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 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
40 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
41 Disable the session-cache too, which might reduce our load. Since we
42 currrectly use a new context for every connection, both as server and
43 client, there is no benefit for these.
44 GnuTLS appears to not support tickets server-side by default (we don't
45 call gnutls_session_ticket_enable_server()) but client side is enabled
46 by default on recent versions (3.1.3 +) unless the PFS priority string
49 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
50 <https://reproducible-builds.org/specs/source-date-epoch/>.
56 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
57 than -2003 did; needs libidn2 in addition to libidn.
59 JH/02 The path option on a pipe transport is now expanded before use.
61 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
62 Patch provided by "Björn", documentation fix added too.
64 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
65 missing a wire-to-host endian conversion.
67 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
68 close after a BDAT command line could be taken as a following command,
69 giving a synch failure. Fix by only checking for synch immediately
70 before acknowledging the chunk.
72 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
73 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
74 macro. Patches provided by Josh Soref.
76 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
77 Previously we did not; the RFC seems ambiguous and VRFY is not listed
78 by IANA as a service extension. However, John Klensin suggests that we
81 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
82 the dkim code may be unix-mode line endings rather than smtp wire-format
83 CRLF, so prepend a CR to any bare LF.
85 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
86 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
88 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
89 the first were themselves being wrongly included in the feed into dkim
90 processing; with most chunk sizes in use this resulted in an incorrect
91 body hash calculated value.
93 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
94 DKIM signature block, for verification. Although advised against by
95 standards it is specifically not ruled illegal.
97 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
99 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
100 missing a body hash (the bh= tag).
102 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
103 It seems that HAProxy sends the Proxy Protocol information in clear and
104 only then does a TLS startup, so do the same.
106 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
107 TCP connections (such as for Spamd) unless the daemon successfully set
108 Fast Open mode on its listening sockets. This fixes breakage seen on
109 too-old kernels or those not configured for Fast Open, at the cost of
110 requiring both directions being enabled for TFO, and TFO never being used
111 by non-daemon-related Exim processes.
113 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
114 endings, at least on the first header line. Try to canonify any that get
115 past that check, despite the cost.
117 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
118 now limited to an arbitrary five deep, while parsing addresses with the
119 strip_excess_angle_brackets option enabled.
121 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
122 instead leave the unprompted TLS handshake in socket buffer for the
123 TLS library to consume.
125 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
127 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
129 JH/16 Drop variables when they go out of scope. Memory management drops a whole
130 region in one operation, for speed, and this leaves assigned pointers
131 dangling. Add checks run only under the testsuite which checks all
132 variables at a store-reset and panics on a dangling pointer; add code
133 explicitly nulling out all the variables discovered. Fixes one known
134 bug: a transport crash, where a dangling pointer for $sending_ip_address
135 originally assigned in a verify callout, is re-used.
137 PP/06 Drop '.' from @INC in various Perl scripts.
139 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
141 PP/08 Reduce a number of compilation warnings under clang; building with
142 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
143 should be warning-free.
145 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
147 HS/01 Fix portability problems introduced by PP/08 for platforms where
148 realloc(NULL) is not equivalent to malloc() [SunOS et al].
150 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
151 chunk. This allows us to accept broken chunked messages. We need a more
152 general solution here.
154 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
155 already-broken messages in the queue.
157 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
159 JH/19 Fix reference counting bug in routing-generated-address tracking.
165 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
166 supports it and a size is available (ie. the sending peer gave us one).
168 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
169 years of being deprecated). The replacements are the ACLs
170 acl_smtp_mime and acl_not_smtp_mime.
172 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
173 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
174 or even an in-clear connection were permitted. Now, if the host lookup
175 was dnssec and dane was requested then the host is only used if the
176 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
177 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
178 if one fails this test.
179 This means that a poorly-configured remote DNS will make it incommunicado;
180 but it protects against a DNS-interception attack on it.
182 JH/04 Bug 1810: make continued-use of an open smtp transport connection
183 non-noisy when a race steals the message being considered.
185 JH/05 If main configuration option tls_certificate is unset, generate a
186 self-signed certificate for inbound TLS connections.
188 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
189 in rewrites and routers.
191 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
192 and logged a warning sing 4.83; now they are a configuration file error.
194 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
195 (lacking @domain). Apply the same qualification processing as RCPT.
197 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
199 JH/10 Support ${sha256:} applied to a string (as well as the previous
202 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
203 a cutthrough deliver is pending, as we always want to make a connection.
204 This also avoids re-routing the message when later placing the cutthrough
205 connection after a verify cache hit.
206 Do not update it with the verify result either.
208 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
209 when routing results in more than one destination address.
211 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
212 signing (which inhibits the cutthrough capability). Previously only
213 the presence of an option was tested; now an expansion evaluating as
214 empty is permissible (obviously it should depend only on data available
215 when the cutthrough connection is made).
217 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
218 the relevant preceding SMTP command did not note the pipelining mode.
220 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
221 Previously they were not counted.
223 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
224 as one having no matching records. Previously we deferred the message
225 that needed the lookup.
227 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
228 distinguished as "(=".
230 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
231 for missing MX records. Previously it only worked for missing A records.
233 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
235 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
236 after the data-go-ahead and data-ack. Patch from Jason Betts.
238 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
239 even for a "none" policy. Patch from Tony Meyer.
241 JH/22 Fix continued use of a connection for further deliveries. If a port was
242 specified by a router, it must also match for the delivery to be
245 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
246 When one of the recipients of a message was unsuitable for the connection
247 (has no matching addresses), we lost track of needing to mark it
248 deferred. As a result mail would be lost.
250 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
252 JH/25 Decoding ACL controls is now done using a binary search; the source code
253 takes up less space and should be simpler to maintain. Merge the ACL
254 condition decode tables also, with similar effect.
256 JH/26 Fix problem with one_time used on a redirect router which returned the
257 parent address unchanged. A retry would see the parent address marked as
258 delivered, so not attempt the (identical) child. As a result mail would
261 JH/27 Fix a possible security hole, wherein a process operating with the Exim
262 UID can gain a root shell. Credit to http://www.halfdog.net/ for
263 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
266 JH/28 Enable {spool,log} filesystem space and inode checks as default.
267 Main config options check_{log,spool}_{inodes,space} are now
268 100 inodes, 10MB unless set otherwise in the configuration.
270 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
271 Previously it only applied to the main-section connection policy
274 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
276 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
277 by me. Added RFC7919 DH primes as an alternative.
279 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
280 are not in the system include path.
282 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
283 GnuTLS, when a session startup failed (eg because the client disconnected)
284 Exim did stdio operations after fclose. This was exposed by a recent
285 change which nulled out the file handle after the fclose.
287 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
288 signed directly by the cert-signing cert, rather than an intermediate
289 OCSP-signing cert. This is the model used by LetsEncrypt.
291 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
293 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
294 an incoming connection.
296 HS/02 Bug 1802: Do not half-close the connection after sending a request
299 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
300 fallback to "prime256v1".
302 JH/34 SECURITY: Use proper copy of DATA command in error message.
303 Could leak key material. Remotely exploitable. CVE-2016-9963.
309 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
310 and 3.4.4 - once the server is enabled to respond to an OCSP request
311 it does even when not requested, resulting in a stapling non-aware
312 client dropping the TLS connection.
314 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
315 support variable-length bit vectors. No functional change.
317 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
318 The I= interface field on outgoing lines is now after the H= remote
319 host field, same as incoming lines. There is a separate
320 outgoing_interface log selector which allows you to disable the
323 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
324 If not running log_selector +smtp_connection the mainlog would be held
325 open indefinitely after a "too many connections" event, including to a
326 deleted file after a log rotate. Leave the per net connection logging
327 leaving it open for efficiency as that will be quickly detected by the
328 check on the next write.
330 HS/01 Bug 1671: Fix post transport crash.
331 Processing the wait-<transport> messages could crash the delivery
332 process if the message IDs didn't exist for some reason. When
333 using 'split_spool_directory=yes' the construction of the spool
334 file name failed already, exposing the same netto behaviour.
336 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
337 mime_regex ACL conditions.
339 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
340 to DSN fail messages (bounces): remote IP, remote greeting, remote response
341 to HELO, local diagnostic string.
343 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
344 log line to debug. Even when configured with a tls authenticator many
345 client connections are expected to not authenticate in this way, so
346 an authenticate fail is not an error.
348 HS/02 Add the Exim version string to the process info. This way exiwhat
349 gives some more detail about the running daemon.
351 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
352 matter for fast-change records such as DNSBLs.
354 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
355 retry record, even if constant. There may be multiple transports with
356 different interface settings and the retry behaviour needs to be kept
359 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
361 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
363 JH/10 Bug 840: fix log_defer_output option of pipe transport
365 JH/11 Bug 830: use same host for all RCPTS of a message, even under
366 hosts_randomize. This matters a lot when combined with mua_wrapper.
368 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
369 ${quote_pgsql:<string>} operator.
371 JH/13 Bug 1708: avoid misaligned access in cached lookup.
373 JH/14 Change header file name for freeradius-client. Relevant if compiling
374 with Radius support; from the Gentoo tree and checked under Fedora.
376 JH/15 Bug 1712: Introduce $prdr_requested flag variable
378 JH/16 Bug 1714: Permit an empty string as expansion result for transport
379 option transport_filter, meaning no filtering.
381 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
383 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
384 defaults to "*" (all hosts). The variable is now available when not built
385 with TLS, default unset, mainly to enable keeping the testsuite sane.
386 If a server certificate is not supplied (via tls_certificate) an error is
387 logged, and clients will find TLS connections fail on startup. Presumably
388 they will retry in-clear.
389 Packagers of Exim are strongly encouraged to create a server certificate
390 at installation time.
392 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
393 with the $config_file variable.
395 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
396 in transport context, after the attempt, and per-recipient. The latter type
397 is per host attempted. The event data is the error message, and the errno
398 information encodes the lookup type (A vs. MX) used for the (first) host,
399 and the trailing two digits of the smtp 4xx response.
401 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
402 to write to mainlog (or rejectlog, paniclog) in the window between file
403 creation and permissions/ownership being changed. Particularly affects
404 installations where exicyclog is run as root, rather than exim user;
405 result is that the running daemon panics and dies.
407 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
409 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
410 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
411 "pri" and "weight". Note that the previous implicit priority given by the
412 list order is no longer honoured.
414 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
417 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
418 by defining SUPPORT_SOCKS.
420 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
421 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
422 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
423 variables are renamed to proxy_{local,external}_{address,port}.
425 JH/27 Move Internationalisation support from Experimental to mainline, enabled
426 for a build by defining SUPPORT_I18N
428 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
429 of the query string, and make ${quote_redis:} do that quoting.
431 JH/29 Move Events support from Experimental to mainline, enabled by default
432 and removable for a build by defining DISABLE_EVENT.
434 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
436 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
437 cached by the daemon.
439 JH/32 Move Redis support from Experimental to mainline, enabled for a build
440 by defining LOOKUP_REDIS. The libhiredis library is required.
442 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
443 keys are given for lookup.
445 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
446 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
447 only supported when built with TLS support. The PolarSSL SHA routines
448 are still used when the TLS library is too old for convenient support.
450 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
451 openssl_options), for security. OpenSSL forces this from version 1.1.0
452 server-side so match that on older versions.
454 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
455 allocation for $value could be released as the expansion processing
456 concluded, but leaving the global pointer active for it.
458 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
459 and to use the domains and local_parts ACL conditions.
461 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
462 incorrectly not doubled on cutthrough transmission, hence seen as a
463 body-termination at the receiving system - resulting in truncated mails.
464 Commonly the sender saw a TCP-level error, and retransmitted the message
465 via the normal store-and-forward channel. This could result in duplicates
466 received - but deduplicating mailstores were liable to retain only the
467 initial truncated version.
469 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
471 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
473 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
474 we're in there, support oversigning also; bug 1309.
476 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
478 HS/04 Add support for keep_environment and add_environment options.
480 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
481 either intentional arithmetic overflow during PRNG, or testing config-
484 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
485 delivery resulted in actual delivery. Cancel cutthrough before DATA
488 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
489 rejects a recipient: pass the reject to the originator.
491 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
492 Many were false-positives and ignorable, but it's worth fixing the
495 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
496 for the new environment-manipulation done at startup. Move the routines
497 from being local to tls.c to being global via the os.c file.
499 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
500 an extract embedded as result-arg for a map, the first arg for extract
501 is unavailable so we cannot tell if this is a numbered or keyed
502 extraction. Accept either.
508 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
511 JH/02 The smtp transport option "multi_domain" is now expanded.
513 JH/03 The smtp transport now requests PRDR by default, if the server offers
516 JH/04 Certificate name checking on server certificates, when exim is a client,
517 is now done by default. The transport option tls_verify_cert_hostnames
518 can be used to disable this per-host. The build option
519 EXPERIMENTAL_CERTNAMES is withdrawn.
521 JH/05 The value of the tls_verify_certificates smtp transport and main options
522 default to the word "system" to access the system default CA bundle.
523 For GnuTLS, only version 3.0.20 or later.
525 JH/06 Verification of the server certificate for a TLS connection is now tried
526 (but not required) by default. The verification status is now logged by
527 default, for both outbound TLS and client-certificate supplying inbound
530 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
533 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
534 Status Notification (bounce) messages are now MIME format per RFC 3464.
535 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
536 under the control of the dsn_advertise_hosts option, and routers may
537 have a dsn_lasthop option.
539 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
540 default, modifiable by a malware= option. The list separator for
541 the options can now be changed in the usual way. Bug 68.
543 JH/10 The smtp_receive_timeout main option is now expanded before use.
545 JH/11 The incoming_interface log option now also enables logging of the
546 local interface on delivery outgoing connections.
548 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
549 if the interface and destination host and port all match.
551 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
554 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
555 Patch from Andrew Lewis.
557 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
558 now supports optional time-restrictions, weighting, and priority
559 modifiers per server. Patch originally by <rommer@active.by>.
561 JH/16 The spamd_address main option now supports a mixed list of local
562 and remote servers. Remote servers can be IPv6 addresses, and
563 specify a port-range.
565 JH/17 Bug 68: The spamd_address main option now supports an optional
566 timeout value per server.
568 JH/18 Bug 1581: Router and transport options headers_add/remove can
569 now have the list separator specified.
