
3639 preserve the sysexits.h value, by assuming that macro definitions were
3640 scanned for macro replacements. I have been disabused of this notion,
3641 so now the code just undefines EX_OK before #including unistd.h.
3643 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3644 in the smtp transport. When a block could not be written in a single
3645 write() function, the timeout was being re-applied to each part-write.
3646 This seems wrong - if the receiver was accepting one byte at a time it
3647 would take for ever. The timeout is now adjusted when this happens. It
3648 doesn't have to be particularly precise.
3650 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3651 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3653 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3654 without a domain if the check to <postmaster@domain> fails.
3656 SC/01 Eximstats: added -xls and the ability to specify output files
3657 (patch written by Frank Heydlauf).
3659 SC/02 Eximstats: use FileHandles for outputting results.
3661 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3663 SC/04 Eximstats: fixed display of large numbers with -nvr option
3665 SC/05 Eximstats: fixed merging of reports with empty tables.
3667 SC/06 Eximstats: added the -include_original_destination flag
3669 SC/07 Eximstats: removed tabs and trailing whitespace.
3671 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3673 TK/06 MBOX spool code: Add real "From " MBOX separator line
3674 so the .eml file is really in mbox format (even though
3675 most programs do not really care). Patch from Alex Miller.
3677 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3678 The latter is generated from $received_to and is only set if the
3679 message has one envelope recipient. SA can use these headers,
3680 obviously out-of-the-box. Patch from Alex Miller.
3682 PH/08 The ${def test on a variable was returning false if the variable's
3683 value was "0", contrary to what the specification has always said!
3684 The result should be true unless the variable is empty.
3686 PH/09 The syntax error of a character other than { following "${if
3687 def:variable_name" (after optional whitespace) was not being diagnosed.
3688 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3689 accidental colon was present, for example, could give incorrect results.
3691 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3692 result is used (not including appendfile, where other changes are about
3695 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3696 This involved changing a lot of size variables from int to off_t. It
3697 should work with maildirs and everything.
3699 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3700 spamd dying while we are connected to it.
3702 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3705 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3706 function for BDB 4.3.
3708 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3710 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3711 the sources for PCRE have been re-arranged and more files are now
3714 PH/15 The code I had for printing potentially long long variables in PH/11
3715 above was not the best (it lost precision). The length of off_t variables
3716 is now inspected at build time, and an appropriate printing format (%ld
3717 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3718 to be "long long int" or "long int". This is needed for the internal
3719 formatting function string_vformat().
3721 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3722 the configuration file to be ":syslog", then the script "guesses" where
3723 the logs files are, rather than using the compiled in default. In our
3724 case the guess is not the same as the compiled default, so the script
3725 suddenly stopped working when I started to use syslog. The patch checks
3726 to see if log_file_path is "". If so, it attempts to read it from exim
3727 with no configuration file to get the compiled in version, before it
3728 falls back to the previous guessing code."
3730 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3731 implementing BATV in an Exim configuration. See NewStuff for the gory
3734 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3735 Makefile that are specific to HP-UX.
3737 PH/18 If the "use_postmaster" option was set for a recipient callout together
3738 with the "random" option, the postmaster address was used as the MAIL
3739 FROM address for the random test, but not for the subsequent recipient
3740 test. It is now used for both.
3742 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3743 patch removes a few documentation additions to RFC 3028, because the
3744 latest draft now contains them. It adds the new en;ascii-case comparator
3745 and a new error check for 8bit text in MIME parts. Comparator and
3746 require names are now matched exactly. I enabled the subaddress
3747 extension, but it is not well tested yet (read: it works for me)."
3749 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3750 rework some of the code of TK/09 above to avoid the hardwired use of
3751 "%lld" and "long long". Replaced the call to snprintf() with a call to
3754 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3755 records point to non-existent hosts", "retry timeout exceeded", and
3756 "retry time not reached for any host after a long failure period".
3758 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3759 experimental DomainKeys support:
3761 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3762 (2) On an error such as an illegally used "control", the wrong name for
3763 the control was given.
3765 These problems did NOT occur unless DomainKeys support was compiled.
3767 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3769 PH/24 Added ${if match_ip condition.
3771 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3772 databases so that it will be absolutely obvious if a crash occurs in the
3773 DB library. This is a regular occurrence (often caused by mis-matched
3776 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3777 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3778 picked up on a user's system that detects such things. There doesn't seem
3779 to be a gcc warning option for this - only an attribute that has to be
3780 put on the function's prototype. It seems that in Fedora Core 4 they have
3781 set this on a number of new functions. No doubt there will be more in due
3784 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3785 specify a transport. However, if an address that was verified by such a
3786 router was the subject of a callout, Exim crashed because it tried to
3787 read the rcpt_include_affixes from the non-existent transport. Now it
3788 just assumes that the setting of that option is false. This bug was
3789 introduced by 4.51/PH/31.
3791 PH/28 Changed -d+all to exclude +memory, because that information is very
3792 rarely of interest, but it makes the output a lot bigger. People tend to
3793 do -d+all out of habit.
3795 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3796 code in os-type was giving problems when libc.so lives in lib64, like on
3799 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3800 aren't the modern standard, and it seems that some systems' include files
3801 don't always have them. Exim was already checking for some of the newer
3802 ones like T_AAAA, and defining it itself. I've added checks for all the
3803 record types that Exim uses.
3805 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3806 not automatically generating a new one, as it is supposed to. This
3807 prevented TLS from working. If the file did exist, but contained invalid
3808 data, a new version was generated, as expected. It was only the case of a
3809 non-existent file that was broken.
3811 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3812 with a change in libdomainkeys > 0.64.
3814 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3815 from DNS. If the selector record carries the flag, it now has
3816 precedence over the domain-wide flag.
3818 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3820 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3821 the use of a port name that isn't defined in /etc/services) occurred, the
3822 message was deferred as in a normal delivery, and thus remained on the
3823 spool, instead of being failed because of the mua_wrapper setting. This
3824 is now fixed, and I tidied up some of the mua_wrapper messages at the
3827 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3828 the messages in a hash of arrays rather than using individual hashes.
3829 This is a bit cleaner and results in dramatic memory savings, albeit
3830 at a slight CPU cost.
3832 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3833 as requested by Marc Sherman.
3835 SC/10 Eximstats: added histograms for user specified patterns as requested
3838 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3840 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3841 fopen() in the content-scanning modules that did not already have it.
3847 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3848 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3850 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3852 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3854 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3855 to test Sieve filters that use "vacation".
3857 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3858 that changes the way the GnuTLS parameters are stored in the cache file.
3859 The new format can be generated externally. For backward compatibility,
3860 if the data in the cache doesn't make sense, Exim assumes it has read an
3861 old-format file, and it generates new data and writes a new file. This
3862 means that you can't go back to an older release without removing the
3865 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3866 work if there are any delivery delays because "one_time" forces the
3867 parent to be marked "delivered", so its unseen clone is never tried
3868 again. For this reason, Exim now forbids the simultaneous setting of
3871 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3872 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3873 entry if you want to know the details. The fix, however, neglected to
3874 consider the case when local delivery batching is involved. The test for
3875 "previously delivered" was not happening when checking to see if an
3876 address could be batched with a previous (undelivered) one; under
3877 certain circumstances this could lead to multiple deliveries to the same
3880 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3881 in its include files, and this causes problems building Exim.
3883 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3884 header_syntax) but Exim was just ignoring anything given after a slash.
3885 In particular, this caused confusion with an attempt to use "verify =
3886 reverse_host_lookup/defer_ok". An error is now given when options are
3887 supplied for verify items that do not have them. (Maybe reverse_host_
3888 lookup should have a defer_ok option, but that's a different point.)
3890 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3891 defined by RFC 821) to 2048, because there were problems with some AUTH
3892 commands, and RFC 1869 says the size should be increased for extended
3893 SMTP commands that take arguments.
3895 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3898 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3899 "unknown" error; now it says that the functionality isn't in the binary.
3901 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3902 an address' error message when a string expansion fails (syntax or
3903 whatever). Otherwise the password may appear in the log. Following change
3904 PH/42 below, there is no longer a chance of it appearing in a bounce
3907 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3909 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3910 of its addresses was ever tried. (Bugzilla bug #2.)
3912 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3913 the result incorrectly in the debug output. (It correctly added a newline
3914 to what was transported.)
3916 TF/01 Added $received_time.
3918 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3919 commented out examples of how to interface to a virus scanner and to
3920 SpamAssassin. Also added commented examples of av_scanner and
3921 spamd_address settings.
3923 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3924 and controls are allowed in which ACLs. There were a couple of minor
3925 errors. Some of the entries in the conditions table (which is a table of
3926 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3927 negation of where the condition IS allowed.
3929 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3931 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3932 header file does not have a version number, so I've had to invent a new
3933 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3934 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3935 radiusclient), but it was contributed by a Radius user.
3937 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3938 files or directories, for OpenSSL.
3940 PH/19 When an Exim process that is running as root has to create an Exim log
3941 file, it does so in a subprocess that runs as exim:exim so as to get the
3942 ownership right at creation (otherwise, other Exim processes might see
3943 the file with the wrong ownership). There was no test for failure of this
3944 fork() call, which would lead to the process getting stuck as it waited
3945 for a non-existent subprocess. Forks do occasionally fail when resources
3946 run out. I reviewed all the other calls to fork(); they all seem to check
3949 PH/20 When checking for unexpected SMTP input at connect time (before writing
3950 the banner), Exim was not dealing correctly with a non-positive return
3951 from the read() function. If the client had disconnected by this time,
3952 the result was a log entry for a synchronization error with an empty
3953 string after "input=" when read() returned zero. If read() returned -1
3954 (an event I could not check), uninitialized data bytes were printed.
3955 There were reports of junk text (parts of files, etc) appearing after
3958 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3960 PH/22 Added support for macro redefinition, and (re)definition in between
3961 driver and ACL definitions.
3963 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3964 forgetting to use the resulting value; it was using the unexpanded value.
3966 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3967 hadn't been configured. The fix is from Juergen Kreileder, who
3968 understands it better than I do:
3970 "Here's what I see happening with three configured cyrus_sasl
3971 authenticators configured (plain, login, cram-md5):
3973 On startup auth_cyrus_sasl_init() gets called for each of these.
3974 This means three calls to sasl_listmech() without a specified mech_list.
3975 => SASL tests which mechs of all available mechs actually work
3976 => three warnings about OTP not working
3977 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3979 With the patch, sasl_listmech() also gets called three times. But now
3980 SASL's mech_list option is set to the server_mech specified in the the
3981 authenticator. Or in other words, the answer from sasl_listmech()
3982 gets limited to just the mech you're testing for (which is different
3984 => the return list contains just 'plain' or 'login', 'cram-md5' or
3985 nothing depending on the value of ob->server_mech.
3987 I've just tested the patch: Authentication still works fine,
3988 unavailable mechs specified in the exim configuration are still
3989 caught, and the auth.log warnings about OTP are gone."
3991 PH/25 When debugging is enabled, the contents of the command line are added
3992 to the debugging output, even when log_selector=+arguments is not
3995 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3996 answer is "GNU", and only if the return is "GNU/something" is the answer
3999 PH/27 $acl_verify_message is now set immediately after the failure of a
4000 verification in an ACL, and so is available in subsequent modifiers. In
4001 particular, the message can be preserved by coding like this:
4003 warn !verify = sender
4004 set acl_m0 = $acl_verify_message
4006 Previously, $acl_verify_message was set only while expanding "message"
4007 and "log_message" when a very denied access.
4009 PH/28 Modified OS/os.c-Linux with
4011 -#ifndef OS_LOAD_AVERAGE
4012 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4014 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4015 nomenclature these days.)
4017 PH/29 Installed patch from the Sieve maintainer that adds the options
4018 sieve_useraddress and sieve_subaddress to the redirect router.
4020 PH/30 In these circumstances:
4021 . Two addresses routed to the same list of hosts;
4022 . First host does not offer TLS;
4023 . First host accepts first address;
4024 . First host gives temporary error to second address;
4025 . Second host offers TLS and a TLS session is established;
4026 . Second host accepts second address.
4027 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4028 and peerdn, if requested) that were in fact used only for the second
4031 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4032 attention to any local part prefix or suffix that was matched by the
4033 router that accepted the address. It now behaves in the same way as it
4034 does for delivery: the affixes are removed from the local part unless
4035 rcpt_include_affixes is set on the transport.
4037 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4038 timeout during the DATA phase of an incoming message.
4040 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4041 applied a patch sanctioned by the Sieve maintainer.
4043 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4044 the uid or gid is negative. A case of a negative gid caused this to be
4045 noticed. The fix allows for either to be negative.
4047 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4048 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4051 PH/36 The change PH/12 above was broken. Fixed it.
4053 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4054 the grounds that routing the same address twice would always produce the
4055 same answer. This might have been true once, but it is certainly no
4056 longer true now. Routing a child address may depend on the previous
4057 routing that produced that child. Some complicated redirection strategies
4058 went wrong when messages had multiple recipients, and made Exim's
4059 behaviour dependent on the order in which the addresses were given.
4061 I have moved the duplicate checking until after the routing is complete.
4062 Exim scans the addresses that are assigned to local and remote
4063 transports, and removes any duplicates. This means that more work will be
4064 done, as duplicates will always all be routed, but duplicates are
4065 presumably rare, so I don't expect this is of any significance.