571 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
574 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
577 JH/21 Support for the A6 type of dns record is withdrawn.
579 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
580 rather than the verbs used.
582 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
583 from 255 to 1024 chars.
585 JH/24 Verification callouts now attempt to use TLS by default.
587 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
588 are generic router options now. The defaults didn't change.
590 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
591 Original patch from Alexander Shikoff, worked over by JH.
593 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
594 files if ZCAT_COMMAND is not executable.
596 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
598 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
600 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
601 Normally benign, it bites when the pair was led to by a CNAME;
602 modern usage is to not canonicalize the domain to a CNAME target
603 (and we were inconsistent anyway for A-only vs AAAA+A).
605 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
607 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
608 when evaluating $sender_host_dnssec.
610 JH/31 Check the HELO verification lookup for DNSSEC, adding new
611 $sender_helo_dnssec variable.
613 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
615 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
617 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
619 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
620 documented as working, but never had. Support all but $spam_report.
622 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
623 added for tls authenticator.
625 HS/03 Add perl_taintmode main config option
631 TL/01 When running the test suite, the README says that variables such as
632 no_msglog_check are global and can be placed anywhere in a specific
633 test's script, however it was observed that placement needed to be near
634 the beginning for it to behave that way. Changed the runtest perl
635 script to read through the entire script once to detect and set these
636 variables, reset to the beginning of the script, and then run through
637 the script parsing/test process like normal.
639 TL/02 The BSD's have an arc4random API. One of the functions to induce
640 adding randomness was arc4random_stir(), but it has been removed in
641 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
642 function when detected.
644 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
645 cause callback expansion.
647 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
648 syntax errors in an expansion can be treated as a string instead of
649 logging or causing an error, due to the internal use of bool_lax
650 instead of bool when processing it.
652 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
653 server certificates when making smtp deliveries.
655 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
657 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
659 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
661 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
662 Merged patch from Sebastian Wiedenroth.
664 JH/05 Fix results-pipe from transport process. Several recipients, combined
665 with certificate use, exposed issues where response data items split
666 over buffer boundaries were not parsed properly. This eventually
667 resulted in duplicates being sent. This issue only became common enough
668 to notice due to the introduction of connection certificate information,
669 the item size being so much larger. Found and fixed by Wolfgang Breyha.
671 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
672 size buffer was used, resulting in syntax errors when an expansion
675 JH/07 Add support for directories of certificates when compiled with a GnuTLS
676 version 3.3.6 or later.
678 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
679 is EXPERIMENTAL_EVENT, the main-configuration and transport options
680 both become "event_action", the variables become $event_name, $event_data
681 and $event_defer_errno. There is a new variable $verify_mode, usable in
682 routers, transports and related events. The tls:cert event is now also
683 raised for inbound connections, if the main configuration event_action
686 TL/06 In test suite, disable OCSP for old versions of openssl which contained
687 early OCSP support, but no stapling (appears to be less than 1.0.0).
689 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
690 server certificate names available under the smtp transport option
691 "tls_verify_cert_hostname" now do not permit multi-component wildcard
694 JH/10 Time-related extraction expansions from certificates now use the main
695 option "timezone" setting for output formatting, and are consistent
696 between OpenSSL and GnuTLS compilations. Bug 1541.
698 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
699 encoded parameter in the incoming message. Bug 1558.
701 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
702 include certificate info, eximon was claiming there were spoolfile
705 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
707 JH/14 Log delivery-related information more consistently, using the sequence
708 "H=<name> [<ip>]" wherever possible.
710 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
711 are problematic for Debian distribution, omit them from the release
714 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
716 JH/16 Fix string representation of time values on 64bit time_t architectures.
719 JH/17 Fix a null-indirection in certextract expansions when a nondefault
720 output list separator was used.
725 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
726 checkers that were complaining about end of non-void function with no
729 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
730 This was a regression introduced in 4.83 by another bugfix.
732 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
734 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
735 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
741 TF/01 Correctly close the server side of TLS when forking for delivery.
743 When a message was received over SMTP with TLS, Exim failed to clear up
744 the incoming connection properly after forking off the child process to
745 deliver the message. In some situations the subsequent outgoing
746 delivery connection happened to have the same fd number as the incoming
747 connection previously had. Exim would try to use TLS and fail, logging
748 a "Bad file descriptor" error.
750 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
751 utilities have not been installed.
753 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
754 temporary space as the ACL may create new global variables.
756 TL/01 LDAP support uses per connection or global context settings, depending
757 upon the detected version of the libraries at build time.
759 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
760 to extract and use the src ip:port in logging and expansions as if it
761 were a direct connection from the outside internet. PPv2 support was
762 updated based on HAProxy spec change in May 2014.
764 JH/02 Add ${listextract {number}{list}{success}{fail}}.
766 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
767 Properly escape header and check for NULL return.
769 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
772 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
774 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
775 characters in header names, implemented as a verify condition.
776 Contributed by Michael Fischer v. Mollard.
778 TL/05 Rename SPF condition results err_perm and err_temp to standardized
779 results permerror and temperror. Previous values are deprecated but
780 still accepted. In a future release, err_perm and err_temp will be
781 completely removed, which will be a backward incompatibility if the
782 ACL tests for either of these two old results. Patch contributed by
783 user bes-internal on the mailing list.
785 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
787 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
788 selectors, in both main and reject logs.
790 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
793 JH/07 Add malware type "sock" for talking to simple daemon.
795 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
797 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
798 routers/transports under cutthrough routing.
800 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
801 numbers. Touch up "bool" conditional to keep the same definition.
803 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
805 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
807 JH/12 Expand items in router/transport headers_add or headers_remove lists
808 individually rather than the list as a whole. Bug 1452.
810 Required for reasonable handling of multiple headers_ options when
811 they may be empty; requires that headers_remove items with embedded
812 colons must have them doubled (or the list-separator changed).
814 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
815 view the policy declared in the DMARC record. Currently, $dmarc_status
816 is a combined value of both the record presence and the result of the
819 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
821 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
822 dnslookup router and the smtp transport (applying to the forward
825 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
826 of ldap servers used for a specific lookup. Patch provided by Heiko
829 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
830 New variable $lookup_dnssec_authenticated for observability.
832 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
833 Patch submitted by Lars Timman.
835 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
837 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
838 Requires trusted mode and valid format message id, aborts otherwise.
839 Patch contributed by Heiko Schlichting.
841 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
842 certextract with support for various fields. Bug 1358.
844 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
845 is requested by default, modifiable by smtp transport option
848 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
849 operate on certificate variables to give certificate fingerprints
850 Also new ${sha256:cert_variable}.
852 JH/23 The PRDR feature is moved from being Experimental into the mainline.
854 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
855 Christian Aistleitner.
857 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
859 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
860 file. Patch from Wolfgang Breyha.
862 JH/25 Expand the coverage of the delivery $host and $host_address to
863 client authenticators run in verify callout. Bug 1476.
865 JH/26 Port service names are now accepted for tls_on_connect_ports, to
866 align with daemon_smtp_ports. Bug 72.
868 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
869 support and error reporting did not work properly.
871 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
872 and is readable. Patch from Andrew Colin Kissa.
874 TL/14 Enhance documentation of ${run expansion and how it parses the
875 commandline after expansion, particularly in the case when an
876 unquoted variable expansion results in an empty value.
878 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
880 PP/02 Fix internal collision of T_APL on systems which support RFC3123
881 by renaming away from it. Addresses GH issue 15, reported by
884 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
886 TL/15 SECURITY: prevent double expansion in math comparison functions
887 (can expand unsanitized data). Not remotely exploitable.
894 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
896 PP/02 Make -n do something, by making it not do something.
897 When combined with -bP, the name of an option is not output.
899 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
902 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
903 $sender_host_name and config options to manage this, and basic check
906 PP/05 DSCP support for outbound connections and control modifier for inbound.
908 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
909 (Only plugin which currently uses this is kerberos4, which nobody should
910 be using, but we should make it available and other future plugins might
911 conceivably use it, even though it would break NAT; stuff *should* be
912 using channel bindings instead).
914 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
915 name; added for Sendmail compatibility; requires admin caller.
916 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
917 just ignore it); requires trusted caller.
918 Also parse but ignore: -Ac -Am -X<logfile>
921 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
923 TL/02 Add +smtp_confirmation as a default logging option.
925 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
926 Patch by Magnus Holmgren from 2007-02-20.
928 TL/04 Bugzilla 1281 - Spec typo.
929 Bugzilla 1283 - Spec typo.
930 Bugzilla 1290 - Spec grammar fixes.
932 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
934 TL/06 Add Experimental DMARC support using libopendmarc libraries.
936 TL/07 Fix an out of order global option causing a segfault. Reported to dev
937 mailing list by by Dmitry Isaikin.
939 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
941 JH/02 Support "G" suffix to numbers in ${if comparisons.
943 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
945 NM/01 Bugzilla 1197 - Spec typo
946 Bugzilla 1196 - Spec examples corrections
948 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
950 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
951 gnutls_enable_pkcs11, but renamed to more accurately indicate its
954 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
955 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
957 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
958 "acl {{name}{arg}...}", and optional args on acl condition
961 JH/05 Permit multiple router/transport headers_add/remove lines.
963 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
965 JH/07 Avoid using a waiting database for a single-message-only transport.
966 Performance patch from Paul Fisher. Bugzilla 1262.
968 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
971 JH/09 Add $headers_added variable, with content from use of ACL modifier
972 add_header (but not yet added to the message). Bugzilla 199.
974 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
975 Pulled from Bugzilla 817 by Wolfgang Breyha.
977 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
979 (nb: this is the same fix as in Exim 4.80.1)
981 JH/11 Add A= logging on delivery lines, and a client_set_id option on
984 JH/12 Add optional authenticated_sender logging to A= and a log_selector
987 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
989 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
990 advertise SMTP AUTH mechanism to us, instead of a generic
991 protocol violation error. Also, make Exim more robust to bad
992 data from the Dovecot auth socket.
994 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
996 When a queue runner is handling a message, Exim first routes the
997 recipient addresses, during which it prunes them based on the retry
998 hints database. After that it attempts to deliver the message to
999 any remaining recipients. It then updates the hints database using
1002 So if a recipient address works intermittently, it can get repeatedly
1003 deferred at routing time. The retry hints record remains fresh so the
1004 address never reaches the final cutoff time.
1006 This is a fairly common occurrence when a user is bumping up against
1007 their storage quota. Exim had some logic in its local delivery code
1008 to deal with this. However it did not apply to per-recipient defers
1009 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1011 This change adds a proper retry rule check during routing so that the
1012 final cutoff time is checked against the message's age. We only do
1013 this check if there is an address retry record and there is not a
1014 domain retry record; this implies that previous attempts to handle
1015 the address had the retry_use_local_parts option turned on. We use
1016 this as an approximation for the destination being like a local
1017 delivery, as in LMTP.
1019 I suspect this new check makes the old local delivery cutoff check
1020 redundant, but I have not verified this so I left the code in place.
1022 TF/02 Correct gecos expansion when From: is a prefix of the username.
1024 Test 0254 submits a message to Exim with the header
1028 When I ran the test suite under the user fanf2, Exim expanded
1029 the header to contain my full name, whereas it should have added
1030 a Resent-Sender: header. It erroneously treats any prefix of the
1031 username as equal to the username.
1033 This change corrects that bug.
1035 GF/01 DCC debug and logging tidyup
1036 Error conditions log to paniclog rather than rejectlog.
1037 Debug lines prefixed by "DCC: " to remove any ambiguity.
1039 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1041 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1042 Bug spotted by Jeremy Harris; was flawed since initial commit.
1043 Would have resulted in OCSP responses post-SNI triggering an Exim
1044 NULL dereference and crash.
1046 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1048 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1049 Bug detection, analysis and fix by Samuel Thibault.
1050 Bugzilla 1331, Debian bug #698092.
1052 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1054 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1055 Server implementation by Todd Lyons, client by JH.
1056 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1057 config variable "prdr_enable" controls whether the server
1058 advertises the facility. If the client requests PRDR a new
1059 acl_data_smtp_prdr ACL is called once for each recipient, after
1060 the body content is received and before the acl_smtp_data ACL.
1061 The client is controlled by both of: a hosts_try_prdr option
1062 on the smtp transport, and the server advertisement.
1063 Default client logging of deliveries and rejections involving
1064 PRDR are flagged with the string "PRDR".
1066 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1067 fclose(). Diagnosis by Todd Lyons.
1069 PP/17 Update configure.default to handle IPv6 localhost better.
1070 Patch by Alain Williams (plus minor tweaks).
1073 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1074 This is now consistent with GnuTLS, and is now documented: the
1075 previous undocumented portable approach to treating the option as
1076 unset was to force an expansion failure. That still works, and
1077 an empty string is now equivalent.
1079 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1080 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1081 not performing validation itself.
1083 PP/20 Added force_command boolean option to pipe transport.
1084 Patch from Nick Koston, of cPanel Inc.
1086 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1089 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1091 PP/21 Fix eximon continuous updating with timestamped log-files.
1092 Broken in a format-string cleanup in 4.80, missed when I repaired the
1093 other false fix of the same issue.
1094 Report and fix from Heiko Schlichting.
1097 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1098 Report from Prashanth Katuri.
1100 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1101 It's SecureTransport, so affects any MacOS clients which use the
1102 system-integrated TLS libraries, including email clients.
1104 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1105 using a MIME ACL for non-SMTP local injection.
1106 Report and assistance in diagnosis by Warren Baker.
1108 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1110 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1112 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1113 last id that failed so it may be referenced in subsequent ACL's.
1115 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1118 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1119 ldap library initialization, allowing self-signed CA's to be
1120 used. Also properly sets require_cert option later in code by
1121 using NULL (global ldap config) instead of ldap handle (per
1122 session). Bug diagnosis and testing by alxgomz.
1124 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1125 the src/util/ subdirectory.
1127 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1128 renamed to Transport Post Delivery Action by Jeremy Harris, as
1131 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1132 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1133 redis_servers = needs to be configured which will be used by the redis
1134 lookup. Patch from Warren Baker, of The Packet Hub.
1136 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1138 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1139 hostname or reverse DNS when processing a host list. Used suggestions
1140 from multiple comments on this bug.
1142 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1144 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1145 Missed a few lines, added it to make the runtest require no keyboard
1148 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1149 contains upper case chars. Make router use caseful_local_part.
1151 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1152 support when GnuTLS has been built with p11-kit.
1158 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1160 This, or similar/improved, will also be change PP/11 of 4.82.
1166 PP/01 Handle short writes when writing local log-files.
1167 In practice, only affects FreeBSD (8 onwards).
1168 Bugzilla 1053, with thanks to Dmitry Isaikin.
1170 NM/01 Bugzilla 949 - Documentation tweak
1172 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1175 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1177 PP/02 Implemented gsasl authenticator.
1179 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1181 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1182 `pkg-config foo` for cflags/libs.
1184 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1185 with rest of GSASL and with heimdal_gssapi.
1187 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1188 `pkg-config foo` for cflags/libs for the TLS implementation.
1190 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1191 properties get this fed in as external SSF. A number of robustness
1192 and debugging improvements to the cyrus_sasl authenticator.
1194 PP/08 cyrus_sasl server now expands the server_realm option.
1196 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1197 Patch by Jeremy Harris.
1199 PP/10 Added dbmjz lookup type.
1201 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1203 PP/12 MAIL args handles TAB as well as SP, for better interop with
1204 non-compliant senders.
1205 Analysis and variant patch by Todd Lyons.
1207 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1208 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1209 Patch from Dirk Mueller <dmueller@suse.com>
1211 PP/13 tls_peerdn now print-escaped for spool files.
1212 Observed some $tls_peerdn in wild which contained \n, which resulted
1213 in spool file corruption.
1215 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1216 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1217 or write after TLS renegotiation, which otherwise led to messages
1220 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1221 as a tracking header (ie: a signed header comes before the signature).
1222 Patch from Wolfgang Breyha.
1224 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1225 comma-sep list; embedded commas doubled.
1227 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1229 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1231 Report and patch from Dmitry Banschikov.
1233 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1234 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1235 failures. We appear to now support TLS1.1+ with Exim.
1237 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1238 lets Exim select keys and certificates based upon TLS SNI from client.
1239 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1240 before an outbound SMTP session. New log_selector, +tls_sni.
1242 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1243 NULL dereference. Report and patch from Alun Jones.
1245 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1246 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1247 Not seeing resolver debug output on NetBSD, but suspect this is a
1248 resolver implementation change.
1250 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1251 Left warnings. Added "eximon gdb" invocation mode.
1253 PP/21 Defaulting "accept_8bitmime" to true, not false.
1255 PP/22 Added -bw for inetd wait mode support.
1257 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1258 locate the relevant includes and libraries. Made this the default.
1260 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1261 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1263 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1264 This may cause build issues on older platforms.
1266 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1267 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1268 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1269 Added SNI support via GnuTLS too.
1270 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1272 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1274 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1275 Applied second patch from Janne, implementing suggestion to default
1276 multiple-strings-in-record handling to match SPF spec.
1278 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1280 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1281 read-only, out of scope).
1282 Patch from Wolfgang Breyha, report from Stuart Northfield.
1284 PP/29 Fix three issues highlighted by clang analyser static analysis.
1285 Only crash-plausible issue would require the Cambridge-specific
1286 iplookup router and a misconfiguration.
1287 Report from Marcin Mirosław.
1289 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1291 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1292 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1293 As part of this, removing so much warning spew let me fix some minor
1294 real issues in debug logging.
1296 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1297 assignment on my part. Fixed.
1299 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1300 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1301 Janne Snabb (who went above and beyond: thank you).
1303 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1304 string otherwise requires a connection and a bunch more work and it's
1305 relatively easy to get wrong. Should also expose TLS library linkage
1308 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1309 64-bit ${eval} (JH/03).
1311 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1312 GNU libc to support some of the 64-bit stuff, should not lead to
1313 conflicts. Defined before os.h is pulled in, so if a given platform
1314 needs to override this, it can.
1316 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1317 protection layer was required, which is not implemented.
1318 Bugzilla 1254, patch from Wolfgang Breyha.
1320 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1321 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1322 tls_dhparam take prime identifiers. Also unbreak combination of
1323 OpenSSL+DH_params+TLSSNI.
1325 PP/39 Disable SSLv2 by default in OpenSSL support.
1331 PP/01 Solaris build fix for Oracle's LDAP libraries.
1332 Bugzilla 1109, patch from Stephen Usher.
1334 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1336 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1339 TF/02 Fix a couple more cases where we did not log the error message
1340 when unlink() failed. See also change 4.74-TF/03.
1342 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1343 lock up or crash if it happened to be inside a call to libc when it
1344 got a SIGUSR1 from exiwhat.
1346 The SIGUSR1 handler appends the current process status to the process
1347 log which is later printed by exiwhat. It used to use the general
1348 purpose logging code to do this, but several functions it calls are
1349 not safe for signals.
1351 The new output code in the SIGUSR1 handler is specific to the process
1352 log, and simple enough that it's easy to inspect for signal safety.
1353 Removing some special cases also simplifies the general logging code.
1354 Removing the spurious timestamps from the process log simplifies
1357 TF/04 Improved ratelimit ACL condition.
1359 The /noupdate option has been deprecated in favour of /readonly which
1360 has clearer semantics. The /leaky, /strict, and /readonly update modes
1361 are mutually exclusive. The update mode is no longer included in the
1362 database key; it just determines when the database is updated. (This
1363 means that when you upgrade Exim will forget old rate measurements.)
1365 Exim now checks that the per_* options are used with an update mode that
1366 makes sense for the current ACL. For example, when Exim is processing a
1367 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1368 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1369 must specify per_mail/readonly. If you omit the update mode it defaults to
1370 /leaky where that makes sense (as before) or /readonly where required.
1372 The /noupdate option is now undocumented but still supported for
1373 backwards compatibility. It is equivalent to /readonly except that in
1374 ACLs where /readonly is required you may specify /leaky/noupdate or
1375 /strict/noupdate which are treated the same as /readonly.
1377 A useful new feature is the /count= option. This is a generalization
1378 of the per_byte option, so that you can measure the throughput of other
1379 aggregate values. For example, the per_byte option is now equivalent
1380 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1382 The per_rcpt option has been generalized using the /count= mechanism
1383 (though it's more complicated than the per_byte equivalence). When it is
1384 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1385 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1386 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1387 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1388 Note that using per_rcpt with a non-readonly update mode in more than
1389 one ACL will cause the recipients to be double-counted. (The per_mail
1390 and per_byte options don't have this problem.)
1392 The handling of very low rates has changed slightly. If the computed rate
1393 is less than the event's count (usually one) then this event is the first
1394 after a long gap. In this case the rate is set to the same as this event's
1395 count, so that the first message of a spam run is counted properly.
1397 The major new feature is a mechanism for counting the rate of unique
1398 events. The new per_addr option counts the number of different
1399 recipients that someone has sent messages to in the last time period. It
1400 behaves like per_rcpt if all the recipient addresses are different, but
1401 duplicate recipient addresses do not increase the measured rate. Like
1402 the /count= option this is a general mechanism, so the per_addr option
1403 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1404 example, measure the rate that a client uses different sender addresses
1405 with the options per_mail/unique=$sender_address. There are further
1406 details in the main documentation.
1408 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1410 TF/06 Removed a few PCRE remnants.
1412 TF/07 Automatically extract Exim's version number from tags in the git
1413 repository when doing development or release builds.
1415 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1416 Bugzilla 879. Patch from Paul Fisher.
1418 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1419 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1422 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1424 PP/05 Variable $av_failed, true if the AV scanner deferred.
1425 Bugzilla 1078. Patch from John Horne.
1427 PP/06 Stop make process more reliably on build failure.
1428 Bugzilla 1087. Patch from Heiko Schlittermann.
1430 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1431 Bugzilla 1089. Patch from Heiko Schlittermann.
1433 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1434 Bugzilla 1131. Patch from Holger Weiß.
1436 PP/09 Handle IPv6 addresses with SPF.
1437 Bugzilla 860. Patch from Wolfgang Breyha.
1439 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1441 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1444 PP/11 match_* no longer expand right-hand-side by default.
1445 New compile-time build option, EXPAND_LISTMATCH_RHS.
1446 New expansion conditions, "inlist", "inlisti".
1448 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1450 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1452 PP/14 fix log_write() format string regression from TF/03.
1453 Bugzilla 1152. Patch from Dmitry Isaikin.
1459 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1461 PP/02 Harmonised TLS library version reporting; only show if debugging.
1462 Layout now matches that introduced for other libraries in 4.74 PP/03.
1464 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1466 PP/04 New "dns_use_edns0" global option.
1468 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1471 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1472 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1474 TK/01 Updated PolarSSL code to 0.14.2.
1475 Bugzilla 1097. Patch from Andreas Metzler.
1477 PP/07 Catch divide-by-zero in ${eval:...}.
1478 Fixes bugzilla 1102.
1480 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1483 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1484 format-string attack -- SECURITY: remote arbitrary code execution.
1486 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1487 time unintentionally subject to list matching rules, letting the header
1488 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1489 arbitrary string expansion). This allowed for information disclosure.
1491 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1492 INT_MIN/-1 -- value coerced to INT_MAX.
1498 NM/01 Workaround for PCRE version dependency in version reporting
1501 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1502 This fixes portability to compilers other than gcc, notably
1503 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1505 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1506 makefiles for portability to HP-UX and POSIX correctness.
1508 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1509 Also via indented variable definition in the Makefile.
1510 (Debugging by Oliver Heesakkers).
1512 PP/02 Restore caching of spamd results with expanded spamd_address.
1513 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1515 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1516 Improves build reliability. Fix from: Frank Elsner
1518 NM/02 Fix wide character breakage in the rfc2047 coding
1519 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1521 NM/03 Allow underscore in dnslist lookups
1522 Fixes bug 1026. Patch from Graeme Fowler
1524 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1525 Code patches from Adam Ciarcinski of NetBSD.
1527 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1530 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1531 is logged, to avoid truncation. Patch from John Horne.
1533 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1534 Patch from Jakob Hirsch.
1536 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1537 SQL string expansion failure details.
1538 Patch from Andrey Oktyabrski.
1540 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1541 Patch from Simon Arlott.
1543 PP/09 New lookups functionality failed to compile on old gcc which rejects
1544 extern declarations in function scope.
1545 Patch from Oliver Fleischmann
1547 PP/10 Use sig_atomic_t for flags set from signal handlers.
1548 Check getgroups() return and improve debugging.
1549 Fixed developed for diagnosis in bug 927 (which turned out to be
1552 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1553 Patch from Mark Zealey.
1555 PP/12 Bugzilla 1056: Improved spamd server selection.
1556 Patch from Mark Zealey.
1558 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1559 Based on patch from Heiko Schlittermann.
1561 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1562 Patch from Uwe Doering, sign-off by Michael Haardt.
1564 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1565 variable declaration deep within a block. Bug and patch from
1568 PP/15 lookups-Makefile IRIX compatibility coercion.
1570 PP/16 Make DISABLE_DKIM build knob functional.
1572 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1573 Patch by Simon Arlott
1575 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1576 variable argument macros. Our copy now differs from upstream.
1582 TF/01 Failure to get a lock on a hints database can have serious
1583 consequences so log it to the panic log.
1585 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1586 controlled using the smtp_confirmation log selector.
1588 TF/03 Include the error message when we fail to unlink a spool file.
1590 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1591 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1592 for maintaining out-of-tree patches for some time.
1594 PP/01 Bugzilla 139: Documentation and portability issues.
1595 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1596 Handle per-OS dynamic-module compilation flags.
1598 PP/02 Let /dev/null have normal permissions.
1599 The 4.73 fixes were a little too stringent and complained about the
1600 permissions on /dev/null. Exempt it from some checks.
1601 Reported by Andreas M. Kirchwitz.
1603 PP/03 Report version information for many libraries, including
1604 Exim version information for dynamically loaded libraries. Created
1605 version.h, now support a version extension string for distributors
1606 who patch heavily. Dynamic module ABI change.
1608 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1609 privilege escalation vulnerability whereby the Exim run-time user
1610 can cause root to append content of the attacker's choosing to
1613 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1616 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1617 If dropping privileges for untrusted macros, we disabled normal logging
1618 on the basis that it would fail; for the Exim run-time user, this is not
1619 the case, and it resulted in successful deliveries going unlogged.
1620 Fixed. Reported by Andreas Metzler.
1626 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1627 only prepend for the Resent-* case. Fixes regression introduced in
1628 Exim 4.70 by NM/22 for Bugzilla 607.
1630 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1631 increasing numbers of administrators be bitten by this.
1633 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1635 PP/03 Bugzilla 994: added openssl_options main configuration option.
1637 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1639 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1641 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1643 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1644 without a peer certificate, leading to a segfault because of an
1645 assumption that peers always have certificates. Be a little more
1646 paranoid. Problem reported by Martin Tscholak.
1648 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1649 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1650 NB: ClamAV planning to remove STREAM in "middle of 2010".
1651 CL also introduces -bmalware, various -d+acl logging additions and
1652 more caution in buffer sizes.
1654 PP/09 Implemented reverse_ip expansion operator.
1656 PP/10 Bugzilla 937: provide a "debug" ACL control.
1658 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1660 PP/12 Bugzilla 973: Implement --version.
1662 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1664 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1666 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1668 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1669 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1670 ignore trailing whitespace.
1672 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1674 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1677 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1678 Notification from Dr Andrew Aitchison.
1680 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1681 ExtendedDetectionInfo response format.
1682 Notification from John Horne.
1684 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1687 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1688 XSL and documented dependency on system catalogs, with examples of how
1691 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1694 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1695 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1696 configuration file which is writeable by the Exim user or group.
1698 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1699 of configuration files to cover files specified with the -C option if
1700 they are going to be used with root privileges, not just the default
1703 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1704 option (effectively making it always true).
1706 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1707 files to be used while preserving root privileges.
1709 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1710 that rogue child processes cannot use them.
1712 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1713 run-time user, instead of root.
1715 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1716 Exim run-time user without dropping privileges.
1718 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1719 result string, instead of calling string_vformat() twice with the same
1722 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1723 for other users. Others should always drop root privileges if they use
1724 -C on the command line, even for a whitelisted configure file.