4067 For deliveries to pipes, files, and autoreplies, the duplicate checking
4068 still happens during the routing process, since they are not going to be
4071 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4072 It corrects a timeout issue with spamd. This is Ian's comment: "The
4073 background is that sometimes spamd either never reads data from a
4074 connection it has accepted, or it never writes response data. The exiscan
4075 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4076 blindly assumes that writes won't block so it may never time out."
4078 PH/39 Allow G after quota size as well as K and M.
4080 PH/40 The value set for $authenticated_id in an authenticator may not contain
4081 binary zeroes or newlines because the value is written to log lines and
4082 to spool files. There was no check on this. Now the value is run through
4083 the string_printing() function so that such characters are converted to
4084 printable escape sequences.
4086 PH/41 $message_linecount is a new variable that contains the total number of
4087 lines in the message. Compare $body_linecount, which is the count for the
4090 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4091 bounce and delay warning messages, except in certain special cases, which
4094 (a) An SMTP error message from a remote host;
4095 (b) A message specified in a :fail: redirection;
4096 (c) A message specified in a "fail" command in a system filter;
4097 (d) A message specified in a FAIL return from the queryprogram router;
4098 (e) A message specified by the cannot_route_message router option.
4100 In these cases only, Exim does include the error details in bounce and
4101 warning messages. There are also a few cases where bland messages such
4102 as "unrouteable address" or "local delivery error" are given.
4104 PH/43 $value is now also set for the "else" part of a ${run expansion.
4106 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4107 being worked on, but at least Exim now implements the latest version to
4110 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4111 process to complete was treated as a delivery failure, a timeout while
4112 writing the message to the pipe was logged, but erroneously treated as a
4113 successful delivery. Such timeouts include transport filter timeouts. For
4114 consistency with the overall process timeout, these timeouts are now
4115 treated as errors, giving rise to delivery failures by default. However,
4116 there is now a new Boolean option for the pipe transport called
4117 timeout_defer, which, if set TRUE, converts the failures into defers for
4118 both kinds of timeout. A transport filter timeout is now identified in
4121 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4122 systems where "make" and "gmake" are different, calling "gmake" at top
4123 level broke things. I've arranged for the value of $(MAKE) to be passed
4124 from the Makefile to this script so that it can call the same version of
4128 A note about Exim versions 4.44 and 4.50
4129 ----------------------------------------
4131 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4132 changes of various kinds. As a consequence, a big documentation update was
4133 needed. This delayed the release for rather longer than seemed good, especially
4134 in the light of a couple of (minor) security issues. Therefore, the changes
4135 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4136 release. So 4.44 and 4.50 are in effect two different branches that both start
4139 I have left the 4.50 change log unchanged; it contains all the changes since
4140 4.43. The change log for 4.44 is below; many of its items are identical to
4141 those for 4.50. This seems to be the most sensible way to preserve the
4142 historical information.
4148 1. Minor wording change to the doc/README.SIEVE file.
4150 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4151 computation of the current number of files was incorrect.
4153 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4154 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4157 4. Give more explanation in the error message when the command for a transport
4158 filter fails to execute.
4160 5. There are several places where Exim runs a non-Exim command in a
4161 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4162 was being done only for the command run by the queryprogram router. It is
4163 now done for all such subprocesses. The other cases are: ${run, transport
4164 filters, and the commands run by the lmtp and pipe transports.
4166 6. Added CONFIGURE_GROUP build-time option.
4168 7. Some older OS have a limit of 256 on the maximum number of file
4169 descriptors. Exim was using setrlimit() to set 1000 as a large value
4170 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4171 systems. I've change it so that if it can't get 1000, it tries for 256.
4173 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4174 was an oversight, and furthermore, ever since the addition of extra
4175 controls (e.g. 4.43/32), the checks on when to allow different forms of
4176 "control" were broken. There should now be diagnostics for all cases when a
4177 control that does not make sense is encountered.
4179 9. Added the /retain_sender option to "control=submission".
4181 10. $recipients is now available in the predata ACL (oversight).
4183 11. Tidy the search cache before the fork to do a delivery from a message
4184 received from the command line. Otherwise the child will trigger a lookup
4185 failure and thereby defer the delivery if it tries to use (for example) a
4186 cached ldap connection that the parent has called unbind on.
4188 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4189 of $address_data from the recipient verification was clobbered by the
4190 sender verification.
4192 13. The value of address_data from a sender verification is now available in
4193 $sender_address_data in subsequent conditions in the ACL statement.
4195 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4197 15. Added a new option "connect=<time>" to callout options, to set a different
4200 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4201 was its contents. (It was OK if the option was not defined at all.)
4203 17. A "Completed" log line is now written for messages that are removed from
4204 the spool by the -Mrm option.
4206 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4207 information about exactly what failed.
4209 19. Added -dd to debug only the daemon process.
4211 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4212 handles timeouts, both on the server side and network timeouts. Renamed the
4213 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4215 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4216 It is now set to "smtps".
4218 22. $host_address is now set to the target address during the checking of
4219 ignore_target_hosts.
4221 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4222 being passed; this would have caused $sender_host_name to have been used if
4223 matching the list had actually called for a host name (not very likely,
4224 since this list is usually IP addresses). A host name is now passed as
4227 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4228 code that specifies a non-restarting handler (typically sigaction() in
4229 modern systems) in an attempt to fix a rare and obscure crash bug.
4231 25. Narrowed the window for a race in the daemon that could cause it to ignore
4232 SIGCHLD signals. This is not a major problem, because they are used only to
4233 wake it up if nothing else does.
4235 26. A malformed maildirsize file could cause Exim to calculate negative values
4236 for the mailbox size or file count. Odd effects could occur as a result.
4237 The maildirsize information is now recalculated if the size or filecount
4240 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4241 support for a long time. Removed HAVE_SYS_VFS_H.
4243 28. Installed the latest version of exipick from John Jetmore.
4245 29. In an address list, if the pattern was not a regular expression, an empty
4246 subject address (from a bounce message) matched only if the pattern was an
4247 empty string. Non-empty patterns were not even tested. This was the wrong
4248 because it is perfectly reasonable to use an empty address as part of a
4249 database query. An empty address is now tested by patterns that are
4250 lookups. However, all the other forms of pattern expect the subject to
4251 contain a local part and a domain, and therefore, for them, an empty
4252 address still always fails if the pattern is not itself empty.
4254 30. Exim went into a mad DNS loop when attempting to do a callout where the
4255 host was specified on an smtp transport, and looking it up yielded more
4256 than one IP address.
4258 31. Re-factored the code for checking spool and log partition space into a
4259 function that finds that data and another that does the check. The former
4260 is then used to implement four new variables: $spool_space, $log_space,
4261 $spool_inodes, and $log_inodes.
4263 32. The RFC2047 encoding function was originally intended for short strings
4264 such as real names; it was not keeping to the 75-character limit for
4265 encoded words that the RFC imposes. It now respects the limit, and
4266 generates multiple encoded words if necessary. To be on the safe side, I
4267 have increased the buffer size for the ${rfc2047: expansion operator from
4270 33. It is now permitted to omit both strings after an "if" condition; if the
4271 condition is true, the result is "true". As before, when the second string
4272 is omitted, a false condition yields an empty string. This makes it less
4273 cumbersome to write custom ACL and router conditions.
4275 34. Failure to deliver a bounce message always caused it to be frozen, even if
4276 there was an errors_to setting on the router. The errors_to setting is now
4279 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4280 canonical form (fully expanded) before being placed in
4281 $sender_host_address.
4283 36. The table in the code that translates DNS record types into text (T_A to
4284 "A" for instance) was missing entries for NS and CNAME. It is just possible
4285 that this could have caused confusion if both these types were looked up
4286 for the same domain, because the text type is used as part of Exim's
4287 per-process caching. But the chance of anyone hitting this buglet seems
4290 37. The dnsdb lookup has been extended in a number of ways.
4292 (1) There is a new type, "zns", which walks up the domain tree until it
4293 finds some nameserver records. It should be used with care.
4295 (2) There is a new type, "mxh", which is like "mx" except that it returns
4296 just the host names, not the priorities.
4298 (3) It is now possible to give a list of domains (or IP addresses) to be
4299 looked up. The behaviour when one of the lookups defers can be
4300 controlled by a keyword.
4302 (4) It is now possible to specify the separator character for use when
4303 multiple records are returned.
4305 38. The dnslists ACL condition has been extended: it is now possible to supply
4306 a list of IP addresses and/or domains to be looked up in a particular DNS
4309 39. Added log_selector=+queue_time_overall.
4311 40. When running the queue in the test harness, wait just a tad after forking a
4312 delivery process, to get repeatability of debugging output.
4314 41. Include certificate and key file names in error message when GnuTLS fails
4315 to set them up, because the GnuTLS error message doesn't include the name
4316 of the failing file when there is a problem reading it.
4318 42. Allow both -bf and -bF in the same test run.
4320 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4322 44. The "Exiscan patch" is now merged into the mainline Exim source.
4324 45. Sometimes the final signoff response after QUIT could fail to get
4325 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4326 < 0 before doing a fflush(). This bug looks as though it goes back to the
4327 introduction of TLS in release 3.20, but "sometimes" must have been rare
4328 because the tests only now provoked it.
4330 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4331 (this can affect the format of dates).
4333 47. exim_tidydb, when checking for the continued existence of a message for
4334 which it has found a message-specific retry record, was not finding
4335 messages that were in split spool directories. Consequently, it was
4336 deleting retry records that should have stayed in existence.
4338 48. Steve fixed some bugs in eximstats.
4340 49. The SPA authentication driver was not abandoning authentication and moving
4341 on to the next authenticator when an expansion was forced to fail,
4342 contradicting the general specification for all authenticators. Instead it
4343 was generating a temporary error. It now behaves as specified.
4345 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4346 (the order specifies the preference for clients). The order is now AES256,
4347 AES128, 3DES, ARCFOUR128.
4349 51. Small patch to Sieve code - explicitly set From: when generating an
4352 52. Exim crashed if a remote delivery caused a very long error message to be
4353 recorded - for instance if somebody sent an entire SpamAssassin report back
4354 as a large number of 550 error lines. This bug was coincidentally fixed by
4355 increasing the size of one of Exim's internal buffers (big_buffer) that
4356 happened as part of the Exiscan merge. However, to be on the safe side, I
4357 have made the code more robust (and fixed the comments that describe what
4360 53. Now that there can be additional text after "Completed" in log lines (if
4361 the queue_time_overall log selector is set), a one-byte patch to exigrep
4362 was needed to allow it to recognize "Completed" as not the last thing in
4365 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4366 patch that reportedly fixes this has been added. I am not expert enough to
4367 create a test for it. This is what the patch creator wrote:
4369 "I found a little strange behaviour of ldap code when working with
4370 Windows 2003 AD Domain, where users was placed in more than one
4371 Organization Units. When I tried to give exim partial DN, the exit code
4372 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4373 But simultaneously result of request was absolutely normal ldap result,
4374 so I produce this patch..."
4376 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4377 REFERENCE, so I have modified the code to exclude the patch when that macro
4380 55. Some experimental protocols are using DNS PTR records for new purposes. The
4381 keys for these records are domain names, not reversed IP addresses. The
4382 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4383 leaves it alone. Component reversal etc. now happens only for IP addresses.
4386 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4388 57. Double the size of the debug message buffer (to 2048) so that more of very
4389 long debug lines gets shown.
4391 58. The exicyclog utility now does better if the number of log files to keep
4392 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4394 59. Two changes related to the smtp_active_hostname option:
4396 (1) $smtp_active_hostname is now available as a variable.
4397 (2) The default for smtp_banner uses $smtp_active_hostname instead
4398 of $primary_hostname.
4400 60. The host_aton() function is supposed to be passed a string that is known
4401 to be a valid IP address. However, in the case of IPv6 addresses, it was
4402 not checking this. This is a hostage to fortune. Exim now panics and dies
4403 if the condition is not met. A case was found where this could be provoked
4404 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4405 components; fortuitously, this particular loophole had already been fixed
4406 by change 4.50/55 above.
4408 If there are any other similar loopholes, the new check in host_aton()
4409 itself should stop them being exploited. The report I received stated that
4410 data on the command line could provoke the exploit when Exim was running as
4411 exim, but did not say which command line option was involved. All I could
4412 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4413 running as the user.
4416 61. There was a buffer overflow vulnerability in the SPA authentication code
4417 (which came originally from the Samba project). I have added a test to the
4418 spa_base64_to_bits() function which I hope fixes it.
4421 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4422 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4424 63. The daemon start-up calls getloadavg() while still root for those OS that
4425 need the first call to be done as root, but it missed one case: when
4426 deliver_queue_load_max is set with deliver_drop_privilege. This is
4427 necessary for the benefit of the queue runner, because there is no re-exec
4428 when deliver_drop_privilege is set.
4430 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4431 This has been fixed.
4433 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4434 was in use, was not putting the data itself into the right store pool;
4435 consequently, it could be overwritten for a subsequent message in the same
4436 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4439 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4441 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4442 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4443 the function were treating the return as a boolean value, which happened to
4444 work because 0=false and not-0=true, but is not correct code.
4446 68. The host_aton() function was not handling scoped IPv6 addresses (those
4447 with, for example, "%eth0" on the end) correctly.
4449 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4450 negated items (that is, ~something) in unsigned ints. Some compilers
4451 apparently mutter when there is no cast.
4453 70. If an address verification called from an ACL failed, and did not produce a
4454 user-specific message (i.e. there was only a "system" message), nothing was
4455 put in $acl_verify_message. In this situation, it now puts the system
4458 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4459 session; change 4.31/43 added the unwanted input to the log line - except
4460 that it did not do this in the start of session case. It now does.
4462 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4463 This is wrong and can cause the other end to generate a synchronization
4464 error if it is another Exim or anything else that does the synchronization
4465 check. A QUIT command is no longer sent after a timeout.
4467 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4468 during host lookups.
4470 74. The defer_ok option of callout verification was not working if it was used
4471 when verifying addresses in header lines, that is, for this case:
4473 verify = header_sender/callout=defer_ok
4475 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4476 those file descriptors could be used for SMTP connections. If anything
4477 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4478 could be sent to the SMTP client, causing chaos. The daemon now opens
4479 stdin, stdout, and stderr to /dev/null when it puts itself into the
4482 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4483 log by default. The user can override this with suitable Perl magic.
4485 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4486 the log message when discard triggers, was not working for the DATA ACL or
4487 for the non-SMTP ACL.
4489 78. Error message wording change in sieve.c.
4491 79. If smtp_accept_max_per_host was set, the number of connections could be
4492 restricted to fewer than expected, because the daemon was trying to set up
4493 a new connection before checking whether the processes handling previous
4494 connections had finished. The check for completed processes is now done
4495 earlier. On busy systems, this bug wouldn't be noticed because something
4496 else would have woken the daemon, and it would have reaped the completed
4499 80. If a message was submitted locally by a user whose login name contained one
4500 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4501 It caused a spool format error. I have fixed the spool reading code. A
4502 related problem was that the "from" clause in the Received: line became
4503 illegal because of the space(s). It is now covered by ${quote_local_part.
4505 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4508 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4510 83. Patch from Sieve maintainer to fix unterminated string problem in
4511 "vacation" handling.
4513 84. Some minor changes to the Linux configuration files to help with other
4514 OS variants using glibc.
4516 85. One more patch for Sieve to update vacation handling to latest spec.
4519 ----------------------------------------------------
4520 See the note above about the 4.44 and 4.50 releases.
4521 ----------------------------------------------------
4527 1. Change 4.43/35 introduced a bug that caused file counts to be
4528 incorrectly computed when quota_filecount was set in an appendfile
4531 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4532 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4535 3. Give more explanation in the error message when the command for a transport
4536 filter fails to execute.
4538 4. There are several places where Exim runs a non-Exim command in a
4539 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4540 was being done only for the command run by the queryprogram router. It is
4541 now done for all such subprocesses. The other cases are: ${run, transport
4542 filters, and the commands run by the lmtp and pipe transports.
4544 5. Some older OS have a limit of 256 on the maximum number of file
4545 descriptors. Exim was using setrlimit() to set 1000 as a large value
4546 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4547 systems. I've change it so that if it can't get 1000, it tries for 256.
4549 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4550 was an oversight, and furthermore, ever since the addition of extra
4551 controls (e.g. 4.43/32), the checks on when to allow different forms of
4552 "control" were broken. There should now be diagnostics for all cases when a
4553 control that does not make sense is encountered.
4555 7. $recipients is now available in the predata ACL (oversight).
4557 8. Tidy the search cache before the fork to do a delivery from a message
4558 received from the command line. Otherwise the child will trigger a lookup
4559 failure and thereby defer the delivery if it tries to use (for example) a
4560 cached ldap connection that the parent has called unbind on.
4562 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4563 of $address_data from the recipient verification was clobbered by the
4564 sender verification.
4566 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4567 was its contents. (It was OK if the option was not defined at all.)
4569 11. A "Completed" log line is now written for messages that are removed from
4570 the spool by the -Mrm option.
4572 12. $host_address is now set to the target address during the checking of
4573 ignore_target_hosts.
4575 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4576 being passed; this would have caused $sender_host_name to have been used if
4577 matching the list had actually called for a host name (not very likely,
4578 since this list is usually IP addresses). A host name is now passed as
4581 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4582 code that specifies a non-restarting handler (typically sigaction() in
4583 modern systems) in an attempt to fix a rare and obscure crash bug.
4585 15. Narrowed the window for a race in the daemon that could cause it to ignore
4586 SIGCHLD signals. This is not a major problem, because they are used only to
4587 wake it up if nothing else does.
4589 16. A malformed maildirsize file could cause Exim to calculate negative values
4590 for the mailbox size or file count. Odd effects could occur as a result.
4591 The maildirsize information is now recalculated if the size or filecount
4594 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4595 support for a long time. Removed HAVE_SYS_VFS_H.
4597 18. Updated exipick to current release from John Jetmore.
4599 19. Allow an empty sender to be matched against a lookup in an address list.
4600 Previously the only cases considered were a regular expression, or an
4603 20. Exim went into a mad DNS lookup loop when doing a callout where the
4604 host was specified on the transport, if the DNS lookup yielded more than
4607 21. The RFC2047 encoding function was originally intended for short strings
4608 such as real names; it was not keeping to the 75-character limit for
4609 encoded words that the RFC imposes. It now respects the limit, and
4610 generates multiple encoded words if necessary. To be on the safe side, I
4611 have increased the buffer size for the ${rfc2047: expansion operator from
4614 22. Failure to deliver a bounce message always caused it to be frozen, even if
4615 there was an errors_to setting on the router. The errors_to setting is now
4618 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4619 canonical form (fully expanded) before being placed in
4620 $sender_host_address.
4622 24. Updated eximstats to version 1.33
4624 25. Include certificate and key file names in error message when GnuTLS fails
4625 to set them up, because the GnuTLS error message doesn't include the name
4626 of the failing file when there is a problem reading it.
4628 26. Expand error message when OpenSSL has problems setting up cert/key files.
4631 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4632 (this can affect the format of dates).
4634 28. exim_tidydb, when checking for the continued existence of a message for
4635 which it has found a message-specific retry record, was not finding
4636 messages that were in split spool directories. Consequently, it was
4637 deleting retry records that should have stayed in existence.
4639 29. eximstats updated to version 1.35
4640 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4641 1.35 - bugfix such that pie charts by volume are generated correctly
4643 30. The SPA authentication driver was not abandoning authentication and moving
4644 on to the next authenticator when an expansion was forced to fail,
4645 contradicting the general specification for all authenticators. Instead it
4646 was generating a temporary error. It now behaves as specified.
4648 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4649 (the order specifies the preference for clients). The order is now AES256,
4650 AES128, 3DES, ARCFOUR128.
4652 31. Small patch to Sieve code - explicitly set From: when generating an
4655 32. Exim crashed if a remote delivery caused a very long error message to be
4656 recorded - for instance if somebody sent an entire SpamAssassin report back
4657 as a large number of 550 error lines. This bug was coincidentally fixed by
4658 increasing the size of one of Exim's internal buffers (big_buffer) that
4659 happened as part of the Exiscan merge. However, to be on the safe side, I
4660 have made the code more robust (and fixed the comments that describe what
4663 33. Some experimental protocols are using DNS PTR records for new purposes. The
4664 keys for these records are domain names, not reversed IP addresses. The
4665 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4666 leaves it alone. Component reversal etc. now happens only for IP addresses.
4669 34. The host_aton() function is supposed to be passed a string that is known
4670 to be a valid IP address. However, in the case of IPv6 addresses, it was
4671 not checking this. This is a hostage to fortune. Exim now panics and dies
4672 if the condition is not met. A case was found where this could be provoked
4673 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4674 components; fortuitously, this particular loophole had already been fixed
4675 by change 4.50/55 or 4.44/33 above.
4677 If there are any other similar loopholes, the new check in host_aton()
4678 itself should stop them being exploited. The report I received stated that
4679 data on the command line could provoke the exploit when Exim was running as
4680 exim, but did not say which command line option was involved. All I could
4681 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4682 running as the user.
4685 35. There was a buffer overflow vulnerability in the SPA authentication code
4686 (which came originally from the Samba project). I have added a test to the
4687 spa_base64_to_bits() function which I hope fixes it.
4690 36. The daemon start-up calls getloadavg() while still root for those OS that
4691 need the first call to be done as root, but it missed one case: when
4692 deliver_queue_load_max is set with deliver_drop_privilege. This is
4693 necessary for the benefit of the queue runner, because there is no re-exec
4694 when deliver_drop_privilege is set.
4696 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4697 was in use, was not putting the data itself into the right store pool;
4698 consequently, it could be overwritten for a subsequent message in the same
4699 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4702 38. Sometimes the final signoff response after QUIT could fail to get
4703 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4704 < 0 before doing a fflush(). This bug looks as though it goes back to the
4705 introduction of TLS in release 3.20, but "sometimes" must have been rare
4706 because the tests only now provoked it.
4712 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4713 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4714 Now it returns an integer. A number of calls were still expecting a T/F
4715 return. Fortuitously, in all cases, the tests worked in OK situations,
4716 which is the norm. However, things would have gone wrong on any write
4717 failures on the smtp file descriptor. This function is used when sending
4718 messages over SMTP and also when doing verify callouts.
4720 2. When Exim is called to do synchronous delivery of a locally submitted
4721 message (the -odf or -odi options), it no longer closes stderr before doing
4724 3. Implemented the mua_wrapper option.
4726 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4728 5. Implemented the functions header_remove(), header_testname(),
4729 header_add_at_position(), and receive_remove_recipient(), and exported them
4732 6. If an ACL "warn" statement specified the addition of headers, Exim already
4733 inserted X-ACL-Warn: at the start if there was no header name. However, it
4734 was not making this test for the second and subsequent header lines if
4735 there were newlines in the string. This meant that an invalid header could
4736 be inserted if Exim was badly configured.
4738 7. Allow an ACL "warn" statement to add header lines at the start or after all
4739 the Received: headers, as well as at the end.
4741 8. Added the rcpt_4xx retry error code.
4743 9. Added postmaster_mailfrom=xxx to callout verification option.
4745 10. Added mailfrom=xxxx to the callout verification option, for verify=
4748 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4749 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4751 12. Inserted some casts to stop certain compilers warning when using pointer
4752 differences as field lengths or precisions in printf-type calls (mostly
4753 affecting debugging statements).
4755 13. Added optional readline() support for -be (dynamically loaded).
4757 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4758 same clock tick as a message's arrival, so that its received time was the
4759 same as the "first fail" time on the retry record, and that message
4760 remained on the queue past the ultimate address timeout, every queue runner
4761 would try a delivery (because it was past the ultimate address timeout) but
4762 after another failure, the ultimate address timeout, which should have then
4763 bounced the address, did not kick in. This was a "< instead of <=" error;
4764 in most cases the first failure would have been in the next clock tick
4765 after the received time, and all would be well.
4767 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4768 being recognized when the domain list was tested by the match_domain
4769 condition in an expansion string.
4771 16. Added the ${str2b64: operator.
4773 17. Exim was always calling setrlimit() to set a large limit for the number of
4774 processes, without checking whether the existing limit was already
4775 adequate. (It did check for the limit on file descriptors.) Furthermore,
4776 errors from getrlimit() and setrlimit() were being ignored. Now they are
4777 logged to the main and panic logs, but Exim does carry on, to try to do its
4778 job under whatever limits there are.
4780 18. Imported PCRE 5.0.
4782 19. Trivial typo in log message " temporarily refused connection" (the leading
4785 20. If the log selector return_path_on_delivery was set and an address was
4786 redirected to /dev/null, the delivery process crashed because it assumed
4787 that a return path would always be set for a "successful" delivery. In this
4788 case, the whole delivery is bypassed as an optimization, and therefore no
4791 21. Internal re-arrangement: the function for sending a challenge and reading
4792 a response while authentication was assuming a zero-terminated challenge
4793 string. It's now changed to take a pointer and a length, to allow for
4794 binary data in such strings.
4796 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4798 23. Exim was not respecting finduser_retries when seeking the login of the
4799 uid under which it was called; it was always trying 10 times. (The default
4800 setting of finduser_retries is zero.) Also, it was sleeping after the final
4801 failure, which is pointless.
4803 24. Implemented tls_on_connect_ports.
4805 25. Implemented acl_smtp_predata.
4807 26. If the domain in control=submission is set empty, Exim assumes that the
4808 authenticated id is a complete email address when it generates From: or
4809 Sender: header lines.
4811 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4812 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4813 chown and chgrp in /bin and hostname in /usr/bin.
4815 28. Exim was keeping the "process log" file open after each use, just as it
4816 does for the main log. This opens the possibility of it remaining open for
4817 long periods when the USR1 signal hits a daemon. Occasional processlog
4818 errors were reported, that could have been caused by this. Anyway, it seems
4819 much more sensible not to leave this file open at all, so that is what now
4822 29. The long-running daemon process does not normally write to the log once it
4823 has entered its main loop, and it closes the log before doing so. This is
4824 so that log files can straightforwardly be renamed and moved. However,
4825 there are a couple of unusual error situations where the daemon does write
4826 log entries, and I had neglected to close the log afterwards.