1726 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1728 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1734 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1735 $data_path, and $header_path variables; fixed documentation bugs and
1738 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1739 exipick to access non-standard spools, including the "frozen" queue
1742 NM/01 Bugzilla 965: Support mysql stored procedures.
1743 Patch from Alain Williams
1745 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1747 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1748 Patch from Andreas Metzler
1750 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1751 Patch from Kirill Miazine
1753 NM/05 Bugzilla 671: Added umask to procmail example.
1755 JJ/03 installed exipick 20100323.0, fixing doc bug
1757 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1758 directory. Notification and patch from Dan Rosenberg.
1760 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1762 TK/02 Improve log output when DKIM signing operation fails.
1764 MH/01 Treat the transport option dkim_domain as a colon separated
1765 list, not as a single string, and sign the message with each element,
1766 omitting multiple occurences of the same signer.
1768 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1769 Bugzilla 985, 986. Patch by Simon Arlott
1771 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1772 Patch by Simon Arlott
1774 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1775 MBX locking. Notification from Dan Rosenberg.
1781 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1783 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1785 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1787 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1789 NM/04 Bugzilla 915: Flag broken perl installation during build.
1795 TK/01 Added patch by Johannes Berg that expands the main option
1796 "spamd_address" if it starts with a dollar sign.
1798 TK/02 Write list of recipients to X-Envelope-Sender header when building
1799 the mbox-format spool file for content scanning (suggested by Jakob
1802 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1803 (http://www.dcc-servers.net/) support via dccifd. Activated by
1804 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1806 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1807 by Mark Daniel Reidel <mr@df.eu>.
1809 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1810 When building exim an external PCRE library is now needed -
1811 PCRE is a system library on the majority of modern systems.
1812 See entry on PCRE_LIBS in EDITME file.
1814 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1815 conversation. Added nologin parameter to request.
1816 Patch contributed by Kirill Miazine.
1818 TF/01 Do not log submission mode rewrites if they do not change the address.
1820 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1822 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1823 log files in place. Contributed by Roberto Lima.
1825 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1827 TF/03 Bugzilla 615: When checking the local_parts router precondition
1828 after a local_part_suffix or local_part_prefix option, Exim now
1829 does not use the address's named list lookup cache, since this
1830 contains cached lookups for the whole local part.
1832 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1833 Robert Millan. Documentation is in experimental-spec.txt.
1835 TF/04 Bugzilla 668: Fix parallel build (make -j).
1837 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1839 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1840 Patch provided by Jan Srzednicki.
1842 TF/05 Leading white space used to be stripped from $spam_report which
1843 wrecked the formatting. Now it is preserved.
1845 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1846 that they are available at delivery time.
1848 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1850 TF/08 TLS error reporting now respects the incoming_interface and
1851 incoming_port log selectors.
1853 TF/09 Produce a more useful error message if an SMTP transport's hosts
1854 setting expands to an empty string.
1856 NM/06 Bugzilla 744: EXPN did not work under TLS.
1857 Patch provided by Phil Pennock.
1859 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1860 Patch provided by Richard Godbee.
1862 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1863 acl_smtp_notquit, added index entry.
1865 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1866 Patch provided by Eugene Bujak.
1868 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1869 accept(). Patch provided by Maxim Dounin.
1871 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1872 Patch provided by Phil Pennock.
1874 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1876 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1877 Patch provided by Brad "anomie" Jorsch.
1879 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1880 Patch provided by Dean Brooks.
1882 TK/05 Add native DKIM support (does not depend on external libraries).
1884 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1885 Patch provided by Graeme Fowler.
1887 NM/16 Bugzilla 851: Documentation example syntax fix.
1889 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1891 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1894 NM/19 Bugzilla 745: TLS version reporting.
1895 Patch provided by Phil Pennock.
1897 NM/20 Bugzilla 167: bool: condition support.
1898 Patch provided by Phil Pennock.
1900 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1901 clients. Patch provided by Phil Pennock.
1903 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1904 Patch provided by Brad "anomie" Jorsch.
1906 NM/23 Bugzilla 687: Fix misparses in eximstats.
1907 Patch provided by Heiko Schlittermann.
1909 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1910 Patch provided by Heiko Schlittermann.
1912 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1913 plus update to original patch.
1915 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1917 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1918 Patch provided by David Brownlee.
1920 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1922 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1924 NM/30 Bugzilla 888: TLS documentation bugfixes.
1926 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1928 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1929 Unlike the original bugzilla I have changed all shell scripts in src tree.
1931 NM/33 Bugzilla 898: Transport filter timeout fix.
1932 Patch by Todd Rinaldo.
1934 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1935 Patch by Serge Demonchaux.
1937 NM/35 Bugzilla 39: Base64 decode bug fixes.
1938 Patch by Jakob Hirsch.
1940 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1942 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1944 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1946 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1952 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1953 ALT-N's libdkim that I have put here:
1954 http://duncanthrax.net/exim-experimental/
1956 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1957 were called 'true' and it seems that C99 defines that as a reserved
1958 keyword to be used with 'bool' variable types. That means you could
1959 not include C99-style headers which use bools without triggering
1960 build errors in sieve.c.
1962 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1963 as mailq or other aliases. Changed the --help handling significantly
1964 to do whats expected. exim_usage() emits usage/help information.
1966 SC/01 Added the -bylocaldomain option to eximstats.
1968 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1970 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1972 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1978 PH/01 Another patch from the Sieve maintainer.
1980 PH/02 When an IPv6 address is converted to a string for single-key lookup
1981 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1982 dots are used instead of colons so that keys in lsearch files need not
1983 contain colons. This was done some time before quoting was made available
1984 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1985 (notated using the quote facility) so as to distinguish them from IPv4
1986 keys. This meant that lookups for IP addresses in host lists did not work
1987 for iplsearch lookups.
1989 This has been fixed by arranging for IPv6 addresses to be expressed with
1990 colons if the lookup type is iplsearch. This is not incompatible, because
1991 previously such lookups could never work.
1993 The situation is now rather anomalous, since one *can* have colons in
1994 ordinary lsearch keys. However, making the change in all cases is
1995 incompatible and would probably break a number of configurations.
1997 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2000 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2001 conversion specification without a maximum field width, thereby enabling
2002 a rogue spamd server to cause a buffer overflow. While nobody in their
2003 right mind would setup Exim to query an untrusted spamd server, an
2004 attacker that gains access to a server running spamd could potentially
2005 exploit this vulnerability to run arbitrary code as the Exim user.
2007 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2008 $primary_hostname instead of what libspf2 thinks the hosts name is.
2010 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2011 a directory entry by the name of the lookup key. Previously, if a
2012 symlink pointed to a non-existing file or a file in a directory that
2013 Exim lacked permissions to read, a lookup for a key matching that
2014 symlink would fail. Now it is enough that a matching directory entry
2015 exists, symlink or not. (Bugzilla 503.)
2017 PH/03 The body_linecount and body_zerocount variables are now exported in the
2020 PH/04 Added the $dnslist_matched variable.
2022 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2023 This means they are set thereafter only if the connection becomes
2026 PH/06 Added the client_condition to authenticators so that some can be skipped
2027 by clients under certain conditions.
2029 PH/07 The error message for a badly-placed control=no_multiline_responses left
2030 "_responses" off the end of the name.
2032 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2034 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2035 (without spaces) instead of just copying the configuration text.
2037 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2039 PH/11 Added $max_received_linelength.
2041 PH/12 Added +ignore_defer and +include_defer to host lists.
2043 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2044 way in which PCRE > 7.0 is built.
2046 PH/14 Implemented queue_only_load_latch.
2048 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2049 MAIL command. The effect was to mangle the value on 64-bit systems.
2051 PH/16 Another patch from the Sieve maintainer.
2053 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2055 PH/18 If a system quota error occurred while trying to create the file for
2056 a maildir delivery, the message "Mailbox is full" was not appended to the
2057 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2058 only to a quota excession during the actual writing of the file.
2060 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2061 characters?) which causes problems in log lines. The DN values are now
2062 passed through string_printing() before being added to log lines.
2064 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2065 and InterBase are left for another time.)
2067 PH/21 Added message_body_newlines option.
2069 PH/22 Guard against possible overflow in moan_check_errorcopy().
2071 PH/23 POSIX allows open() to be a macro; guard against that.
2073 PH/24 If the recipient of an error message contained an @ in the local part
2074 (suitably quoted, of course), incorrect values were put in $domain and
2075 $local_part during the evaluation of errors_copy.
2081 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2082 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2085 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2086 issue a MAIL command.
2088 PH/02 In an ACL statement such as
2090 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2092 if a client was not listed at all, or was listed with a value other than
2093 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2094 the condition was not true (as it should be), so access was not denied.
2095 The bug was that the ! inversion was incorrectly passed on to the second
2096 item. This has been fixed.
2098 PH/03 Added additional dnslists conditions == and =& which are different from
2099 = and & when the dns lookup returns more than one IP address.
2101 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2102 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2104 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2105 FSYNC, which compiles an option called disable_fsync that allows for
2106 bypassing fsync(). The documentation is heavily laced with warnings.
2108 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2110 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2111 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2112 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2113 including adding "make clean"; (3) Added -fPIC when compiling the test
2114 dynamically loaded module, to get rid of a warning.
2116 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2117 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2118 = 0s. The bug is otherwise harmless.
2120 PH/07 There was a bug in the dovecot authenticator such that the value of
2121 $auth1 could be overwritten, and so not correctly preserved, after a
2122 successful authentication. This usually meant that the value preserved by
2123 the server_setid option was incorrect.
2125 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2127 PH/09 Installed PCRE release 7.0.
2129 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2130 run for batched SMTP input. It is now run at the start of every message
2131 in the batch. While fixing this I discovered that the process information
2132 (output by running exiwhat) was not always getting set for -bs and -bS
2133 input. This is fixed, and it now also says "batched" for BSMTP.
2135 PH/11 Added control=no_pipelining.
2137 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2138 patch, slightly modified), and move the expansion of helo_data till after
2139 the connection is made in the smtp transport (so it can use these
2142 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2144 PH/14 Added log_selector = +pid.
2146 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2148 PH/16 Add ${if forany and ${if forall.
2150 PH/17 Added dsn_from option to vary the From: line in DSNs.
2152 PH/18 Flush SMTP output before performing a callout, unless control =
2153 no_callout_flush is set.
2155 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2156 was true (the default) a successful delivery failed to delete the retry
2157 item, thus causing premature timeout of the address. The bug is now
2160 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2162 PH/21 Long custom messages for fakedefer and fakereject are now split up
2163 into multiline responses in the same way that messages for "deny" and
2164 other ACL rejections are.
2166 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2167 with slight modification.
2169 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2170 draft, changing the syntax and factoring some duplicate code".
2172 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2173 for deliveries of the second and subsequent messages over the same SMTP
2176 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2177 ${reduce, with only minor "tidies".
2179 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2181 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2182 expansion side effects.
2184 PH/27 When a message times out after an over-quota error from an Exim-imposed
2185 quota, the bounce message says "mailbox is full". This message was not
2186 being given when it was a system quota that was exceeded. It now should
2189 MH/03 Made $recipients available in local_scan(). local_scan() already has
2190 better access to the recipient list through recipients_list[], but
2191 $recipients can be useful in postmaster-provided expansion strings.
2193 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2194 in the case of a MAIL command with additional options following the
2195 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2196 were accidentally chopped off.
2198 PH/29 SMTP synchronization checks are implemented when a command is read -
2199 there is a check that no more input is waiting when there shouldn't be
2200 any. However, for some commands, a delay in an ACL can mean that it is
2201 some time before the response is written. In this time, more input might
2202 arrive, invalidly. So now there are extra checks after an ACL has run for
2203 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2204 pipelining has not been advertised.
2206 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2208 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2209 returned in the SMTP response when smtp_return_error_details was false.
2210 This has been fixed.
2212 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2213 instead of the C I/O that was originally supplied, because problems were
2214 reported on Solaris.
2216 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2217 Exim which did not show up earlier: it was assuming that a call to
2218 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2219 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2220 was a macro that became an assignment, so it seemed to work. This has
2221 changed to a proper function call with a void return, hence the compile
2222 error. Exim's code has been fixed.
2224 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2227 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2229 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2231 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2232 because it turns out to be too big or there is a timeout) there is no
2233 "Completed" line in the log. When some messages of this type were
2234 selected by exigrep, they were listed as "not completed". Others were
2235 picked up by some special patterns. I have improved the selection
2236 criteria to be more general.
2238 PH/38 The host_find_failed option in the manualroute router can now be set
2239 to "ignore", to completely ignore a host whose IP address cannot be
2240 found. If all hosts are ignored, the behaviour is controlled by the new
2241 host_all_ignored option.
2243 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2244 homing or because of multiple MX records with /mx) generated more than
2245 one IP address, and the following item turned out to be the local host,
2246 all the secondary addresses of the first item were incorrectly removed
2247 from the list, along with the local host and any following hosts (which
2248 is what is supposed to happen).
2250 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2251 whoever called Exim into the -H file. In the case of the daemon it was
2252 behaving confusingly. When first started, it used values for whoever
2253 started the daemon, but after a SIGHUP it used the Exim user (because it
2254 calls itself on a restart). I have changed the code so that it now always
2257 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2258 message are rejected with the same error (e.g. no authentication or bad
2259 sender address), and a DATA command is nevertheless sent (as can happen
2260 with PIPELINING or a stupid MUA), the error message that was given to the
2261 RCPT commands is included in the rejection of the DATA command. This is
2262 intended to be helpful for MUAs that show only the final error to their
2265 PH/42 Another patch from the Sieve maintainer.
2267 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2268 Eximstats - Fixed some broken HTML links and added missing column headers
2270 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2271 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2273 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2275 PH/43 Yet another patch from the Sieve maintainer.
2277 PH/44 I found a way to check for a TCP/IP connection going away before sending
2278 the response to the final '.' that terminates a message, but only in the
2279 case where the client has not sent further data following the '.'
2280 (unfortunately, this is allowed). However, in many cases there won't be
2281 any further data because there won't be any more messages to send. A call
2282 to select() can be used: if it shows that the input is "ready", there is
2283 either input waiting, or the socket has been closed. An attempt to read
2284 the next input character can distinguish the two cases. Previously, Exim
2285 would have sent an OK response which the client would never have see.
2286 This could lead to message repetition. This fix should cure that, at
2287 least in a lot of common cases.
2289 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2290 advertised in response to EHLO.
2296 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2297 fixed by 4.65/MH/01 (is this a record?) are fixed:
2299 (i) An empty string was always treated as zero by the numeric comparison
2300 operators. This behaviour has been restored.