4828 30. The text of an SMTP error response that was received during a remote
4829 delivery was being truncated at 512 bytes. This is too short for some of
4830 the long messages that one sometimes sees. I've increased the limit to
4833 31. It is now possible to make retry rules that apply only when a message has a
4834 specific sender, in particular, an empty sender.
4836 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4837 it possible to be selective about when SMTP synchronization is enforced.
4839 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4841 32. Implemented hosts_connection_nolog.
4843 33. Added an ACL for QUIT.
4845 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4848 35. Added mailbox_size and mailbox_filecount to appendfile.
4850 36. Added control = no_multiline_responses to ACLs.
4852 37. There was a bug in the logic of the code that waits for the clock to tick
4853 in the case where the clock went backwards by a substantial amount such
4854 that the microsecond fraction of "now" was more than the microsecond
4855 fraction of "then" (but the whole seconds number was less).
4857 38. Added support for the libradius Radius client library this is found on
4858 FreeBSD (previously only the radiusclient library was supported).
4864 1. When certain lookups returned multiple values in the form name=value, the
4865 quoting of the values was not always being done properly. Specifically:
4866 (a) If the value started with a double quote, but contained no whitespace,
4868 (b) If the value contained whitespace other than a space character (i.e.
4869 tabs or newlines or carriage returns) it was not quoted.
4870 This fix has been applied to the mysql and pgsql lookups by writing a
4871 separate quoting function and calling it from the lookup code. The fix
4872 should probably also be applied to nisplus, ibase and oracle lookups, but
4873 since I cannot test any of those, I have not disturbed their existing code.
4875 2. A hit in the callout cache for a specific address caused a log line with no
4876 reason for rejecting RCPT. Now it says "Previous (cached) callout
4877 verification failure".
4879 3. There was an off-by-one bug in the queryprogram router. An over-long
4880 return line was truncated at 256 instead of 255 characters, thereby
4881 overflowing its buffer with the terminating zero. As well as fixing this, I
4882 have increased the buffer size to 1024 (and made a note to document this).
4884 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4885 when Exim is waiting for an SMTP response from a remote server, Exim
4886 restarts its select() call on the socket, thereby resetting its timeout.
4887 This is not a problem when such interrupts are rare. Somebody set up a cron
4888 job to run exiwhat every 2 minutes, which is less than the normal select()
4889 timeout (5 or 10 minutes). This meant that the select() timeout never
4890 kicked in because it was always reset. I have fixed this by comparing the
4891 time when an interrupt arrives with the time at the start of the first call
4892 to select(). If more time than the timeout has elapsed, the interrupt is
4893 treated as a timeout.
4895 5. Some internal re-factoring in preparation for the addition of Sieve
4896 extensions (by MH). In particular, the "personal" test is moved to a
4897 separate function, and given an option for scanning Cc: and Bcc: (which is
4898 not set for Exim filters).
4900 6. When Exim created an email address using the login of the caller as the
4901 local part (e.g. when creating a From: or Sender: header line), it was not
4902 quoting the local part when it contained special characters such as @.
4904 7. Installed new OpenBSD configuration files.
4906 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4907 try to make them clearer.
4909 9. Callout options, other than the timeout value, were being ignored when
4910 verifying sender addresses in header lines. For example, when using
4912 verify = header_sender/callout=no_cache
4914 the cache was (incorrectly) being used.
4916 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4917 only the Cygwin environment.
4919 11. When return_path_on_delivery was set as a log selector, if different remote
4920 addresses in the same message used different return paths and parallel
4921 remote delivery occurred, the wrong values would sometimes be logged.
4922 (Whenever a remote delivery process finished, the return path value from
4923 the most recently started remote delivery process was logged.)
4925 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4926 lines when AUTH and/or TLS are in use. This is the "received protocol"
4927 field. Exim used to use "asmtp" for authenticated SMTP, without any
4928 indication (in the protocol name) for TLS use. Now it follows the RFC and
4929 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4930 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4931 names appear in log lines as well as in Received: header lines.
4933 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4934 extensions, and comparison tests, and to fix some bugs.
4936 14. Changes to the "personal" filter test:
4938 (1) The test was buggy in that it was just doing the equivalent of
4939 "contains" tests on header lines. For example, if a user's address was
4940 anne@some.where, the "personal" test would incorrectly be true for
4942 To: susanne@some.where
4944 This test is now done by extracting each address from the header in turn,
4945 and checking the entire address. Other tests that are part of "personal"
4946 are now done using regular expressions (for example, to check local parts
4947 of addresses in From: header lines).
4949 (2) The list of non-personal local parts in From: addresses has been
4950 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4951 taken from the Sieve specification recommendations.
4953 (3) If the message contains any header line starting with "List-" it is
4954 treated as non-personal.
4956 (4) The test for "circular" in the Subject: header line has been removed
4957 because it now seems ill-conceived.
4959 15. Minor typos in src/EDITME comments corrected.
4961 16. Installed latest exipick from John Jetmore.
4963 17. If headers_add on a router specified a text string that was too long for
4964 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4965 of string_sprintf() is now avoided.
4967 18. $message_body_size was not set (it was always zero) when running the DATA
4968 ACL and the local_scan() function.
4970 19. For the "mail" command in an Exim filter, no default was being set for
4971 the once_repeat time, causing a random time value to be used if "once" was
4972 specified. (If the value happened to be <= 0, no repeat happened.) The
4973 default is now 0s, meaning "never repeat". The "vacation" command was OK
4974 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4975 (I found it when inspecting the code).
4977 20. There is now an overall timeout for performing a callout verification. It
4978 defaults to 4 times the callout timeout, which applies to individual SMTP
4979 commands during the callout. The overall timeout applies when there is more
4980 than one host that can be tried. The timeout is checked before trying the
4981 next host. This prevents very long delays if there are a large number of
4982 hosts and all are timing out (e.g. when the network connections are timing
4983 out). The value of the overall timeout can be changed by specifying an
4984 additional sub-option for "callout", called "maxwait". For example:
4986 verify = sender/callout=5s,maxwait=20s
4988 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4989 to the end before writing, but this should make it even safer).
4991 22. Exim was forgetting that it had advertised PIPELINING for the second and
4992 subsequent messages on an SMTP connection. It was also not resetting its
4993 memory on STARTTLS and an internal HELO.
4995 23. When Exim logs an SMTP synchronization error within a session, it now
4996 records whether PIPELINING has been advertised or not.
4998 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4999 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5000 rather than long int.
5002 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5004 26. Added the never_mail option to autoreply.
5010 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5011 crash if the getsockname() call failed; this can happen if a connection is
5012 closed very soon after it is established. The problem was simply in the
5013 order in which certain operations were done, causing Exim to try to write
5014 to the SMTP stream before it had set up the file descriptor. The bug has
5015 been fixed by making things happen in the correct order.
5021 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5022 before the connection was closed, thus losing the rejection response.
5024 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5025 some early Solaris releases, but causes trouble in current releases where
5026 socklen_t is defined.
5028 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5031 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5034 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5035 that releasing the top of it at the end releases what was used for sub-
5036 expansions (unless the block got too big). However, discard this block if
5037 the first thing is a variable or header, so that we can use its block when
5038 it is dynamic (useful for very large $message_headers, for example).
5040 6. Lookups now cache *every* query, not just the most recent. A new, separate
5041 store pool is used for this. It can be recovered when all lookup caches are
5042 flushed. Lookups now release memory at the end of their result strings.
5043 This has involved some general refactoring of the lookup sources.
5045 7. Some code has been added to the store_xxx() functions to reduce the amount
5046 of flapping under certain conditions.
5048 8. log_incoming_interface used to affect only the <= reception log lines. Now
5049 it causes the local interface and port to be added to several more SMTP log
5050 lines, for example "SMTP connection from", and rejection lines.
5052 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5054 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5056 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5057 could be overwritten at the end of the current message (or the start of a
5058 new message if it was set in a HELO ACL). The value is now preserved for
5059 the duration of the SMTP connection.
5061 12. If a transport had a headers_rewrite setting, and a matching header line
5062 contained an unqualified address, that address was qualified, even if it
5063 did not match any rewriting rules. The underlying bug was that the values
5064 of the flags that permit the existence of unqualified sender and recipient
5065 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5066 non-local messages, and by -bnq for local messages) were not being
5067 preserved with the message after it was received.
5069 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5070 "next input=" as part of the text comprising the host identity instead of
5071 the correct text. The code was using the same buffer for two different
5072 strings. However, depending on which order the printing function evaluated
5073 its arguments, the bug did not always show up. Under Linux, for example, my
5074 test suite worked just fine.
5076 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5077 doesn't work with some older versions of Perl. It has been changed to "my",
5078 which in any case is probably the better facility to use.
5080 15. A really picky compiler found some instances of statements for creating
5081 error messages that either had too many or two few arguments for the format
5084 16. The size of the buffer for calls to the DNS resolver has been increased
5085 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5086 for addresses that have a lot of PTR records. This alleviates a problem; it
5087 does not fully solve it.
5089 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5090 buffer now truncates the list and logs the incident, which is the same
5091 action as happens when Exim is looking up a host name and its aliases.
5092 Previously in this situation something unpredictable would happen;
5093 sometimes it was "internal error: store_reset failed".
5095 18. If a server dropped the connection unexpectedly when an Exim client was
5096 using GnuTLS and trying to read a response, the client delivery process
5097 crashed while trying to generate an error log message.
5099 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5100 string, for example:
5102 warn message = H1: something\nH2: something
5104 the text was added as a single header line from Exim's point of view
5105 though it ended up OK in the delivered message. However, searching for the
5106 second and subsequent header lines using $h_h2: did not work. This has been
5107 fixed. Similarly, if a system filter added multiple headers in this way,
5108 the routers could not see them.
5110 20. Expanded the error message when iplsearch is called with an invalid key to
5111 suggest using net-iplsearch in a host list.
5113 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5114 ACLs are no longer actually imposed (and a message to that effect is
5117 22. If a "gecos" field in a passwd entry contained escaped characters, in
5118 particular, if it contained a \" sequence, Exim got it wrong when building
5119 a From: or a Sender: header from that name. A second bug also caused
5120 incorrect handling when an unquoted " was present following a character
5121 that needed quoting.
5123 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5124 was not being matched caselessly.
5126 24. Arranged for all hyphens in the exim.8 source to be escaped with
5129 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5130 properly. Recipient callout cache records were still being keyed to include
5131 the sender, even when use_sender was set false. This led to far more
5132 callouts that were necessary. The sender is no longer included in the key
5133 when use_sender is false.
5135 26. Added "control = submission" modifier to ACLs.
5137 27. Added the ${base62d: operator to decode base 62 numbers.
5139 28. dnsdb lookups can now access SRV records.
5141 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5142 the configuration file.
5144 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5145 (-v) mode. This makes the output for a verbose queue run more intelligible.
5147 31. Added a use_postmaster feature to recipient callouts.
5149 32. Added the $body_zerocount variable, containing the number of binary zero
5150 bytes in the message body.
5152 33. The time of last modification of the "new" subdirectory is now used as the
5153 "mailbox time last read" when there is a quota error for a maildir
5156 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5158 35. Added +ignore_unknown as a special item in host lists.
5160 36. Code for decoding IPv6 addresses in host lists is now included, even if
5161 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5162 address was recognized as an IP address, but was then not correctly decoded
5163 into binary, causing unexpected and incorrect effects when compared with
5170 1. Very minor rewording of debugging text in manualroute to say "list of
5171 hosts" instead of "hostlist".
5173 2. If verify=header_syntax was set, and a header line with an unqualified
5174 address (no domain) and a large number of spaces between the end of the
5175 name and the colon was received, the reception process suffered a buffer
5176 overflow, and (when I tested it) crashed. This was caused by some obsolete
5177 code that should have been removed. The fix is to remove it!
5179 3. When running in the test harness, delay a bit after writing a bounce
5180 message to get a bit more predictability in the log output.
5182 4. Added a call to search_tidyup() just before forking a reception process. In
5183 theory, someone could use a lookup in the expansion of smtp_accept_max_
5184 per_host which, without the tidyup, could leave open a database connection.
5186 5. Added the variables $recipient_data and $sender_data which get set from a
5187 lookup success in an ACL "recipients" or "senders" condition, or a router
5188 "senders" option, similar to $domain_data and $local_part_data.
5190 6. Moved the writing of debug_print from before to after the "senders" test
5193 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5194 problems for message scanning, either using a data ACL, or using
5195 local_scan() because the Received: header was not generated till after they
5196 were called (in order to set the time as the time of reception completion).
5197 I have revised the way this works. The header is now generated after the
5198 body is received, but before the ACL or local_scan() are called. After they
5199 are run, the timestamp in the header is updated.
5205 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5206 before starting a queue runner without re-exec. This happened only when
5207 deliver_drop_privilege was set or when the Exim user was set to root. The
5208 effect of the bug was that timeouts during subsequent deliveries caused
5209 crashes instead of being properly handled. The handler is now left at its
5210 default (and expected) setting.
5212 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5213 message, again when deliver_drop_privilege is set or Exim is run as root.
5214 The bug described in (1) was not present in this case, but the tidying up
5215 of the other signals was missing. I have made the two cases consistent.
5217 3. The ignore_target_hosts setting on a manualroute router was being ignored
5218 for hosts that were looked up using the /MX notation.
5220 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5223 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5224 operated on the sender address. After changing the $sender_address to <>
5225 for the sender address verify, Exim was re-instated it as the original
5226 (before rewriting) address, but remembering that it had rewritten it, so it
5227 wasn't rewriting it again. This bug also had the effect of breaking the
5228 sender address verification caching when the sender address was rewritten.
5230 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5231 This has been changed so that if the ip literal address matches
5232 ignore_target_hosts, the router declines.
5234 7. Added expansion conditions match_domain, match_address, and match_local_
5235 part (NOT match_host).
5237 8. The placeholder for the Received: header didn't have a length field set.
5239 9. Added code to Exim itself and to exim_lock to test for a specific race
5240 condition that could lead to file corruption when using MBX delivery. The
5241 issue is with the lockfile that is created in /tmp. If this file is removed
5242 after a process has opened it but before that process has acquired a lock,
5243 there is the potential for a second process to recreate the file and also
5244 acquire a lock. This could lead to two Exim processes writing to the file
5245 at the same time. The added code performs the same test as UW imapd; it
5246 checks after acquiring the lock that its file descriptor still refers to
5247 the same named file.
5249 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5250 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5253 11. Added the smtp_active_hostname option. If used, this will typically be made
5254 to depend on the incoming interface address. Because $interface_address is
5255 not set up until the daemon has forked a reception process, error responses
5256 that can happen earlier (such as "too many connections") no longer contain
5259 12. If an expansion in a condition on a "warn" statement fails because a lookup
5260 defers, the "warn" statement is abandoned, and the next ACL statement is
5261 processed. Previously this caused the whole ACL to be aborted.
5263 13. Added the iplsearch lookup type.
5265 14. Added ident_timeout as a log selector.
5267 15. Added tls_certificate_verified as a log selector.
5269 16. Added a global option tls_require_ciphers (compare the smtp transport
5270 option of the same name). This controls incoming TLS connections.
5272 17. I finally figured out how to make tls_require_ciphers do a similar thing
5273 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5274 before starting the TLS session.
5276 18. Tabs are now shown as \t in -bP output.
5278 19. If the log selector return_path_on_delivery was set, Exim crashed when
5279 bouncing a message because it had too many Received: header lines.
5281 20. If two routers both had headers_remove settings, and the first one included
5282 a superfluous trailing colon, the final name in the first list and the
5283 first name in the second list were incorrectly joined into one item (with a
5284 colon in the middle).
5290 1. Added -C and -D options to the exinext utility, mainly to make it easier
5291 to include in the automated testing, but these could be helpful when
5292 multiple configurations are in use.
5294 2. The exinext utility was not formatting the output nicely when there was
5295 an alternate port involved in the retry record key, nor when there was a
5296 message id as well (for retries that were specific to a specific message
5297 and a specific host). It was also confused by IPv6 addresses, because of
5298 the additional colons they contain. I have fixed the IPv4 problem, and
5299 patched it up to do a reasonable job for IPv6.
5301 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5302 delivery, the log line now contains "pipelined" if PIPELINING was used.
5304 4. An SMTP transport process used to panic and die if the bind() call to set
5305 an explicit outgoing interface failed. This has been changed; it is now
5306 treated in the same way as a connect() failure.
5308 5. A reference to $sender_host_name in the part of a conditional expansion
5309 that was being skipped was still causing a DNS lookup. This no longer
5312 6. The def: expansion condition was not recognizing references to header lines
5313 that used bh_ and bheader_.
5315 7. Added the _cache feature to named lists.
5317 8. The code for checking quota_filecount in the appendfile transport was
5318 allowing one more file than it should have been.
5320 9. For compatibility with Sendmail, the command line option
5328 and sets the incoming protocol and host name (for trusted callers). The
5329 host name and its colon can be omitted when only the protocol is to be set.
5330 Note the Exim already has two private options, -pd and -ps, that refer to
5331 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5332 "s", but I don't think that's a major issue.
5334 10. A number of refactoring changes to the code, none of which should affect
5337 (a) The number of logging options was getting close to filling up the
5338 32-bit word that was used as a bit map. I have split them into two classes:
5339 those that are passed in the argument to log_write(), and those that are
5340 only ever tested independently outside of that function. These are now in
5341 separate 32-bit words, so there is plenty of room for expansion again.
5342 There is no change in the user interface or the logging behaviour.
5344 (b) When building, for example, log lines, the code previously used a
5345 macro that called string_cat() twice, in order to add two strings. This is
5346 not really sufficiently general. Furthermore, there was one instance where
5347 it was actually wrong because one of the argument was used twice, and in
5348 one call a function was used. (As it happened, calling the function twice
5349 did not affect the overall behaviour.) The macro has been replaced by a
5350 function that can join an arbitrary number of extra strings onto a growing
5353 (c) The code for expansion conditions now uses a table and a binary chop
5354 instead of a serial search (which was left over from when there were very
5355 few conditions). Also, it now recognizes conditions like "pam" even when
5356 the relevant support is not compiled in: a suitably worded error message is
5357 given if an attempt is made to use such a condition.
5359 11. Added ${time_interval:xxxxx}.
5361 12. A bug was causing one of the ddress fields not to be passed back correctly
5362 from remote delivery subprocesses. The field in question was not being
5363 subsequently used, so this caused to problems in practice.
5365 13. Added new log selectors queue_time and deliver_time.
5367 14. Might have fixed a bug in maildirsizefile handling that threw up
5368 "unexpected character" debug warnings, and recalculated the data
5369 unnecessarily. In any case, I expanded the warning message to give more
5372 15. Added the message "Restricted characters in address" to the statements in
5373 the default ACL that block characters like @ and % in local parts.
5375 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5376 Three changes have been made:
5378 (a) There was a serious bug; a negative response to MAIL caused the whole
5379 recipient domain to be cached as invalid, thereby blocking all messages
5380 to all local parts at the same domain, from all senders. This bug has
5381 been fixed. The domain is no longer cached after a negative response to
5382 MAIL if the sender used is not empty.
5384 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5387 (c) A new callout option, "use_sender" has been added for people who want
5388 the modified behaviour.
5394 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5397 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5398 indeed breaks things for older releases.
5400 3. Added additional logging to the case where there is a problem reading data
5401 from a filter that is running in a subprocess using a pipe, in order to
5402 try to track down a specific problem.
5404 4. Testing facility fudge: when running in the test harness and attempting
5405 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5406 getting "No route to host". Convert this to a timeout.
5408 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5411 6. Some OS don't have socklen_t but use size_t instead. This affects the
5412 fifth argument of getsockopt() amongst other things. This is now
5413 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5414 can be set for individual OS. I have set it for SunOS5, OSF1, and
5415 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5416 some earlier ones do not.
5418 7. Change 4.30/15 was not doing the test caselessly.
5420 8. The standard form for an IPv6 address literal was being rejected by address
5421 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5422 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5423 this, as well as the form without the "IPv6" on the front (but only when
5424 address literals are enabled, of course).
5426 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5428 10. Exim crashed if a message with an empty sender address specified by -f
5429 encountered a router with an errors_to setting. This could be provoked only
5430 by a command such as
5434 where an empty string was supplied; "<>" did not hit this bug.
5436 11. Installed PCRE release 4.5.
5438 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5439 remained set. It is now erased.
5441 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5442 times from message ids (which are base 36 rather than the normal 62).
5444 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5445 were being counted as actual protocol errors, and logged if the log
5446 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5447 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5448 and DATA following a set of rejected RCPTs do not count as protocol errors.
5449 In other words, Exim assumes they were pipelined, though this may not
5450 actually be the case. Of course, in all cases the client gets an
5451 appropriate error code.
5453 15. If a lookup fails in an ACL condition, a message about the failure may
5454 be available; it is used if testing the ACL cannot continue, because most
5455 such messages specify what the cause of the deferral is. However, some
5456 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5457 that caused an old message to be retained and used if a later statement
5458 caused a defer, replacing the real cause of the deferral.
5460 16. If an IP address had so many PTR records that the DNS lookup buffer
5461 was not large enough to hold them, Exim could crash while trying to process
5462 the truncated data. It now detects and logs this case.
5464 17. Further to 4.21/58, another change has been made: if (and only if) the
5465 first line of a message (the first header line) ends with CRLF, a bare LF
5466 in a subsequent header line has a space inserted after it, so as not to
5467 terminate the header.
5469 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5470 unnecessarily, used atoi() instead of strtol(), and didn't check the
5471 termination when getting file sizes from file names by regex.
5473 19. Completely re-implemented the support for maildirsize files, in the light
5474 of a number of problems with the previous contributed implementation
5475 (4.30/29). In particular:
5477 . If the quota is zero, the maildirsize file is maintained, but no quota is
5480 . If the maildir directory does not exist, it is created before any attempt
5481 to write a maildirsize file.
5483 . The quota value in the file is just a cache; if the quota is changed in
5484 the transport, the new value overrides.
5486 . A regular expression is available for excluding directories from the
5489 20. The autoreply transport checks the characters in options that define the
5490 message's headers; it allows continued headers, but it was checking with
5491 isspace() after an embedded newline instead of explicitly looking for a
5494 21. If all the "regular" hosts to which an address was routed had passed their
5495 expiry times, and had not reached their retry times, the address was
5496 bounced, even if fallback hosts were defined. Now Exim should go on to try
5499 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5500 equivalent code in the SMTP transport. Some hosts send humungous responses
5501 to HELO/EHLO, more than 1024 it seems.
5503 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5504 gives compiler warnings in some environments. I've now done it "properly",
5507 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5508 (because of problems with the built-in one) was declared to return uschar *
5509 instead of char *, causing compiler failure.
5511 25. Fixed a file descriptor leak when processing alias/forward files.
5513 26. Fixed a minor format string issue in dbfn.c.
5515 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5517 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5518 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5521 29. When a sender address is verified, it is cached, to save repeating the test
5522 when there is more than one recipient in a message. However, when the
5523 verification involves a callout, it is possible for different callout
5524 options to be set for different recipients. It is too complicated to keep
5525 track of this in the cache, so now Exim always runs a verification when a
5526 callout is required, relying on the callout cache for the optimization.
5527 The overhead is duplication of the address routing, but this should not be
5530 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5531 to be verified with callout=postmaster, and the main callout worked but the
5532 postmaster check failed, the verification correctly failed. However, if a
5533 subsequent RCPT command asked for sender verification *without* the
5534 postmaster check, incorrect caching caused this verification also to fail,
5537 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5538 it was not caching the DNS options (qualify_single, search_parents) that
5539 were used when the lookup failed. A subsequent lookup with different
5540 options therefore always gave the same answer, though there were cases
5541 where it should not have. (Example: a "domains = !$mx_any" option on a
5542 dnslookup router: the "domains" option is always processed without any
5543 widening, but the router might have qualify_single set.) Now Exim uses the
5544 cached value only when the same options are set.
5546 32. Added John Jetmore's "exipick" utility to the distribution.
5548 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5549 than a timeout (e.g. a certificate is required, and is not provided), an
5550 Exim server now closes the connection immediately. Previously it waited for
5551 the client to close - but if the client is SSL, it seems that they each
5552 wait for each other, leading to a delay before one of them times out.
5554 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5555 maintained 0.8.x compatibility because I don't think many are using it, and
5556 it is clearly obsolete.
5558 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5561 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5562 client certificate was expired. A simple patch fixes this, though I don't
5563 understand the full logic of why the verify callback is called multiple
5566 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5567 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5568 which causes problems with some clients (such as the Certicom SSL Plus
5569 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5570 disables the coutermeasure allowing Eudora to connect."
5572 38. Exim was not checking that a write() to a log file succeeded. This could
5573 lead to Bad Things if a log got too big, in particular if it hit a file
5574 size limit. Exim now panics and dies if it cannot write to a log file, just
5575 as it does if it cannot open a log file.
5577 39. Modified OS/Makefile-Linux so that it now contains
5579 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5581 The two -D definitions ensure that Exim is compiled with large file
5582 support, which makes it possible to handle log files that are bigger than
5585 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5586 instance) a domain was checked against a named list that involved a lookup,
5587 causing $domain_data to be set, then another domain was checked against the
5588 same list, then the first domain was re-checked, the value of $domain_data
5589 after the final check could be wrong. In particular, if the second check
5590 failed, it could be set empty. This bug probably also applied to
5593 41. The strip_trailing_dot option was not being applied to the address given
5594 with the -f command-line option.
5596 42. The code for reading a message's header from the spool was incrementing
5597 $received_count, but never initializing it. This meant that the value was
5598 incorrect (doubled) while delivering a message in the same process in which
5599 it was received. In the most common configuration of Exim, this never
5600 happens - a fresh exec is done - but it can happen when
5601 deliver_drop_privilege is set.
5603 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5604 it now includes up to 150 characters of the unexpected data in the log
5607 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5608 and building data strings. The size of both of these buffers was 10 000
5609 bytes - far larger than anybody would *ever* want, thought I. Needless to
5610 say, somebody hit the limit. I have increased the maximum line length to
5611 20 000 and the maximum data length of concatenated lines to 100 000. I have
5612 also fixed two bugs, because there was no checking on these buffers. Tsk,
5613 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5614 buffer is too small.
5616 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5617 lsearch lookups. Now it does.
5619 46. When parsing a route_list item in a manualroute router, a fixed-length
5620 buffer was used for the list of hosts. I made this 1024 bytes long,
5621 thinking that nobody would ever have a list of hosts that long. Wrong.