2302 (ii) It is documented that the numeric comparison operators always treat
2303 their arguments as decimal numbers. This was broken in that numbers
2304 starting with 0 were being interpreted as octal.
2306 While fixing these problems I realized that there was another issue that
2307 hadn't been noticed. Values of message_size_limit (both the global option
2308 and the transport option) were treated as octal if they started with 0.
2309 The documentation was vague. These values are now always treated as
2310 decimal, and I will make that clear in the documentation.
2316 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2317 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2320 MH/01 Don't check that the operands of numeric comparison operators are
2321 integers when their expansion is in "skipping" mode (fixes bug
2322 introduced by 4.64-PH/07).
2324 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2325 child addresses, Exim now panics and dies. Previously, because the count
2326 is held in a short int, deliveries were likely to be lost. As such a
2327 large number of recipients for a single message is ridiculous
2328 (performance will be very, very poor), I have chosen to impose a limit
2329 rather than extend the field.
2335 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2336 leftover -K file (the existence of which was triggered by #402).
2337 While we were at it, introduced process PID as part of the -K
2338 filename. This should rule out race conditions when creating
2341 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2342 processing considerably. Previous code took too long for large mails,
2343 triggering a timeout which in turn triggers #401.
2345 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2346 in the DK code in transports.c. sendfile() is not really portable,
2347 hence the _LINUX specificness.
2349 TF/01 In the add_headers option to the mail command in an Exim filter,
2350 there was a bug that Exim would claim a syntax error in any
2351 header after the first one which had an odd number of characters
2354 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2355 callout verification, Exim cached a "reject" for the entire domain. This
2356 is correct for most verifications, but it is not correct for a recipient
2357 verification with use_sender or use_postmaster set, because in that case
2358 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2359 case of MAIL FROM:<> rejection from other early rejections (e.g.
2360 rejection of HELO). When verifying a recipient using a non-null MAIL
2361 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2362 Whatever the result of the callout, the value of the domain cache is
2363 left unchanged (for any other kind of callout, getting as far as trying
2364 RCPT means that the domain itself is ok).
2366 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2369 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2370 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2371 session, the daemon ignores ECONNECT errors and logs others; it now
2372 ignores EPIPE as well.
2374 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2375 (quoted-printable decoding).
2377 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2378 later the small subsequent patch to fix an introduced bug.
2380 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2382 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2384 PH/08 An error is now given if message_size_limit is specified negative.
2386 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2387 to be given (somewhat) arbitrary names.
2389 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2392 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2393 miscellaneous code fixes
2395 PH/10 Added the log_reject_target ACL modifier to specify where to log
2398 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2399 hostname. This is wrong, because it relates to the incoming message (and
2400 probably the interface on which it is arriving) and not to the outgoing
2401 callout (which could be using a different interface). This has been
2402 changed to use the value of the helo_data option from the smtp transport
2403 instead - this is what is used when a message is actually being sent. If
2404 there is no remote transport (possible with a router that sets up host
2405 addresses), $smtp_active_hostname is used.
2407 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2408 tweaks were necessary in order to get it to work (see also 21 below):
2409 (a) The code assumed that strncpy() returns a negative number on buffer
2410 overflow, which isn't the case. Replaced with Exim's string_format()
2412 (b) There were several signed/unsigned issues. I just did the minimum
2413 hacking in of casts. There is scope for a larger refactoring.
2414 (c) The code used strcasecmp() which is not a standard C function.
2415 Replaced with Exim's strcmpic() function.
2416 (d) The code set only $1; it now sets $auth1 as well.
2417 (e) A simple test gave the error "authentication client didn't specify
2418 service in request". It would seem that Dovecot has changed its
2419 interface. Fortunately there's a specification; I followed it and
2420 changed what the client sends and it appears to be working now.
2422 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2425 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2426 address A is aliased to B and C, where B exists and C does not. Without
2427 -v the output is "A verified" because verification stops after a
2428 successful redirection if more than one address is generated. However,
2429 with -v the child addresses are also verified. Exim was outputting "A
2430 failed to verify" and then showing the successful verification for C,
2431 with its parentage. It now outputs "B failed to verify", showing B's
2432 parentage before showing the successful verification of C.
2434 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2435 look up a TXT record in a specific list after matching in a combined
2438 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2439 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2440 they consult the DNS. I had assumed they would set it the way they
2441 wanted; and indeed my experiments on Linux seem to show that in some
2442 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2443 To be on the safe side, however, I have now made the interface to
2444 host_find_byname() similar to host_find_bydns(), with an argument
2445 containing the DNS resolver options. The host_find_byname() function now
2446 sets these options at its start, just as host_find_bydns() does. The smtp
2447 transport options dns_qualify_single and dns_search_parents are passed to
2448 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2449 of host_find_byname() use the default settings of RES_DEFNAMES
2450 (qualify_single) but not RES_DNSRCH (search_parents).
2452 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2453 spool_read_header() do less string testing, by means of a preliminary
2454 switch on the second character of optional "-foo" lines. (This is
2455 overdue, caused by the large number of possibilities that now exist.
2456 Originally there were few.) While I was there, I also converted the
2457 str(n)cmp tests so they don't re-test the leading "-" and the first
2458 character, in the hope this might squeeze out yet more improvement.
2460 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2461 flag allowing group syntax was set by the header_syntax check but not
2462 turned off, possible causing trouble later; (2) The flag was not being
2463 set at all for the header_verify test, causing "group"-style headers to
2464 be rejected. I have now set it in this case, and also caused header_
2465 verify to ignore an empty address taken from a group. While doing this, I
2466 came across some other cases where the code for allowing group syntax
2467 while scanning a header line wasn't quite right (mostly, not resetting
2468 the flag correctly in the right place). These bugs could have caused
2469 trouble for malformed header lines. I hope it is now all correct.
2471 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2472 with the "reply" argument non-NULL. The code, however (which originally
2473 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2474 but it didn't always do it. This confused somebody who was copying the
2475 code for some other use. I have removed all the tests.
2477 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2478 feature that was used to support insecure browsers during the U.S. crypto
2479 embargo. It requires special client support, and Exim is probably the
2480 only MTA that supported it -- and would never use it because real RSA is
2481 always available. This code has been removed, because it had the bad
2482 effect of slowing Exim down by computing (never used) parameters for the
2483 RSA_EXPORT functionality.
2485 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2486 authenticator to fail if there's a tab character in the incoming data
2487 (there should never be unless someone is messing about, as it's supposed
2488 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2489 if the connection is using TLS or if the remote IP is the same as the
2490 local IP, and the "valid-client-cert option" if a client certificate has
2493 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2494 authenticators. This can be used for authorization after authentication
2495 succeeds. (In the case of plaintext, it servers for both authentication
2498 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2499 if any retry times were supplied.
2501 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2502 connection, where there is no client IP address to check. In this
2503 situation, the verify now always succeeds.
2505 PH/25 Applied John Jetmore's -Mset patch.
2507 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2509 PH/27 In a string expansion for a processed (not raw) header when multiple
2510 headers of the same name were present, leading whitespace was being
2511 removed from all of them, but trailing whitespace was being removed only
2512 from the last one. Now trailing whitespace is removed from each header
2513 before concatenation. Completely empty headers in a concatenation (as
2514 before) are ignored.
2516 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2517 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2519 PH/29 [Removed. This was a change that I later backed out, and forgot to
2520 correct the ChangeLog entry (that I had efficiently created) before
2521 committing the later change.]
2523 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2524 address errors (4xx response to RCPT) over the same connection as other
2525 messages routed to the same hosts. Such deliveries are always "forced",
2526 so retry times are not inspected. This resulted in far too many retries
2527 for the affected addresses. The effect occurred only when there were more
2528 hosts than the hosts_max_try setting in the smtp transport when it had
2529 the 4xx errors. Those hosts that it had tried were not added to the list
2530 of hosts for which the message was waiting, so if all were tried, there
2531 was no problem. Two fixes have been applied:
2533 (i) If there are any address or message errors in an SMTP delivery, none
2534 of the hosts (tried or untried) are now added to the list of hosts
2535 for which the message is waiting, so the message should not be a
2536 candidate for sending over the same connection that was used for a
2537 successful delivery of some other message. This seems entirely
2538 reasonable: after all the message is NOT "waiting for some host".
2539 This is so "obvious" that I'm not sure why it wasn't done
2540 previously. Hope I haven't missed anything, but it can't do any
2541 harm, as the worst effect is to miss an optimization.
2543 (ii) If, despite (i), such a delivery is accidentally attempted, the
2544 routing retry time is respected, so at least it doesn't keep
2545 hammering the server.
2547 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2548 in ${readsocket because some servers need this prod.
2550 PH/32 Added some extra debug output when updating a wait-xxx database.
2552 PH/33 The hint "could be header name not terminated by colon", which has been
2553 given for certain expansion errors for a long time, was not being given
2554 for the ${if def:h_colon_omitted{... case.
2556 PH/34 The spec says: "With one important exception, whenever a domain list is
2557 being scanned, $domain contains the subject domain." There was at least
2558 one case where this was not true.
2560 PH/35 The error "getsockname() failed: connection reset by peer" was being
2561 written to the panic log as well as the main log, but it isn't really
2562 panic-worthy as it just means the connection died rather early on. I have
2563 removed the panic log writing for the ECONNRESET error when getsockname()
2566 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2567 runs only) independently of the message's sender address. This meant
2568 that, if the 4xx error was in fact related to the sender, a different
2569 message to the same recipient with a different sender could confuse
2570 things. In particular, this can happen when sending to a greylisting
2571 server, but other circumstances could also provoke similar problems.
2572 I have changed the default so that the retry time for these errors is now
2573 based a combination of the sender and recipient addresses. This change
2574 can be overridden by setting address_retry_include_sender=false in the
2577 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2578 remote server are returned as part of bounce messages. This was not
2579 happening for LMTP over a pipe (the lmtp transport), but now it is the
2580 same for both kinds of LMTP.
2582 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2583 in header lines that were in fact CNAMEs. This is no longer the case.
2585 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2586 and queue runs started by the daemon processed all messages. This has
2587 been fixed so that -R and -S can now usefully be given with -q<time>.
2589 PH/40 Import PCRE release 6.7 (fixes some bugs).
2591 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2593 PH/42 Give an error if -q is specified more than once.
2595 PH/43 Renamed the variables $interface_address and $interface_port as
2596 $received_ip_address and $received_port, to make it clear that these
2597 values apply to message reception, and not to the outgoing interface when
2598 a message is delivered. (The old names remain recognized, of course.)
2600 PH/44 There was no timeout on the connect() call when using a Unix domain
2601 socket in the ${readsocket expansion. There now is.
2603 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2604 be meaningful with "accept".
2606 SC/01 Eximstats V1.43
2607 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2609 SC/02 Eximstats V1.44
2610 Use a glob alias rather than an array ref in the generated
2611 parser. This improves both readability and performance.
2613 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2614 Collect SpamAssassin and rejection statistics.
2615 Don't display local sender or destination tables unless
2616 there is data to show.
2617 Added average volumes into the top table text output.
2619 SC/04 Eximstats V1.46
2620 Collect data on the number of addresses (recipients)
2621 as well as the number of messages.
2623 SC/05 Eximstats V1.47
2624 Added 'Message too big' to the list of mail rejection
2625 reasons (thanks to Marco Gaiarin).
2627 SC/06 Eximstats V1.48
2628 Mainlog lines which have GMT offsets and are too short to
2629 have a flag are now skipped.
2631 SC/07 Eximstats V1.49 (Alain Williams)
2632 Added the -emptyok flag.
2634 SC/08 Eximstats V1.50
2635 Fixes for obtaining the IP address from reject messages.
2637 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2638 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2639 whitespace changes from 4.64-PH/27
2641 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2644 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2645 are found, allow negative numbers in numeric criteria)
2647 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2649 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2650 to match changes made in 4.64-PH/43
2652 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2654 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2655 "domain" is actually an IP address, and force a failure. This locks out
2656 those revolvers/nameservers that support "A-for-A" lookups, in
2657 contravention of the specifications.
2659 PH/48 When a host name was looked up from an IP address, and the subsequent
2660 forward lookup of the name timed out, the host name was left in
2661 $sender_host_name, contrary to the specification.
2663 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2664 restricted to single-key lookups, Exim was not diagnosing an error if
2665 * or *@ was used with a query-style lookup.
2667 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2669 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2670 long ago, but noone interested enough thought of it. Let's just say that
2671 the "1.1" means that there are some new functions that weren't there at
2672 some point in the past.
2674 PH/51 Error processing for expansion failure of helo_data from an smtp
2675 transport during callout processing was broken.
2677 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2678 tested/used via the -bh/-bhc/-bs options.
2680 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2681 bug, fixed in subsequent PCRE releases).
2683 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2684 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2686 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2692 SC/01 Use a glob alias rather than an array ref in eximstats generated
2693 parser. This improves both readability and performance.
2695 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2696 Don't display local sender or destination tables in eximstats unless
2697 there is data to show.
2698 Added average volumes into the eximstats top table text output.
2700 SC/03 Collect data on the number of addresses (recipients) as well
2701 as the number of messages in eximstats.
2703 TF/01 Correct an error in the documentation for the redirect router. Exim
2704 does (usually) call initgroups() when daemonizing.
2706 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2707 with consistent privilege compared to when running as a daemon.
2709 TF/03 Note in the spec that $authenticated_id is not set for local
2710 submissions from trusted users.
2712 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2713 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2715 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2716 by adding some example configuration directives to the default
2717 configuration file. A little bit of work is required to uncomment the
2718 directives and define how usernames and passwords are checked, but
2719 there is now a framework to start from.
2721 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2722 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2723 without this. I don't know how relevant this is to other LDAP libraries.
2725 PH/02 Add the verb name to the "unknown ACL verb" error.
2727 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2729 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2731 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2732 directory not expanded when it should be if an expanded home directory
2733 was set for the address (which is overridden by the transport).
2735 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2738 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2739 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2740 because it is too late at that time, and has no effect.
2742 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2743 security issue with \' (bugzilla #107). I could not use the
2744 PQescapeStringConn() function, because it needs a PGconn value as one of
2747 PH/08 When testing addresses using -bt, indicate those final addresses that
2748 are duplicates that would not cause an additional delivery. At least one
2749 person was confused, thinking that -bt output corresponded to deliveries.
2750 (Suppressing duplicates isn't a good idea as you lose the information
2751 about possibly different redirections that led to the duplicates.)
2753 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2754 systems where poll() doesn't work, in particular OS X.
2756 PH/10 Added more information to debugging output for retry time not reached.
2758 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2759 operations in malware.c.
2761 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2764 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2765 syslog_duplication set false, log lines that would normally be written
2766 both the the main log and to the reject log were not written to syslog at
2769 PH/14 In the default configuration, change the use of "message" in ACL warn
2770 statements to "add_header".
2772 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2773 not followed by a command (e.g. "seen endif").
2775 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2776 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2779 PH/17 Added extra conditions to the default value of delay_warning_condition
2783 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2784 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2785 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2788 The Auto-Submitted: and various List- headers are standardised, whereas I
2789 don't think Precedence: ever was.
2791 PH/18 Refactored debugging code in route_finduser() to show more information,
2792 in particular, the error code if getpwnam() issues one.
2794 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2795 This is apparently needed in addition to the PH/07 change above to avoid
2796 any possible encoding problems.
2798 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2799 but not after initializing Perl.
2801 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2802 output them only if debugging. By default they are written stderr,
2803 apparently, which is not desirable.
2805 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2808 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2811 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2813 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2814 authenticated or an ident call has been made. Suppress the default
2815 values for $authenticated_id and $authenticated_sender (but permit -oMai
2816 and -oMas) when testing with -bh.
2818 PH/24 Re-jigged the order of the tests in the default configuration so that the
2819 tests for valid domains and recipients precede the DNS black list and CSA
2820 tests, on the grounds that those ones are more expensive.
2822 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2823 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2824 This bug exists in every version of Exim that I still have, right back to
2827 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2828 However, an attempt to turn on case-sensitivity in a regex key by
2829 including (?-i) didn't work because the subject string was already
2830 lowercased, and the effects were non-intuitive. It turns out that a
2831 one-line patch can be used to allow (?-i) to work as expected.
2837 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2838 other effects) broke the use of negated acl sub-conditions.
2840 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2843 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2844 "Deny" causes Exim to reject the incoming connection with a 554 error.
2845 Unfortunately, if there is a major crisis, such as a disk failure,
2846 tcp-wrappers gives "deny", whereas what one would like would be some
2847 kind of temporary error. A kludge has been added to help with this.
2848 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2849 554 error is used if errno is still zero or contains ENOENT (which occurs
2850 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2853 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2855 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2856 errors. Otherwise a message that provokes a temporary error (when other
2857 messages do not) can cause a whole host to time out.
2859 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2860 addresses were routed directly to files or pipes from a redirect router.
2861 File deliveries just didn't batch; pipe deliveries might have suffered
2864 PH/06 A failure to get a lock for a hints database would erroneously always say
2865 "Failed to get write lock", even when it was really a read lock.
2867 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2868 of 0600. This has been changed to use the value of the lockfile_mode
2869 option (which defaults to 0600).
2871 PH/08 Applied small patch from the Sieve maintainer.
2873 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2874 folder from quota calculations, a direct delivery into this folder messed
2875 up the contents of the maildirsize file. This was because the regex was
2876 used only to exclude .Trash (or whatever) when the size of the mailbox
2877 was calculated. There was no check that a delivery was happening into an
2878 excluded directory. This bug has been fixed by ignoring all quota
2879 processing for deliveries into excluded directories.
2881 PH/10 Added the maildirfolder_create_regex option to appendfile.
2887 PH/01 The code for finding all the local interface addresses on a FreeBSD
2888 system running IPv6 was broken. This may well have applied to all BSD
2889 systems, as well as to others that have similar system calls. The broken
2890 code found IPv4 interfaces correctly, but gave incorrect values for the
2891 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2892 that it would not match correctly against @[] and not recognize the IPv6
2895 PH/02 The ipliteral router was not recognizing addresses of the form user@
2896 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2898 PH/03 Added disable_ipv6.
2900 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2901 decoded form, because it is most often used to construct To: headers
2902 lines in autoreplies, and the decoded form may well be syntactically
2903 invalid. However, $reply_address has leading white space removed, and all
2904 newlines turned into spaces so that the autoreply transport does not
2907 PH/05 If group was specified without a user on a router, and no group or user
2908 was specified on a transport, the group from the router was ignored.
2910 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2911 for visible compile-time settings that can be used to change these
2912 numbers, for those that want even more. Backwards compatibility with old
2913 spool files has been maintained. However, going back to a previous Exim
2914 release will lost any variables that are in spool files.
2916 PH/07 Two small changes when running in the test harness: increase delay when
2917 passing a TCP/IP connection to a new process, in case the original
2918 process has to generate a bounce, and remove special handling of
2919 127.0.0.2 (sic), which is no longer necessary.
2921 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2922 be the same on different OS.
2924 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2927 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2928 whether --show-vars was specified or not
2930 JJ/02 exipick: Added support for new ACL variable spool format introduced
2933 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2934 syntactically invalid From: or Reply-to: line, and a filter used this to
2935 generate an autoreply, and therefore failed to obtain an address for the
2936 autoreply, Exim could try to deliver to a non-existent relative file
2937 name, causing unrelated and misleading errors. What now happens is that
2938 it logs this as a hard delivery error, but does not attempt to create a
2941 PH/11 The exinext utility has a -C option for testing purposes, but although
2942 the given file was scanned by exinext itself; it wasn't being passed on
2943 when Exim was called.
2945 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2946 an end-of-file indication when reading a command response.
2948 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2949 compiled. In many other places in Exim, IPv6 addresses are always
2950 recognized, so I have changed this. It also means that IPv4 domain
2951 literals of the form [IPV4:n.n.n.n] are now always recognized.
2953 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2954 used if the router is not running as root, for example, when verifying at
2955 ACL time, or when using -bh. The debugging output from this situation was
2956 non-existent - all you got was a failure to exec. I have made two
2959 (a) Failures to set uid/gid, the current directory, or a process leader
2960 in a subprocess such as that created by queryprogram now generate
2961 suitable debugging output when -d is set.
2963 (b) The queryprogram router detects when it is not running as root,
2964 outputs suitable debugging information if -d is set, and then runs
2965 the subprocess without attempting to change uid/gid.
2967 PH/15 Minor change to Makefile for building test_host (undocumented testing
2970 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2971 additional section of a DNS packet that returns MX or SRV records.
2972 Instead, it always explicitly searches for A/AAAA records. This avoids
2973 major problems that occur when a DNS server includes only records of one
2974 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2975 fixed another bug: if SRV records were looked up and the corresponding
2976 address records were *not* found in the additional section, the port
2977 values from the SRV records were lost.
2979 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2980 using the correct key (the original address) when searching the retry
2981 rules in order to find which one to use for generating the retry hint.
2983 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2984 adding the default one. Similarly, if it contains a Reply-To: header, the
2985 errors_reply_to option, if set, is not used.
2987 PH/19 When calculating a retry time, Exim used to measure the "time since
2988 failure" by looking at the "first failed" field in the retry record. Now
2989 it does not use this if it is later than than the arrival time of the
2990 message. Instead it uses the arrival time. This makes for better
2991 behaviour in cases where some deliveries succeed, thus re-setting the
2992 "first failed" field. An example is a quota failure for a huge message
2993 when small messages continue to be delivered. Without this change, the
2994 "time since failure" will always be short, possible causing more frequent
2995 delivery attempts for the huge message than are intended.
2996 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2998 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2999 $1, $2, $3) because the numerical variables can be reset during some
3000 expansion items (e.g. "match"), thereby losing the authentication data.
3002 PH/21 Make -bV show the size of off_t variables so that the test suite can
3003 decide whether to run tests for quotas > 2G.
3005 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3006 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3007 filecount value is greater than 2G or if a quota value is greater than 2G
3008 on a system where the size of off_t is not greater than 4, a panic error
3011 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3012 never match. The debug and -bh output now contains an explicit error
3013 message indicating a malformed IPv4 address or mask.
3015 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3016 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3017 PH/23 above applies.
3019 PH/25 Do not write to syslog when running in the test harness. The only
3020 occasion when this arises is a failure to open the main or panic logs
3021 (for which there is an explicit test).
3023 PH/26 Added the /no_tell option to "control=freeze".
3025 PH/27 If a host name lookup failed very early in a connection, for example, if
3026 the IP address matched host_lookup and the reverse lookup yielded a name
3027 that did not have a forward lookup, an error message of the form "no IP
3028 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3029 could be logged. Now it outputs the IP address instead of "NULL".
3031 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3032 allows the sender and the authenticated sender to be set when
3033 submitting a message from within Exim. Since child_open_exim() is
3034 documented for local_scan(), the new function should be too.
3036 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3037 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3038 results in an empty string is now treated as unset.
3040 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3042 PH/31 Added sender_verify_fail logging option.
3044 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3045 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3046 tidied the source and removed it altogether.
3048 PH/33 When a queue run was abandoned because the load average was too high, a
3049 log line was always written; now it is written only if the queue_run log
3050 selector is set. In addition, the log line for abandonment now contains
3051 information about the queue run such as the pid. This is always present
3052 in "start" and "stop" lines but was omitted from the "abandon" line.
3054 PH/34 Omit spaces between a header name and the colon in the error message that
3055 is given when verify = headers_syntax fails (if there are lots of them,
3056 the message gets confusing).
3058 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3059 names, as there are now some PTR records that contain slashes. This check
3060 is only to protect against broken name servers that fall over on strange
3061 characters, so the fact that it applies to all lookups doesn't matter.
3063 PH/36 Now that the new test suite is complete, we can remove some of the
3064 special code in Exim that was needed for the old test suite. For example,
3065 sorting DNS records because real resolvers return them in an arbitrary
3066 order. The new test suite's fake resolver always returns records in the
3069 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3070 bounces) except when queue_only is set, to avoid logging races between
3071 the different processes.
3073 PH/38 Panic-die if .include specifies a non-absolute path.
3075 PH/39 A tweak to the "H" retry rule from its user.
3077 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3078 a label. They prevented compilation on older perls.
3080 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3081 a warning to be raised on newish perls.
3083 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3084 on queue. Changes to match documented behaviour of showing count of
3085 messages matching specified criteria.
3087 PH/40 Changed the default ident timeout from 30s to 5s.
3089 PH/41 Added support for the use of login_cap features, on those BSD systems
3090 that have them, for controlling the resources used by pipe deliveries.
3092 PH/42 The content-scanning code uses fopen() to create files in which to put
3093 message data. Previously it was not paying any attention to the mode of
3094 the files. Exim runs with umask(0) because the rest of the code creates
3095 files with open(), and sets the required mode explicitly. Thus, these
3096 files were ending up world-writeable. This was not a big issue, because,
3097 being within the spool directory, they were not world-accessible. I have
3098 created a function called modefopen, which takes an additional mode
3099 argument. It sets umask(777), creates the file, chmods it to the required
3100 mode, then resets the umask. All the relevant calls to fopen() in the
3101 content scanning code have been changed to use this function.
3103 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3104 to 24 hours. This avoids potential overflow problems when processing G
3105 and H retry rules. I suspect nobody ever tinkers with this value.
3107 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3109 PH/45 When the plaintext authenticator is running as a client, the server's
3110 challenges are checked to ensure they are valid base64 strings. By
3111 default, the authentication attempt is cancelled if an invalid string is
3112 received. Setting client_ignore_invalid_base64 true ignores these errors.
3113 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3114 they are received. Thus, the responses can be made to depend on the
3115 challenges. If an invalid string is ignored, an empty string is placed in
3118 PH/46 Messages that are created by the autoreply transport now contains a
3119 References: header, in accordance with RFCs 2822 and 3834.
3121 PH/47 Added authenticated_sender_force to the smtp transport.
3123 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3125 PH/49 Installed latest patch from the Sieve maintainer.
3127 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3128 was also set, the appendfile transport was unnecessarily scanning a
3129 directory of message files (e.g. for maildir delivery) to find the count
3130 of files (along with the size), even though it did not need this
3131 information. It now does the scan only if it needs to find either the
3132 size of the count of files.
3134 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3136 PH/52 Two bugs concerned with error handling when the smtp transport is
3139 (i) Exim was not creating retry information for temporary errors given
3140 for individual recipients after the DATA command when the smtp transport
3141 was used in LMTP mode. This meant that they could be retried too
3142 frequently, and not timed out correctly.
3144 (ii) Exim was setting the flag that allows error details to be returned
3145 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3146 recipients that were returned after the DATA command.
3148 PH/53 This is related to PH/52, but is more general: for any failing address,
3149 when detailed error information was permitted to be returned to the
3150 sender, but the error was temporary, then after the final timeout, only
3151 "retry timeout exceeded" was returned. Now it returns the full error as
3152 well as "retry timeout exceeded".
3154 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3155 do this, and (what is worse) MTAs that accept it.
3157 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3158 will now be deprecated.
3160 PH/56 New os.c-cygwin from the Cygwin maintainer.
3162 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3163 formats (previously only available in exim formats via -bpr, -bpru,
3164 and -bpra. Now also available in native and exiqgrep formats)
3166 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3167 with very large, slow to parse queues
3169 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3171 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3173 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3174 responses to authentication challenges, though it was showing the
3175 challenges; (ii) I've removed the CR characters from the debug output for
3178 PH/58 Allow for the insertion of a newline as well as a space when a string
3179 is turned into more than one encoded-word during RFC 2047 encoding. The
3180 Sieve code now uses this.
3182 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3183 data_4xx, lost_connection, tls_required.
3185 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3186 message was being sent as an SMTP response.
3188 PH/61 Add -l and -k options to exicyclog.
3190 PH/62 When verifying, if an address was redirected to one new address, so that
3191 verification continued, and the new address failed or deferred after
3192 having set something in $address_data, the value of $address_data was not
3193 passed back to the ACL. This was different to the case when no
3194 redirection occurred. The value is now passed back in both cases.
3196 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3197 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3198 use login_cap.h, so on its own it isn't the distinguishing feature. The
3199 new name refers directly to the setclassresources() function.
3201 PH/65 Added configuration files for NetBSD3.
3203 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3205 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3206 is preferred over IPv4.
3208 PH/68 The bounce_return_message and bounce_return_body options were not being
3209 honoured for bounces generated during the reception of non-SMTP messages.