5622 Somebody had a whole pile of complicated expansion conditions, and the
5623 string was silently truncated, leading to an expansion error. It turns out
5624 that it is easier to change to an unlimited length (owing to other changes
5625 that have happened since this code was originally written) than to build
5626 structure for giving a limitation error. The length of the item that
5627 expands into the list of hosts is now unlimited.
5629 47. The lsearch lookup could not handle data where the length of text line was
5630 more than 4095 characters. Such lines were truncated, leading to shortened
5631 data being returned. It should now handle lines of any length.
5633 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5634 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5637 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5638 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5639 no longer happens in comment lines. A list of replacements is now placed
5640 at the head of all of the source files, except those whose only change is
5641 to replace PERL_COMMAND in the very first #! line.
5643 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5644 messages on the queue, with a bottom-up merge sort, using code contributed
5645 by Michael Haardt. This should make operations like -bp somewhat faster on
5646 large queues. It won't affect queue runners, except when queue_run_in_order
5649 51. Installed eximstats 1.31 in the distribution.
5651 52. Added support for SRV lookups to the dnslookup router.
5653 53. If an ACL referred to $message_body or $message_body_end, the value was not
5654 reset for any messages that followed in the same SMTP session.
5656 54. The store-handling optimization for building very long strings was not
5657 differentiating between the different store pools. I don't think this
5658 actually made any difference in practice, but I've tidied it.
5660 55. While running the routers to verify a sender address, $sender_address
5661 was still set to the sender address. This is wrong, because when routing to
5662 send a bounce to the sender, it would be empty. Therefore, I have changed
5663 it so that, while verifying a sender address, $sender_address is set to <>.
5664 (There is no change to what happens when verifying a recipient address.)
5666 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5667 target A or AAAA records (if not already returned) without resetting the
5668 qualify_single or search_parents options of the DNS resolver. These are
5669 inappropriate in this case because the targets of MX and SRV records must
5670 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5671 target that, when qualified, matched something in the local domain. These
5672 two options are now turned off when doing these lookups.
5674 57. It seems that at least some releases of Reiserfs (which does not have the
5675 concept of a fixed number of inodes) returns zero and not -1 for the
5676 number of available inodes. This interacted badly with check_spool_inodes,
5677 which assumed that -1 was the "no such thing" setting. What I have done is
5678 to check that the total number of inodes is greater than zero before doing
5679 the test of how many are available.
5681 58. When a "warn" ACL statement has a log_message modifier, the message is
5682 remembered, and not repeated. This is to avoid a lot of repetition when a
5683 message has many recipients that cause the same warning to be written.
5684 However, Exim was preserving the list of already written lines for an
5685 entire SMTP session, which doesn't seem right. The memory is now reset if a
5686 new message is started.
5688 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5689 debugging output unless log_rewrite was also set.
5691 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5692 of (int)(handle) when we know that handle contains (void *)(-1).
5694 61. The Exim daemon panic-logs an error return when it closes the incoming
5695 connection. However "connection reset by peer" seems to be common, and
5696 isn't really an error worthy of noting specially, so that particular error
5699 62. When Exim is trying to find all the local interfaces, it used to panic and
5700 die if the ioctl to get the interface flags failed. However, it seems that
5701 on at least one OS (Solaris 9) it is possible to have an interface that is
5702 included in the list of interfaces, but for which you get a failure error
5703 for this call. This happens when the interface is not "plumbed" into a
5704 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5705 failure of the "get flags" call assumes that the interface is down.
5707 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5708 makes life easier for people who are doing arithmetic on fields extracted
5709 from dates, where you often get leading zeros that should not be
5710 interpreted as octal.
5712 64. Added qualify_domain to the redirect router, to override the global
5715 65. If a pathologically long header line contained very many addresses (the
5716 report of this problem mentioned 10 000) and each of them was rewritten,
5717 Exim could use up a very large amount of memory. (It kept on making new
5718 copies of the header line as it rewrote, and never released the old ones.)
5719 At the expense of a bit more processing, the header rewriting function has
5720 been changed so that it no longer eats memory in this way.
5722 66. The generation of the Received: header has been moved from the time that a
5723 message starts to be received, to the time that it finishes. The timestamp
5724 in the Received: header should now be very close to that of the <= log
5725 line. There are two side-effects of this change:
5727 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5728 logged header lines no longer include the local Received: line, because
5729 it has not yet been created. The same applies to a copy of the message
5730 that is returned to a non-SMTP sender when a message is rejected.
5732 (b) When a filter file is tested using -bf, no additional Received: header
5733 is added to the test message. After some thought, I decided that this
5736 This change does not affect the value of $received_for. It is still set
5737 after address rewriting, but before local_scan() is called.
5739 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5741 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5742 gave an unhelpful panic error message, and a defer error. I have managed to
5743 change this behaviour so that it now rejects any supplied certificate,
5744 which seems right, as the list of acceptable certificates is empty.
5746 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5747 gave an unhelpful defer error. I have not managed to make this reject any
5748 supplied certificates, but the error message it gives is "no certificate
5749 supplied", which is not helpful.
5751 70. exigrep's output now also includes lines that are not associated with any
5752 message, but which match the given pattern. Implemented by a patch from
5753 Martin Sluka, which also tidied up the Perl a bit.
5755 71. Recipient callout verification, like sender verification, was using <> in
5756 the MAIL FROM command. This isn't really the right thing, since the actual
5757 sender may affect whether the remote host accepts the recipient or not. I
5758 have changed it to use the actual sender in the callout; this means that
5759 the cache record is now keyed on a recipient/sender pair, not just the
5760 recipient address. There doesn't seem to be a real danger of callout loops,
5761 since a callout by the remote host to check the sender would use <>.
5762 [SEE ABOVE: changed after hitting problems.]
5764 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5765 temporary errors. However, in the case of such a code being given after
5766 the end of a data transmission (i.e. after ".") Exim was failing to write
5767 a retry record for the message. (Yes, there was some broken host that was
5768 actually sending 8xx at this point.)
5770 73. An unknown lookup type in a host list could cause Exim to panic-die when
5771 the list was checked. (An example that provoked this was putting <; in the
5772 middle of a list instead of at the start.) If this happened during a DATA
5773 ACL check, a -D file could be left lying around. This kind of configuration
5774 error no longer causes Exim to die; instead it causes a defer error. The
5775 incident is still logged to the main and panic logs.
5777 74. Buglet left over from Exim 3 conversion. The message "too many messages
5778 in one connection" was written to the rejectlog but not the mainlog, except
5779 when address rewriting (yes!) was being logged.
5781 75. Added write_rejectlog option.
5783 76. When a system filter was run not as root (that is, when system_filter_user
5784 was set), the values of the $n variables were not being returned to the
5785 main process; thus, they were not subsequently available in the $sn
5788 77. Added +return_path_on_delivery log selector.
5790 78. A connection timeout was being treated differently from recipients deferred
5791 when testing hosts_max_try with a message that was older than the host's
5792 retry timeout. (The host should not be counted, thus allowing all hosts to
5793 be tried at least once before bouncing.) This may have been the cause of an
5794 occasionally reported bug whereby a message would remain on the queue
5795 longer than the retry timeout, but would be bounced if a delivery was
5796 forced. I say "may" because I never totally pinned down the problem;
5797 setting up timeout/retry tests is difficult. See also the next item.
5799 79. The ultimate address timeout was not being applied to errors that involved
5800 a combination of host plus message (for example, a timeout on a MAIL
5801 command). When an address resolved to a number of possible hosts, and they
5802 were not all tried for each delivery (e.g. because of hosts_max_try), a
5803 message could remain on the queue longer than the retry timeout.
5805 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5808 81. Fixed an obscure SMTP outgoing bug which required at least the following
5809 conditions: (a) there was another message waiting for the same server;
5810 (b) the server returned 5xx to all RCPT commands in the first message so
5811 that the message was not completed; (c) the server dropped the connection
5812 or gave a negative response to the RSET that Exim sends to abort the
5813 transaction. The observed case was a dropped connection after DATA that had
5814 been sent in pipelining mode. That is, the server had advertised PIPELINING
5815 but was not implementing it correctly. The effect of the bug was incorrect
5816 behaviour, such as trying another host, and this could lead to a crash.
5822 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5823 and daemon.c were passed as pointers to ints; they should have been
5824 pointers to socklen_t variables (which are typically unsigned ints).
5826 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5829 3. Fixed a really odd bug that affected only the testing scheme; patching a
5830 certain fixed string in the binary changed the value of another string that
5831 happened to be identical to the end of the original first string.
5833 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5834 name", it returns that address as the IP address. On some operating
5835 systems (e.g. Solaris), it also passes back the IP address string as the
5836 "host name". However, on others (e.g. Linux), it passes back an empty
5837 string. Exim wasn't checking for this, and was changing the host name to an
5838 empty string, assuming it had been canonicalized.
5840 5. Although rare, it is permitted to have more than one PTR record for a given
5841 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5842 all the names associated with an address, because they do in Solaris.
5843 However, it seems that they do not in Linux for data that comes from the
5844 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5845 I found this out when I moved to a new Linux workstation and tried to run
5846 the Exim test suite.
5848 To get round this problem I have changed the code so that it now does its
5849 own call to the DNS to look up PTR records when searching for a host name.
5850 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5851 addresses that are only in /etc/hosts are still found.
5853 This behaviour is, however, controlled by an option called host_lookup_
5854 order, which defaults to "bydns:byaddr". If people want to use the other
5855 order, or indeed, just use one or the other means of lookup, they can
5856 specify it in this variable.
5858 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5859 some operating systems, this comes back from gethostbyaddr() as an empty
5860 string, and this is what Exim used to test for. However, it seems that in
5861 other systems, "." is yielded. Exim now tests for this case too.
5863 7. The values of check_spool_space and check_log_space are now held internally
5864 as a number of kilobytes instead of an absolute number of bytes. If a
5865 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5866 kilobyte. This means that much larger values can be stored.
5868 8. Exim monitor: an attempt to get the action menu when not actually pointing
5869 at a message produces an empty menu entitled "No message selected". This
5870 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5871 no entries in it ("Shell widget menu has zero width and/or height"). So I
5872 have added a single, blank menu entry in this case.
5874 9. Added ${quote_local_part.
5876 10. MIME decoding is now applied to the contents of Subject: header lines when
5879 11. Now that a reference to $sender_host_address automatically causes a reverse
5880 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5881 host lookup before query-style lookups in lists that might use this
5882 variable. This has therefore been abolished, and the "net-" prefix is no
5883 longer necessary for query-style lookups.
5885 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5886 have been a typo for LFLAGS, so it has been changed.
5888 13. The install script calls Exim with "-C /dev/null" in order to find the
5889 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5890 to be output. However, since Exim outputs its version number before the
5891 error, it didn't break the script. It just looked ugly. I fixed this by
5892 always allowing "-C /dev/null" if the caller is root.
5894 14. Ignore overlarge ACL variable number when reading spool file - insurance
5895 against a later release with more variables having written the file.
5897 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5898 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5899 this, as well as the form without the "IPv6" on the front.
5901 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5902 OS/Makefile-Darwin file.
5904 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5905 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5907 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5908 variable. However, some OS use ints for inodes. I've added cast to long int
5909 to get rid of the compiler warning.
5911 19. I had forgotten to lock out "/../" in configuration file names when
5912 ALT_CONFIG_PREFIX was set.
5914 20. Routers used for verification do not need to specify transports. However,
5915 if such a router generated a host list, and callout was configured, Exim
5916 crashed, because it could not find a port number from the (non-existent)
5917 transport. It now assumes port 25 in this circumstance.
5919 21. Added the -t option to exigrep.
5921 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5922 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5923 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5924 them is set, LOOKUP_LSEARCH is forced.
5926 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5928 24. Added sender and host information to the "rejected by local_scan()" log
5929 line; previously there was no indication of these.
5931 25. Added .include_if_exists.
5933 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5934 Linux. It turns out that not all file systems support this. Apparently some
5935 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5936 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5937 not supported on the file descriptor, is now ignored when Exim is trying to
5938 sync a directory. This applies only to Linux.
5940 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5942 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5943 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5946 29. Added support for maildirsize files from supplied patch (modified a bit).
5948 30. The use of :fail: followed by an empty string could lead Exim to respond to
5949 sender verification failures with (e.g.):
5951 550 Verification failed for <xxx>
5952 550 Sender verify failed
5954 where the first response line was missing the '-' that indicates it is not
5955 the final line of the response.
5957 31. The loop for finding the name of the user that called Exim had a hardwired
5958 limit of 10; it now uses the value of finduser_retries, which is used for
5959 all other user lookups.
5961 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5964 33. Exim was neglecting to zero errno before one call of strtol() when
5965 expanding a string and expecting an integer value. On some systems this
5966 resulted in spurious "integer overflow" errors. Also, it was casting the
5967 result into an int without checking.
5969 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5970 did not work. The code looks as if it has *never* worked, though it appears
5971 to have been documented since at least release 1.62. I have made it work.
5973 35. The "timeout_DNS" error in retry rules, also documented since at least
5974 1.62, also never worked. As it isn't clear exactly what this means, and
5975 clearly it isn't a major issue, I have abolished the feature by treating it
5976 as "timeout", and writing a warning to the main and panic logs.
5978 36. The display of retry rules for -brt wasn't always showing the error code
5981 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5982 timeout_connect_A, timeout_connect_MX.