3210 In particular, this applied to messages rejected by the ACL. This bug has
3211 been fixed. However, if bounce_return_message is true and bounce_return_
3212 body is false, the headers that are returned for a non-SMTP message
3213 include only those that have been read before the error was detected.
3214 (In the case of an ACL rejection, they have all been read.)
3216 PH/69 The HTML version of the specification is now built in a directory called
3217 spec_html instead of spec.html, because the latter looks like a path with
3218 a MIME-type, and this confuses some software.
3220 PH/70 Catch two compiler warnings in sieve.c.
3222 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3223 function verify_get_ident() calls ip_connect() to connect a socket, but
3224 if the "connect()" function timed out, ip_connect() used to close the
3225 socket. However, verify_get_ident() also closes the socket later, and in
3226 between Exim writes to the log, which may get opened at this point. When
3227 the socket was closed in ip_connect(), the log could get the same file
3228 descriptor number as the socket. This naturally causes chaos. The fix is
3229 not to close the socket in ip_connect(); the socket should be closed by
3230 the function that creates it. There was only one place in the code where
3231 this was missing, in the iplookup router, which I don't think anybody now
3232 uses, but I've fixed it anyway.
3234 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3235 well as to direct DNS lookups. Otherwise the handling of names in host
3236 lists is inconsistent and therefore confusing.
3242 PH/01 Two changes to the default runtime configuration:
3244 (1) Move the checks for relay_from_hosts and authenticated clients from
3245 after to before the (commented out) DNS black list checks.
3247 (2) Add control=submission to the relay_from_hosts and authenticated
3248 clients checks, on the grounds that messages accepted by these
3249 statements are most likely to be submissions.
3251 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3253 (1) Generate an error if the third argument for the ${prvs expansion is
3256 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3259 (3) Reset the variables that are obtained from the first argument of
3260 ${prvscheck and used in the second argument before leaving the code,
3261 because their memory is reclaimed, so using them afterwards may do
3264 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3265 one (it's much easier than Tom thought :-).
3267 (5) Because of (4), we can now allow for the use of $prvscheck_result
3268 inside the third argument.
3270 PH/03 For some reason, the default setting of PATH when running a command from
3271 a pipe transport was just "/usr/bin". I have changed it to
3274 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3275 anything to be listed in the output from -bV.
3277 PH/05 When a filter generated an autoreply, the entire To: header line was
3278 quoted in the delivery log line, like this:
3280 => >A.N.Other <ano@some.domain> <original@ddress> ...
3282 This has been changed so that it extracts the operative address. There
3283 may be more than one such address. If so, they are comma-separated, like
3286 => >ano@some.domain,ona@other.domain <original@ddress> ...
3288 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3289 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3290 not being looked up in the rDNS to get a host name, Exim was showing the
3291 IP address twice in Received: lines, even though the IP addresses were
3292 identical. For example:
3294 Received: from [1.2.3.4] (helo=[1.2.3.4])
3296 However, if the real host name was known, it was omitting the HELO data
3297 if it matched the actual IP address. This has been tidied up so that it
3298 doesn't show the same IP address twice.
3300 PH/07 When both +timestamp and +memory debugging was on, the value given by
3301 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3302 called by the debug printing, thereby overwriting the timestamp buffer.
3303 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3305 PH/08 When the original message was included in an autoreply transport, it
3306 always said "this is a copy of the message, including all the headers",
3307 even if body_only or headers_only was set. It now gives an appropriate
3310 PH/09 Applied a patch from the Sieve maintainer which:
3312 o fixes some comments
3313 o adds the (disabled) notify extension core
3314 o adds some debug output for the result of if/elsif tests
3315 o points to the current vacation draft in the documentation
3316 and documents the missing references header update
3320 o fixes a bug in processing the envelope test (when testing
3321 multiple envelope elements, the last element determined the
3324 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3325 Electronic Mail") by including:
3327 Auto-submitted: auto-generated
3329 in the messages that it generates (bounce messages and others, such as
3330 warnings). In the case of bounce messages for non-SMTP messages, there was
3331 also a typo: it was using "Auto_submitted" (underscore instead of
3332 hyphen). Since every message generated by Exim is necessarily in response
3333 to another message, thes have all been changed to:
3335 Auto-Submitted: auto-replied
3337 in accordance with these statements in the RFC:
3339 The auto-replied keyword:
3341 - SHOULD be used on messages sent in direct response to another
3342 message by an automatic process,
3344 - MUST NOT be used on manually-generated messages,
3346 - MAY be used on Delivery Status Notifications (DSNs) and Message
3347 Disposition Notifications (MDNs),
3349 - MUST NOT be used on messages generated by automatic or periodic
3350 processes, except for messages which are automatic responses to
3353 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3354 to the default Received: header definition.
3356 PH/12 Added log selector acl_warn_skipped (default on).
3358 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3359 variables because (a) they are in the wrong storage pool and (b) even if
3360 they were copied, it wouldn't work properly because of the caching.
3362 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3363 checking when decoding. Apparently there are clients that generate
3364 overlong encoded strings. Why am I not surprised?
3366 PH/15 If the first argument of "${if match_address" was not empty, but did not
3367 contain an "@" character, Exim crashed. Now it writes a panic log message
3368 and treats the condition as false.
3370 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3372 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3373 extension "envelope-auth". The code is finished and in agreement with
3374 other implementations, but there is no documentation so far and in fact,
3375 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3376 not changing the active code.
3378 Print executed "if" and "elsif" statements when debugging is used. This
3379 helps a great deal to understand what a filter does.
3381 Document more things not specified clearly in RFC3028. I had all this
3382 sorted out, when out of a sudden new issues came to my mind. Oops."
3384 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3387 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3388 canonical form (as documented). However, after a host name lookup from
3389 the IP address, check_host() was doing a simple string comparison with
3390 addresses acquired from the DNS when checking that the found name did
3391 have the original IP as one of its addresses. Since any found IPv6
3392 addresses are likely to be in abbreviated form, the comparison could
3393 fail. Luckily, there already exists a function for doing the comparison
3394 by converting both addresses to binary, so now that is used instead of
3395 the text comparison.
3397 PH/20 There was another similar case to PH/19, when a complete host name was
3398 given in a host list; looking up its IP address could give an abbreviated
3399 form, whereas the current host's name might or might not be abbreviated.
3400 The same fix has been applied.
3406 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3407 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3410 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3411 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3413 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3415 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3416 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3417 Also, for "Auto-Submitted", treat anything other than "no" as
3418 non-personal, in accordance with RFC 3834. (Previously it treated
3419 anything starting "auto-" as non-personal.)
3421 TF/01 The control=submission/name=... option had a problem with syntax
3422 errors if the name included a slash character. The /name= option
3423 now slurps the rest of the string, so it can include any characters
3424 but it must come last in the list of options (after /sender_retain
3427 PH/05 Some modifications to the interface to the fake nameserver for the new
3435 TK/01 Added the "success_on_redirect" address verification option. See
3436 NewStuff for rationale and an example.
3438 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3440 PH/02 Patch to exigrep to allow it to work on syslog lines.
3442 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3443 fread() to skip over the body file's header line, because in Cygwin the
3444 header line is locked and is inaccessible.
3446 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3447 co-exist for some time) to make it clear that it is the Exim ID that is
3448 referenced, not the Message-ID: header line.
3450 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3451 string_format() function, because snprintf() does not exist on all
3454 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3455 ${stat: expansion item.
3457 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3458 protocol synchronization error", to keep the pedants happy.
3460 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3461 well as for IRIX systems, when gcc is being used. See the host.c source
3464 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3466 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3469 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3470 transport and to the smtp transport in LMTP mode.
3472 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3474 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3475 run a filter in a subprocess. This could lead to confusion in subsequent
3476 lookups in the parent process. There should also be a search_tidyup() at
3477 the end of the subprocess.
3479 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3480 only if the host matched helo_try_verify_hosts, which caused the
3481 verification to occur when the EHLO/HELO command was issued. The ACL just
3482 tested the remembered result. Now, if a previous verification attempt has
3483 not happened, "verify = helo" does it there and then.
3485 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3487 TK/03 Fix log output including CR from clamd.
3489 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3490 exist provoked a memory error which could cause a segfault.
3492 PH/15 Installed PCRE 6.2
3494 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3496 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3497 of the problem. Specifically, suggested +O2 rather than +O1 for the
3500 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3502 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3503 record, the port was not correctly logged when the outgoing_port log
3504 selector was set (it logged the transort's default port).
3506 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3507 fallback_hosts, and "hosts" in the smtp transport.
3509 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3510 host errors such as "Connection refused".
3512 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3513 authentication with radiusclient 0.4.9:
3515 - Error returned from rc_read_config was caught wrongly
3516 - Username/password not passed on to radius server due to wrong length.
3518 The presumption is that some radiusclient API changes for 4.51/PH/17
3519 were not taken care of correctly. The code is still untested by me (my
3520 Linux distribution still has 0.3.2 of radiusclient), but it was
3521 contributed by a Radius user.
3523 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3524 expanding the "port" option of the smtp transport.
3526 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3527 while reading a MIME header. Thanks to Tom Hughes for a patch.
3529 PH/24 Include config.h inside local_scan.h so that configuration settings are
3532 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3533 that in an ACL for RCPT (for example), you can examine exactly what was
3536 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3537 commands, but it was not correctly comparing the address with the actual
3538 client host address. Thus, it would show the EHLO address in Received:
3539 header lines when this was not necessary.
3541 PH/27 Added the % operator to ${eval:}.
3543 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3544 it should be ignoring failures (because with -C, for example, it has lost
3545 privilege). It wasn't ignoring creation failures other than "already
3548 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3551 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3552 because an input line was too long, either on its own, or by virtue of
3553 too many continuations, the temporary file was not being removed, and the
3554 return code was incorrect.
3556 PH/31 Missing "BOOL" in function definition in filtertest.c.
3558 PH/32 Applied Sieve patches from the maintainer.
3560 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3562 PH/33 Added "verify = not_blind".
3564 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3565 Local/Makefile (with some defaults set). These are used in built scripts
3566 such as exicyclog, but they have never been used in the exim_install
3567 script (though there are many overriding facilities there). I have
3568 arranged that the exim_install script now takes note of these two
3571 PH/35 Installed configuration files for Dragonfly.
3573 PH/36 When a locally submitted message by a trusted user did not contain a
3574 From: header, and the sender address was obtained from -f or from an SMTP
3575 MAIL command, and the trusted user did not use -F to supply a sender
3576 name, $originator_name was incorrectly used when constructing a From:
3577 header. Furthermore, $originator_name was used for submission mode
3578 messages from external hosts without From: headers in a similar way,
3579 which is clearly wrong.
3581 PH/37 Added control=suppress_local_fixups.
3583 PH/38 When log_selector = +received_sender was set, and the addition of the
3584 sender made the log line's construction buffer exactly full, or one byte
3585 less than full, an overflow happened when the terminating "\n" was
3588 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3589 when the result of a list match is failure because a DNS lookup failed.
3591 PH/40 RM_COMMAND is now used in the building process.
3593 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3594 the "build-* directories that it finds.
3596 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3597 address in a domain literal was a prefix of an interface address.
3599 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3600 when verifying a sender address, unless rewrite_headers is false.
3602 PH/44 Wrote a long comment about why errors_to addresses are verified as
3603 recipients, not senders.
3605 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3606 the ratelimit ACL was added.
3608 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3610 PH/46 Added extra information about PostgreSQL errors to the error string.
3612 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3613 avoiding the need to install special zones in a real server. This is
3614 backwards compatible; if it can't find the fake resolver, it drops back.
3615 Thus, both old and new test suites can be run.
3617 TF/02 Added util/ratelimit.pl
3619 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3620 clock is set back in time.
3622 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3623 Candler <B.Candler@pobox.com>.
3625 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3626 OK for addresses that are the result of redirections.
3628 PH/48 A number of further additions for the benefit of the new test suite,
3629 including a fake gethostbyname() that interfaces to the fake DNS resolver
3632 TF/06 The fix for widen_domains has also been applied to qualify_single and
3633 search_parents which are the other dnslookup options that can cause
3636 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3639 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3641 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3642 service" if it encountered an MX record with an empty target hostname.
3643 The message is now "an MX or SRV record indicated no SMTP service".
3645 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3646 if the DNS of the sending site is misconfigured. This is quite a
3647 common situation. This change restores the behaviour of treating a
3648 helo verification defer as a failure.
3650 PH/51 If self=fail was set on a router, the bounce message did not include the
3651 actual error message.
3657 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3659 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3660 command itself ended in error, the underlying message about the transport
3661 filter timeout was being overwritten with the pipe command error. Now the
3662 underlying error message should be appended to the second error message.
3664 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3666 PH/02 Reduce the amount of output that "make" produces by default. Full output
3667 can still be requested.
3669 PH/03 The warning log line about a condition test deferring for a "warn" verb
3670 was being output only once per connection, rather than after each
3671 occurrence (because it was using the same function as for successful
3672 "warn" verbs). This seems wrong, so I have changed it.
3674 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3675 it should not have, which might have caused a crash in the right
3676 circumstances, but probably never did.
3678 PH/04 Installed a modified version of Tony Finch's patch to make submission
3679 mode fix the return path as well as the Sender: header line, and to
3680 add a /name= option so that you can make the user's friendly name appear
3683 TF/03 Added the control = fakedefer ACL modifier.
3685 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3686 Mark Lowes for thorough testing.
3688 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3690 TK/03 Merged latest SRS patch from Miles Wilton.
3692 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3693 with the definition in sysexits.h (which is #included earlier).
3694 Fortunately, Exim does not actually use EX_OK. The code used to try to
3695 preserve the sysexits.h value, by assuming that macro definitions were
3696 scanned for macro replacements. I have been disabused of this notion,
3697 so now the code just undefines EX_OK before #including unistd.h.
3699 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3700 in the smtp transport. When a block could not be written in a single
3701 write() function, the timeout was being re-applied to each part-write.
3702 This seems wrong - if the receiver was accepting one byte at a time it
3703 would take for ever. The timeout is now adjusted when this happens. It
3704 doesn't have to be particularly precise.
3706 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3707 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3709 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3710 without a domain if the check to <postmaster@domain> fails.
3712 SC/01 Eximstats: added -xls and the ability to specify output files
3713 (patch written by Frank Heydlauf).
3715 SC/02 Eximstats: use FileHandles for outputting results.
3717 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3719 SC/04 Eximstats: fixed display of large numbers with -nvr option
3721 SC/05 Eximstats: fixed merging of reports with empty tables.
3723 SC/06 Eximstats: added the -include_original_destination flag
3725 SC/07 Eximstats: removed tabs and trailing whitespace.
3727 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3729 TK/06 MBOX spool code: Add real "From " MBOX separator line
3730 so the .eml file is really in mbox format (even though
3731 most programs do not really care). Patch from Alex Miller.
3733 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3734 The latter is generated from $received_to and is only set if the
3735 message has one envelope recipient. SA can use these headers,
3736 obviously out-of-the-box. Patch from Alex Miller.
3738 PH/08 The ${def test on a variable was returning false if the variable's
3739 value was "0", contrary to what the specification has always said!
3740 The result should be true unless the variable is empty.
3742 PH/09 The syntax error of a character other than { following "${if
3743 def:variable_name" (after optional whitespace) was not being diagnosed.
3744 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3745 accidental colon was present, for example, could give incorrect results.
3747 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3748 result is used (not including appendfile, where other changes are about
3751 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3752 This involved changing a lot of size variables from int to off_t. It
3753 should work with maildirs and everything.
3755 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3756 spamd dying while we are connected to it.
3758 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3761 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3762 function for BDB 4.3.
3764 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3766 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3767 the sources for PCRE have been re-arranged and more files are now
3770 PH/15 The code I had for printing potentially long long variables in PH/11
3771 above was not the best (it lost precision). The length of off_t variables
3772 is now inspected at build time, and an appropriate printing format (%ld
3773 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3774 to be "long long int" or "long int". This is needed for the internal
3775 formatting function string_vformat().
3777 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3778 the configuration file to be ":syslog", then the script "guesses" where
3779 the logs files are, rather than using the compiled in default. In our
3780 case the guess is not the same as the compiled default, so the script
3781 suddenly stopped working when I started to use syslog. The patch checks
3782 to see if log_file_path is "". If so, it attempts to read it from exim
3783 with no configuration file to get the compiled in version, before it
3784 falls back to the previous guessing code."
3786 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3787 implementing BATV in an Exim configuration. See NewStuff for the gory
3790 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3791 Makefile that are specific to HP-UX.
3793 PH/18 If the "use_postmaster" option was set for a recipient callout together
3794 with the "random" option, the postmaster address was used as the MAIL
3795 FROM address for the random test, but not for the subsequent recipient
3796 test. It is now used for both.
3798 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3799 patch removes a few documentation additions to RFC 3028, because the
3800 latest draft now contains them. It adds the new en;ascii-case comparator
3801 and a new error check for 8bit text in MIME parts. Comparator and
3802 require names are now matched exactly. I enabled the subaddress
3803 extension, but it is not well tested yet (read: it works for me)."
3805 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3806 rework some of the code of TK/09 above to avoid the hardwired use of
3807 "%lld" and "long long". Replaced the call to snprintf() with a call to
3810 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3811 records point to non-existent hosts", "retry timeout exceeded", and
3812 "retry time not reached for any host after a long failure period".
3814 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3815 experimental DomainKeys support:
3817 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3818 (2) On an error such as an illegally used "control", the wrong name for
3819 the control was given.
3821 These problems did NOT occur unless DomainKeys support was compiled.
3823 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3825 PH/24 Added ${if match_ip condition.
3827 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3828 databases so that it will be absolutely obvious if a crash occurs in the
3829 DB library. This is a regular occurrence (often caused by mis-matched
3832 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3833 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3834 picked up on a user's system that detects such things. There doesn't seem
3835 to be a gcc warning option for this - only an attribute that has to be
3836 put on the function's prototype. It seems that in Fedora Core 4 they have
3837 set this on a number of new functions. No doubt there will be more in due
3840 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3841 specify a transport. However, if an address that was verified by such a
3842 router was the subject of a callout, Exim crashed because it tried to
3843 read the rcpt_include_affixes from the non-existent transport. Now it
3844 just assumes that the setting of that option is false. This bug was
3845 introduced by 4.51/PH/31.
3847 PH/28 Changed -d+all to exclude +memory, because that information is very
3848 rarely of interest, but it makes the output a lot bigger. People tend to
3849 do -d+all out of habit.
3851 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3852 code in os-type was giving problems when libc.so lives in lib64, like on
3855 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3856 aren't the modern standard, and it seems that some systems' include files
3857 don't always have them. Exim was already checking for some of the newer
3858 ones like T_AAAA, and defining it itself. I've added checks for all the
3859 record types that Exim uses.
3861 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3862 not automatically generating a new one, as it is supposed to. This
3863 prevented TLS from working. If the file did exist, but contained invalid
3864 data, a new version was generated, as expected. It was only the case of a
3865 non-existent file that was broken.
3867 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3868 with a change in libdomainkeys > 0.64.
3870 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3871 from DNS. If the selector record carries the flag, it now has
3872 precedence over the domain-wide flag.
3874 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3876 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3877 the use of a port name that isn't defined in /etc/services) occurred, the
3878 message was deferred as in a normal delivery, and thus remained on the
3879 spool, instead of being failed because of the mua_wrapper setting. This
3880 is now fixed, and I tidied up some of the mua_wrapper messages at the
3883 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3884 the messages in a hash of arrays rather than using individual hashes.
3885 This is a bit cleaner and results in dramatic memory savings, albeit
3886 at a slight CPU cost.
3888 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3889 as requested by Marc Sherman.
3891 SC/10 Eximstats: added histograms for user specified patterns as requested
3894 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3896 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3897 fopen() in the content-scanning modules that did not already have it.
3903 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3904 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3906 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3908 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3910 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3911 to test Sieve filters that use "vacation".
3913 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3914 that changes the way the GnuTLS parameters are stored in the cache file.
3915 The new format can be generated externally. For backward compatibility,
3916 if the data in the cache doesn't make sense, Exim assumes it has read an
3917 old-format file, and it generates new data and writes a new file. This
3918 means that you can't go back to an older release without removing the
3921 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3922 work if there are any delivery delays because "one_time" forces the
3923 parent to be marked "delivered", so its unseen clone is never tried
3924 again. For this reason, Exim now forbids the simultaneous setting of
3927 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3928 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3929 entry if you want to know the details. The fix, however, neglected to
3930 consider the case when local delivery batching is involved. The test for
3931 "previously delivered" was not happening when checking to see if an
3932 address could be batched with a previous (undelivered) one; under
3933 certain circumstances this could lead to multiple deliveries to the same
3936 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3937 in its include files, and this causes problems building Exim.
3939 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3940 header_syntax) but Exim was just ignoring anything given after a slash.
3941 In particular, this caused confusion with an attempt to use "verify =
3942 reverse_host_lookup/defer_ok". An error is now given when options are
3943 supplied for verify items that do not have them. (Maybe reverse_host_
3944 lookup should have a defer_ok option, but that's a different point.)
3946 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3947 defined by RFC 821) to 2048, because there were problems with some AUTH
3948 commands, and RFC 1869 says the size should be increased for extended
3949 SMTP commands that take arguments.
3951 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3954 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3955 "unknown" error; now it says that the functionality isn't in the binary.
3957 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3958 an address' error message when a string expansion fails (syntax or
3959 whatever). Otherwise the password may appear in the log. Following change
3960 PH/42 below, there is no longer a chance of it appearing in a bounce
3963 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3965 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3966 of its addresses was ever tried. (Bugzilla bug #2.)
3968 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3969 the result incorrectly in the debug output. (It correctly added a newline
3970 to what was transported.)
3972 TF/01 Added $received_time.
3974 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3975 commented out examples of how to interface to a virus scanner and to
3976 SpamAssassin. Also added commented examples of av_scanner and
3977 spamd_address settings.
3979 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3980 and controls are allowed in which ACLs. There were a couple of minor
3981 errors. Some of the entries in the conditions table (which is a table of
3982 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3983 negation of where the condition IS allowed.
3985 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3987 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3988 header file does not have a version number, so I've had to invent a new
3989 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3990 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3991 radiusclient), but it was contributed by a Radius user.
3993 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3994 files or directories, for OpenSSL.
3996 PH/19 When an Exim process that is running as root has to create an Exim log
3997 file, it does so in a subprocess that runs as exim:exim so as to get the
3998 ownership right at creation (otherwise, other Exim processes might see
3999 the file with the wrong ownership). There was no test for failure of this
4000 fork() call, which would lead to the process getting stuck as it waited
4001 for a non-existent subprocess. Forks do occasionally fail when resources
4002 run out. I reviewed all the other calls to fork(); they all seem to check
4005 PH/20 When checking for unexpected SMTP input at connect time (before writing
4006 the banner), Exim was not dealing correctly with a non-positive return
4007 from the read() function. If the client had disconnected by this time,
4008 the result was a log entry for a synchronization error with an empty
4009 string after "input=" when read() returned zero. If read() returned -1
4010 (an event I could not check), uninitialized data bytes were printed.
4011 There were reports of junk text (parts of files, etc) appearing after
4014 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4016 PH/22 Added support for macro redefinition, and (re)definition in between
4017 driver and ACL definitions.
4019 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4020 forgetting to use the resulting value; it was using the unexpanded value.
4022 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4023 hadn't been configured. The fix is from Juergen Kreileder, who
4024 understands it better than I do:
4026 "Here's what I see happening with three configured cyrus_sasl
4027 authenticators configured (plain, login, cram-md5):
4029 On startup auth_cyrus_sasl_init() gets called for each of these.
4030 This means three calls to sasl_listmech() without a specified mech_list.
4031 => SASL tests which mechs of all available mechs actually work
4032 => three warnings about OTP not working
4033 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4035 With the patch, sasl_listmech() also gets called three times. But now
4036 SASL's mech_list option is set to the server_mech specified in the the
4037 authenticator. Or in other words, the answer from sasl_listmech()
4038 gets limited to just the mech you're testing for (which is different
4040 => the return list contains just 'plain' or 'login', 'cram-md5' or
4041 nothing depending on the value of ob->server_mech.
4043 I've just tested the patch: Authentication still works fine,
4044 unavailable mechs specified in the exim configuration are still
4045 caught, and the auth.log warnings about OTP are gone."
4047 PH/25 When debugging is enabled, the contents of the command line are added
4048 to the debugging output, even when log_selector=+arguments is not
4051 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4052 answer is "GNU", and only if the return is "GNU/something" is the answer
4055 PH/27 $acl_verify_message is now set immediately after the failure of a
4056 verification in an ACL, and so is available in subsequent modifiers. In
4057 particular, the message can be preserved by coding like this:
4059 warn !verify = sender
4060 set acl_m0 = $acl_verify_message
4062 Previously, $acl_verify_message was set only while expanding "message"
4063 and "log_message" when a very denied access.
4065 PH/28 Modified OS/os.c-Linux with
4067 -#ifndef OS_LOAD_AVERAGE
4068 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4070 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4071 nomenclature these days.)
4073 PH/29 Installed patch from the Sieve maintainer that adds the options
4074 sieve_useraddress and sieve_subaddress to the redirect router.
4076 PH/30 In these circumstances:
4077 . Two addresses routed to the same list of hosts;
4078 . First host does not offer TLS;
4079 . First host accepts first address;
4080 . First host gives temporary error to second address;
4081 . Second host offers TLS and a TLS session is established;
4082 . Second host accepts second address.
4083 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4084 and peerdn, if requested) that were in fact used only for the second
4087 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4088 attention to any local part prefix or suffix that was matched by the
4089 router that accepted the address. It now behaves in the same way as it
4090 does for delivery: the affixes are removed from the local part unless
4091 rcpt_include_affixes is set on the transport.
4093 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4094 timeout during the DATA phase of an incoming message.
4096 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4097 applied a patch sanctioned by the Sieve maintainer.
4099 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4100 the uid or gid is negative. A case of a negative gid caused this to be
4101 noticed. The fix allows for either to be negative.
4103 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4104 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4107 PH/36 The change PH/12 above was broken. Fixed it.
4109 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4110 the grounds that routing the same address twice would always produce the
4111 same answer. This might have been true once, but it is certainly no
4112 longer true now. Routing a child address may depend on the previous
4113 routing that produced that child. Some complicated redirection strategies
4114 went wrong when messages had multiple recipients, and made Exim's
4115 behaviour dependent on the order in which the addresses were given.
4117 I have moved the duplicate checking until after the routing is complete.
4118 Exim scans the addresses that are assigned to local and remote
4119 transports, and removes any duplicates. This means that more work will be
4120 done, as duplicates will always all be routed, but duplicates are
4121 presumably rare, so I don't expect this is of any significance.
4123 For deliveries to pipes, files, and autoreplies, the duplicate checking
4124 still happens during the routing process, since they are not going to be
4127 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4128 It corrects a timeout issue with spamd. This is Ian's comment: "The
4129 background is that sometimes spamd either never reads data from a
4130 connection it has accepted, or it never writes response data. The exiscan
4131 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4132 blindly assumes that writes won't block so it may never time out."
4134 PH/39 Allow G after quota size as well as K and M.
4136 PH/40 The value set for $authenticated_id in an authenticator may not contain
4137 binary zeroes or newlines because the value is written to log lines and
4138 to spool files. There was no check on this. Now the value is run through
4139 the string_printing() function so that such characters are converted to
4140 printable escape sequences.
4142 PH/41 $message_linecount is a new variable that contains the total number of
4143 lines in the message. Compare $body_linecount, which is the count for the
4146 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4147 bounce and delay warning messages, except in certain special cases, which
4150 (a) An SMTP error message from a remote host;
4151 (b) A message specified in a :fail: redirection;
4152 (c) A message specified in a "fail" command in a system filter;
4153 (d) A message specified in a FAIL return from the queryprogram router;
4154 (e) A message specified by the cannot_route_message router option.
4156 In these cases only, Exim does include the error details in bounce and
4157 warning messages. There are also a few cases where bland messages such
4158 as "unrouteable address" or "local delivery error" are given.
4160 PH/43 $value is now also set for the "else" part of a ${run expansion.
4162 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4163 being worked on, but at least Exim now implements the latest version to
4166 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4167 process to complete was treated as a delivery failure, a timeout while
4168 writing the message to the pipe was logged, but erroneously treated as a
4169 successful delivery. Such timeouts include transport