5984 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5985 to the empty sender.
5987 39. The daemon was not analysing the content of -oX till after it had closed
5988 stderr and disconnected from the controlling terminal. This meant that any
5989 syntax errors were only noted on the panic log, and the return code from
5990 the command was 0. By re-arranging the code a little, I've made the
5991 decoding happen first, so such errors now appear on stderr, and the return
5992 code is 1. However, the actual setting up of the sockets still happens in
5993 the disconnected process, so errors there are still only recorded on the
5996 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5997 connections (as happens on some IP stacks) was logged at start up time as
5998 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5999 it from "IPv6 and IPv4", which means that two separate sockets are being
6002 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6003 says whether AF_INET or AF_INET6 was passed as an argument.
6005 42. Exiwhat output was messed up when time zones were included in log
6008 43. Exiwhat now gives more information about the daemon's listening ports,
6009 and whether -tls-on-connect was used.
6011 44. The "port" option of the smtp transport is now expanded.
6013 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6014 silently ignored. Now an error message is written to the main and panic
6017 46. There's a new ACL modifier called "logwrite" which writes to a log file
6018 as soon as it is encountered.
6020 47. Added $local_user_uid and $local_user_gid at routing time.
6022 48. Exim crashed when trying to verify a sender address that was being
6025 49. Exim was recognizing only a space character after ".include". It now also
6026 recognizes a tab character.
6028 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6029 extracting the relevant information from the specification. The man page no
6030 longer contains scrambled data for the -d option, and I've added a section
6031 at the front about calling Exim under different names.
6033 51. Added "extra_headers" argument to the "mail" command in filter files.
6035 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6038 53. Installed eximstats 1.29.
6040 54. Added transport_filter_timeout as a generic transport option.
6042 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6043 Cc: header lines. This was required by RFC 822, but it not required by RFC
6046 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6047 incoming messages that did not have them. Now it does so only if the
6048 message originates locally, that is, if there is no associated remote host
6049 address. When Resent- header lines are present, this applies to the Resent-
6050 lines rather than the non-Resent- lines.
6052 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6053 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6055 58. Messages for configuration errors now include the name of the main
6056 configuration file - useful now that there may be more than one file in a
6057 list (.included file names were always shown).
6059 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6060 for those rare installations that do not start the daemon as root or run it
6061 setuid root. I've cut out the call to initgroups() if the daemon is not
6064 60. The Exim user and group can now be bound into the binary as text strings
6065 that are looked up at the start of Exim's processing.
6067 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6069 62. Added $mailstore_basename variable.
6071 63. Installed patch to sieve.c from Michael Haardt.
6073 64. When Exim failed to open the panic log after failing to open the main log,
6074 the original message it was trying to log was written to stderr and debug
6075 output, but if they were not available (the usual case in production), it
6076 was lost. Now it is written to syslog before the two lines that record the
6077 failures to open the logs.
6079 65. Users' Exim filters run in subprocesses under the user's uid. It is
6080 possible for a "deliver" command or an alias in a "personal" command to
6081 provoke an address rewrite. If logging of address rewriting is configured,
6082 this fails because the process is not running as root or exim. There may be
6083 a better way of dealing with this, but for the moment (because 4.30 needs
6084 to be released), I have disabled address rewrite logging when running a
6085 filter in a non-root, non-exim process.
6091 1. The buildconfig auxiliary program wasn't quoting the value set for
6092 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6093 not defined. This bug was masked in 4.22 by the effect that was fixed in
6096 2. Some messages that were rejected after a message id was allocated were
6097 shown as "incomplete" by exigrep. It no longer does this for messages that
6098 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6100 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6101 have allow_domain_literals set, the ID did not get logged in the <= line.
6102 Domain literals are now always recognized in Message-ID: header lines.
6104 4. The first argument for a ${extract expansion item is the key name or field
6105 number. Leading and trailing spaces in this item were not being ignored,
6106 causing some misleading effects.
6108 5. When deliver_drop_privilege was set, single queue runner processes started
6109 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6110 same command in the process it spins off) were not dropping privilege.
6112 6. When the daemon running as "exim" started a queue runner, it always
6113 re-executed Exim in the spun-off process. This is a waste of effort when
6114 deliver_drop_privilege is set. The new process now just calls the
6115 queue-runner function directly.
6121 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6124 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6125 ignored. Though the use of -oP was forcing the writing of a pid file, it
6126 was always written to the default place.
6128 3. If the message "no IP address found for host xxxx" is generated during
6129 incoming verification, it is now followed by identification of the incoming
6130 connection (so you can more easily find what provoked it).
6132 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6134 5. Added some features to "harden" Exim a bit more against certain attacks:
6136 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6137 be put in Local/Makefile. This is like the never_users runtime option,
6138 but it cannot be overridden. The default setting is "root".
6140 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6141 prefix string with which any file named in a -C command line option
6144 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6145 is retained for -C and -D only if the caller of Exim is root. Without
6146 it, the exim user may also use -C and -D and retain privilege.
6148 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6149 command line option is disabled.
6151 6. Macro names set by the -D option must start with an upper case letter, just
6152 like macro names defined in the configuration file.
6154 7. Added "dereference=" facility to LDAP.
6156 8. Two instances of the typo "uknown" in the source files are fixed.
6158 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6159 the Configure-Makefile script screwed up while processing it.
6161 10. Incorporated PCRE 4.4.
6163 11. The SMTP synchronization check was not operating right at the start of an
6164 SMTP session. For example, it could not catch a HELO sent before the client
6165 waited for the greeting. There is now a check for outstanding input at the
6166 point when the greeting is written. Because of the duplex, asynchronous
6167 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6168 way, but not yet received, when the check is performed.
6170 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6171 on TCP/IP sockets, because this apparently causes some broken clients to
6174 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6175 unchanged) from the Cygwin maintainer.
6177 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6178 was supported instead of testing for mbx. Effectively a typo.
6180 15. The spa authenticator server code was not checking that the input it
6181 received was valid base64.
6183 16. The debug output line for the "set" modifier in ACLs was not showing the
6184 name of the variable that was being set.
6186 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6188 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6189 lookup on its own. Something else was needed to trigger the lookup. For
6190 example, a match in host_lookup or the need for a host name in a host list.
6191 Now, if $sender_host_name is referenced and the host name has not yet been
6192 looked up, a lookup is performed. If the lookup fails, the variable remains
6193 empty, and $host_lookup_failed is set to "1".
6195 19. Added "eqi" as a case-independent comparison operator.
6197 20. The saslauthd authentication condition could segfault if neither service
6198 nor realm was specified.
6200 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6201 error message that was logged was misleading, and incoming SMTP
6202 connections were dropped. The message is now more accurate, and temporary
6203 errors are given to SMTP connections.
6205 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6206 (except RSET) with a 421 error, until QUIT is received. However, it was
6207 failing to send a response to QUIT.
6209 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6210 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6211 ACL is now run after the helo_try_verify_hosts code.
6213 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6214 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6215 case-independent, so other case variants are also recognized). Apparently
6216 some systems use these upper case variants.
6218 25. If more than two messages were waiting for the same host, and a transport
6219 filter was specified for the transport, Exim sent two messages over the
6220 same TCP/IP connection, and then failed with "socket operation on non-
6221 socket" when it tried to send the third.
6223 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6225 27. The extern definition of crypt16() in expand.c was not being excluded when
6226 the OS had its own crypt16() function.
6228 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6229 as a preferred synonym for return_size_limit, both as an option and as an
6232 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6234 30. Changed the default configuration ACL to relax the local part checking rule
6235 for addresses that are not in any local domains. For these addresses,
6236 slashes and pipe symbols are allowed within local parts, but the sequence
6237 /../ is explicitly forbidden.
6239 31. SPA server authentication was not clearing the challenge buffer before
6242 32. log_message in a "warn" ACL statement was writing to the reject log as
6243 well as to the main log, which contradicts the documentation and doesn't
6244 seem right (because no rejection is happening). So I have stopped it.
6246 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6247 However, I am unable to do any testing of this.
6249 34. Fixed an infelicity in the appendfile transport. When checking directories
6250 for a mailbox, to see if any needed to be created, it was accidentally
6251 using path names with one or more superfluous leading slashes; tracing
6252 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6254 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6255 contents are added to the log line that is written for every discarded
6256 recipient. (Previously a log_message setting was ignored.)
6258 36. The ${quote: operator now quotes the string if it is empty.
6260 37. The install script runs exim in order to find its version number. If for
6261 some reason other than non-existence or emptiness, which it checks, it
6262 could not run './exim', it was installing it with an empty version number,
6263 i.e. as "exim-". This error state is now caught, and the installation is
6266 38. An argument was missing from the function that creates an error message
6267 when Exim fails to connect to the socket for saslauthd authentication.
6268 This could cause Exim to crash, or give a corrupted message.
6270 39. Added isip, isip4, and isip6 to ${if conditions.
6272 40. The ACL variables $acl_xx are now saved with the message, and can be
6273 accessed later in routers, transports, and filters.
6275 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6276 strings in the file are not string-expanded.
6278 42. If a MAIL command specified a SIZE value that was too large to fit into an
6279 int variable, the check against message_size_limit failed. Such values are
6280 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6281 day this will have to be increased, but I don't think I want to be around
6282 when emails are that large.
6289 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6290 iconv() is not standard in FreeBSD.
6292 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6293 IPv6 enabled. The observed symptom was a segmentation fault on return from
6294 the function os_common_find_running_interfaces() in src/os.c.
6296 3. In the check_special_case() function in daemon.c I had used "errno" as an
6297 argument name, which causes warnings on some systems. This was basically a
6298 typo, since it was named "eno" in the comments!
6300 4. The code that waits for the clock to tick (at a resolution of some fraction
6301 of a second) so as to ensure message-id uniqueness was always waiting for
6302 at least one whole tick, when it could have waited for less. [This is
6303 almost certainly not relevant at current processor speeds, where it is
6304 unlikely to ever wait at all. But we try to future-proof.]
6306 5. The function that sleeps for a time interval that includes fractions of a
6307 second contained a race. It did not block SIGALRM between setting the
6308 timer, and suspending (a couple of lines later). If the interval was short
6309 and the sigsuspend() was delayed until after it had expired, the suspension
6310 never ended. On busy systems this could lead to processes getting stuck for
6313 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6314 process, before it forks any delivery processes. The open lookup caching
6315 mechanism meant that the open file or database connection was passed into
6316 the delivery process. The problem was that delivery processes always tidy
6317 up cached lookup data. This could cause a problem for the next delivery
6318 process started by the queue runner, because the external queue runner
6319 process does not know about the closure. So the next delivery process
6320 still has data in the lookup cache. In the case of a file lookup, there was
6321 no problem because closing a file descriptor in a subprocess doesn't affect
6322 the parent. However, if the lookup was caching a connection to a database,
6323 the connection was closed, and the second delivery process was likely to
6324 see errors such as "PGSQL: query failed: server closed the connection
6325 unexpectedly". The problem has been fixed by closing all cached lookups
6326 in a queue runner before running a delivery process.
6328 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6329 seem to have the "const" qualifier which it has on other OS. I've
6332 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6333 *of the same type* (client or server) with the same public name that an
6334 error should be diagnosed.
6336 9. When Exim looked up a host name for an IP address, but failed to find the
6337 original IP address when looking up the host name (a safety check), it
6338 output the message "<ip address> does not match any IP for NULL", which was
6339 confusing, to say the least. The bug was that the host name should have
6340 appeared instead of "NULL".
6342 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6343 user that is built into the binary, and the -C or -D options is used, root
6344 privilege is dropped before the configuration file is read. In addition,
6345 logging is switched to stderr instead of the normal log files. If the
6346 configuration then re-defines the Exim user, the unprivileged environment
6347 is probably not what is expected, so Exim logs a panic warning message (but
6350 However, if deliver_drop_privilege is set, the unprivileged state may well
6351 be exactly what is intended, so the warning has been cut out in that case,
6352 and Exim is allowed to try to write to its normal log files.
6358 1. smtp_return_error_details was not giving details for temporary sender
6359 or receiver verification errors.
6361 2. Diagnose a configuration error if two authenticators have the same public
6364 3. Exim used not to create the message log file for a message until the first
6365 delivery attempt. This could be confusing when incoming messages were held
6366 for policy or load reasons. The message log file is now created at the time
6367 the message is received, and an initial "Received" line is written to it.
6369 4. The automatically generated man page for command line options had a minor
6370 bug that caused no ill effects; however, a more serious problem was that
6371 the procedure for building the man page automatically didn't always
6372 operate. Consequently, release 4.20 contains an out-of-date version. This
6373 shouldn't happen again.
6375 5. When building Exim with embedded Perl support, the script that builds the
6376 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6377 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6379 6. The freeze_tell option was not being used for messages that were frozen on
6380 arrival, either by an ACL or by local_scan().
6382 7. Added the smtp_incomplete_transaction log selector.
6384 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6385 was accepting AUTH without a new EHLO.
6387 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6388 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6391 10. Logging of TCP/IP connections (when configured) now happens in the main
6392 daemon process instead of the child process, so that the TCP/IP connection
6393 count is more accurate (but it can never be perfect).
6395 11. The use of "drop" in a nested ACL was not being handled correctly in the
6396 outer ACL. Now, if condition failure induced by the nested "drop" causes
6397 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6398 or "require"), the connection is dropped.
6400 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6401 that yield "discard" can now be used with an "accept" or a "discard" verb,
6402 but an error is generated for any others (because I can't see a useful way
6403 to define what should happen).
6405 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6406 are now processed in the same way as lines in the Exim configuration file.
6407 In particular, continuation lines are supported.
6409 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6411 15. Added -ti meaning -t -i.
6413 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6414 domains, and warn by logging if others are found.
6416 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6417 returned from GIFCONF when Exim is trying to find the list of interfaces on
6418 a host. The code in os.c has been modified to copy each ifreq to an aligned
6419 structure in all cases.
6421 Also, in some cases, the returned ifreq's were being copied to a 'struct
6422 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6423 means the last couple of bytes of an IPv6 address could be chopped if the
6424 ifreq contained only a normal sockaddr (14 bytes storage).
6426 18. Named domain lists were not supported in the hosts_treat_as_local option.
6427 An entry such as +xxxx was not recognized, and was treated as a literal
6430 19. Ensure that header lines added by a DATA ACL are included in the reject log
6431 if the ACL subsequently rejects the message.
6433 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6434 MD5 (which is deprecated).
6436 21. When testing a filter file using -bf, Exim was writing a message when it
6437 took the sender from a "From " line in the message, but it was not doing so
6438 when it took $return_path from a Return-Path: header line. It now does.
6440 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6441 with a valid header line field name (a series of printing characters
6442 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6444 23. Changed "disc" in the source to "disk" to conform to the documentation and
6445 the book and for uniformity.
6447 24. Ignore Sendmail's -Ooption=value command line item.
6449 25. When execve() failed while trying to run a command in a pipe transport,
6450 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6451 could be confused with a return value of 69 from the command itself. This
6452 has been changed to 127, the value the shell returns if it is asked to run
6453 a non-existent command. The wording for the related log line suggests a
6454 non-existent command as the problem.
6456 26. If received_header_text expands to an empty string, do not add a Received:
6457 header line to the message. (Well, it adds a token one on the spool, but
6458 marks it "old" so that it doesn't get used or transmitted.)
6460 27. Installed eximstats 1.28 (addition of -nt option).
6462 28. There was no check for failure on the call to getsockname() in the daemon
6463 code. This can fail if there is a shortage of resources on the system, with
6464 ENOMEM, for example. A temporary error is now given on failure.
6466 29. Contrary to the C standard, it seems that in some environments, the
6467 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6468 program. Exim now does this explicitly; it affects the formatting of
6469 timestamps using strftime().
6471 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6472 complaints. I've now initialized all the variables, to avoid this.
6474 32. Header lines added by a system filter were not being "seen" during
6475 transport-time rewrites.
6477 33. The info_callback() function passed to OpenSSL is set up with type void
6478 (*)(SSL *, int, int), as described somewhere. However, when calling the
6479 function (actually a macro) that sets it up, the type void(*)() is
6480 expected. I've put in a cast to prevent warnings from picky compilers.
6482 34. If a DNS black list lookup found a CNAME record, but there were no A
6483 records associated with the domain it pointed at, Exim crashed.
6485 35. If a DNS black list lookup returned more than one A record, Exim ignored
6486 all but the first. It now scans all returned addresses if a particular IP
6487 value is being sought. In this situation, the contents of the
6488 $dnslist_value variable are a list of all the addresses, separated by a
6491 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6492 to accept a host name and all its aliases if the forward lookup for any of
6493 them yielded the IP address of the incoming connection. Now it accepts only
6494 those names whose forward lookup yields the correct IP address. Any other
6495 names are discarded. This closes a loophole whereby a rogue DNS
6496 administrator could create reverse DNS records to break through a
6497 wildcarded host restriction in an ACL.
6499 37. If a user filter or a system filter that ran in a subprocess used any of
6500 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6501 the wrong values were passed to the pipe command ($thisaddress had the
6502 value of $0, $0 had the value of $1, etc). This bug was introduced by
6503 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6505 38. Improved the line breaking for long SMTP error messages from ACLs.
6506 Previously, if there was no break point between 40 and 75 characters, Exim
6507 left the rest of the message alone. Two changes have been made: (a) I've
6508 reduced the minimum length to 35 characters; (b) if it can't find a break
6509 point between 35 and 75 characters, it looks ahead and uses the first one
6510 that it finds. This may give the occasional overlong line, but at least the
6511 remaining text gets split now.
6513 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6514 file descriptors might be low, and that setting 1000 would always raise it.
6515 It turns out that in some environments, the limit is already over 1000 and
6516 that lowering it causes trouble. So now Exim takes care not to decrease it.
6518 40. When delivering a message, the value of $return_path is set to $sender_
6519 address at the start of routing (routers may change the value). By an
6520 oversight, this default was not being set up when an address was tested by
6521 -bt or -bv, which affected the outcome if any router or filter referred to
6524 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6525 log only when "message" or "log_message" are set. However, if one of the
6526 conditions was an address verification, or a call to a nested ACL, the
6527 messages generated by the underlying test were being passed through. This
6528 no longer happens. The underlying message is available in $acl_verify_
6529 message for both "message" and "log_message" expansions, so it can be
6530 passed through if needed.
6532 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6533 new expansion $bh_ to give the encoded byte string without charset
6534 translation. Translation happens only if iconv() is available; HAVE_ICONV
6535 indicates this at build time. HEADERS_CHARSET gives the charset to
6536 translate to; headers_charset can change it in the configuration, and
6537 "headers charset" can change it in an individual filter file.
6539 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6540 that creates RFC 2047 encoded "words" labels them as that charset instead
6541 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6542 expansion operator; (ii) when Exim creates a From: line for a local
6543 message; (iii) when a header line is rewritten to include a "phrase" part.
6545 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6546 buggy, causing it to skip the first lines of messages whose message ID
6547 ended in 'D'. This would not have bitten before Exim release 4.14, because
6548 message IDs were unlikely to end in 'D' before then. The effect was to have
6549 incorrect size information for certain domains.
6551 45. #include "config.h" was missing at the start of the crypt16.c module. This
6552 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6555 46. If there was a timeout during a "random" callout check, Exim treated it as
6556 a failure of the random address, and carried on sending RSET and the real
6557 address. If the delay was just some slowness somewhere, the response to the
6558 original RCPT would be taken as a response to RSET and so on, causing
6559 mayhem of various kinds.
6561 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6562 when I implemented it. It didn't allow for the fact that some option values
6563 may legitimately be negative (e.g. size_addition), and it didn't even do
6564 the right test for positive values.
6566 48. Domain names in DNS records are case-independent. Exim always looks them up
6567 in lower case. Some resolvers return domain names in exactly the case they
6568 appear in the zone file, that is, they may contain uppercase letters. Not
6569 all resolvers do this - some return always lower case. Exim was treating a
6570 change of case by a resolver as a change of domain, similar to a widening
6571 of a domain abbreviation. This triggered its re-routing code and so it was
6572 trying to route what was effectively the same domain again. This normally
6573 caused routing to fail (because the router wouldn't handle the domain
6574 twice). Now Exim checks for this case specially, and just changes the
6575 casing of the domain that it ultimately uses when it transmits the message
6578 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6581 50. If a filter generated a file delivery with a non-absolute name (possible if
6582 no home directory exists for the router), the forbid_file option was not
6585 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6586 the existing equality matching.
6588 52. Exim was using ints instead of ino_t variables in some places where it was
6589 dealing with inode numbers.
6591 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6592 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6593 called TMPDIR, and if it finds it is different, it changes its value.
6595 54. The smtp_printf() function is now made available to local_scan() so
6596 additional output lines can be written before returning. There is also an
6597 smtp_fflush() function to enable the detection of a dropped connection.
6598 The variables smtp_input and smtp_batched_input are exported to
6601 55. Changed the default runtime configuration: the message "Unknown user"
6602 has been removed from the ACL, and instead placed on the localuser router,
6603 using the cannot_route_message feature. This means that any verification
6604 failures that generate their own messages won't get overridden. Similarly,
6605 the "Unrouteable address" message that was in the ACL for unverifiable
6606 relay addresses has also been removed.
6608 56. Added hosts_avoid_esmtp to the smtp transport.
6610 57. The exicyclog script was not checking for the esoteric option
6611 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6612 will work only if exicyclog is run under the appropriate euid.
6614 58. Following a discussion on the list, the rules by which Exim recognises line
6615 endings on incoming messages have been changed. The -dropcr and drop_cr
6616 options are now no-ops, retained only for backwards compatibility. The
6617 following line terminators are recognized: LF CRLF CR. However, special
6618 processing applies to CR:
6620 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6621 nor a local message in the state where . is a terminator.
6623 (ii) If a bare CR is encountered in a header line, an extra space is added
6624 after the line terminator so as not to end the header. The reasoning
6625 behind this is that bare CRs in header lines are most likely either
6626 to be mistakes, or people trying to play silly games.
6628 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6629 was being incorrectly given as 18 bytes larger than it should have been.
6630 This is a VOB (very old bug).
6632 60. This may never have affected anything current, but just in case it has:
6633 When the local host is found other than at the start of a list of hosts,
6634 the local host, those with the same MX, and any that follow, are discarded.
6635 When the list in question was part of a longer list of hosts, the following
6636 hosts (not currently being processed) were also being discarded. This no
6637 longer happens. I'm not sure if this situation could ever has previously
6640 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6643 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6644 header. This is something that is recommended in a new Internet Draft, and
6645 is something that is documented as being done by Sendmail. There are two
6646 possible values. For messages generated by the autoreply transport, Exim
6649 Auto-Submitted: auto-replied
6651 whereas for all other generated messages (e.g. bounces) it adds
6653 Auto-Submitted: auto-generated
6655 63. The "personal" condition in filters now includes a test for the
6656 Auto-Submitted: header. If it contains the string "auto-" the message it
6657 not considered personal.
6659 64. Added rcpt_include_affixes as a generic transport option.
6661 65. Added queue_only_override (default true).
6663 66. Added the syslog_duplication option.
6665 67. If what should have been the first header line of a message consisted of
6666 a space followed by a colon, Exim was mis-interpreting it as a header line.
6667 It isn't of course - it is syntactically invalid and should therefore be
6668 treated as the start of the message body. The misbehaviour could have
6669 caused a number of strange effects, including loss of data in subsequent
6670 header lines, and spool format errors.
6672 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6673 client host had authenticated. This control can now be exercised by an ACL
6674 for more flexibility.
6676 69. By default, callouts do not happen when testing with -bh. There is now a
6677 variant, -bhc, which does actually run the callout code, including
6678 consulting and updating the callout cache.
6680 70. Added support for saslauthd authentication, courtesy of Alexander
6683 71. If statvfs() failed on the spool or log directories while checking their
6684 size for availability, Exim confusingly gave the error "space shortage".
6685 Furthermore, in debugging mode it crashed with a floating point exception.
6686 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6687 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6688 serious problem, Exim now writes to the main and panic logs when this
6689 happens, with details of the failure. It then refuses to accept the
6690 incoming message, giving the message "spool directory problem" or "log
6691 directory problem" with a 421 code for SMTP messages.
6693 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6694 0, 1, and 2 exist, because some OS complain for execs without them (see
6695 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6696 descriptors. However, the code omitted to check that the open succeeded,
6697 causing mysterious errors if for some reason the permissions on /dev/null
6698 got screwed. Now Exim writes a message to the main and panic logs, and
6699 bombs out if it can't open /dev/null.
6701 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6702 interact so that it is all more flexible. It is supposed to remain
6703 backwards compatible. Also added extra_local_interfaces.
6705 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6706 to bomb out with an assertion failure - to the client this appears as a
6707 connection drop. This problem occurs in the part of the code that was taken
6708 from the Samba project. Fortunately, the assertion is in a very simple
6709 function, so I have fixed this by reproducing the function inline in the
6710 one place where it is called, and arranging for authentication to fail
6711 instead of killing the process with assert().
6713 75. The SPA client code was not working when the server requested OEM rather
6714 than Unicode encoding.
6716 76. Added code to make require_files with a specific uid setting more usable in
6717 the case where statting the file as root fails - usually a non-root-mounted
6718 NFS file system. When this happens and the failure is EACCES, Exim now
6719 forks a subprocess and does the per-uid checking as the relevant uid.
6721 77. Added process_log_path.
6723 78. If log_file_path was not explicitly set, a setting of check_log_space or
6724 check_log_inodes was ignored.
6726 79. If a space check for the spool or log partitions fails, the incident is now
6727 logged. Of course, in the latter case the data may get lost...
6729 80. Added the %p formatting code to string_format() so that it can be used to
6730 print addresses in debug_print(). Adjusted all the address printing in the
6731 debugging in store.c to use %p rather than %d.
6733 81. There was a concern that a line of code in smtp_in.c could overflow a
6734 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6735 initially expressed, the concern was not well-founded, because trailing
6736 spaces are removed early. However, if the trailing spaces were followed by
6737 a NULL, they did not get removed, so the overflow was possible. Two fixes
6740 (a) I re-wrote the offending code in a cleaner fashion.
6741 (b) If an incoming SMTP command contains a NULL character, it is rejected
6744 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6745 runs initgroups(), so that if the Exim user is in any additional groups,
6746 they will be used during message reception.
6752 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob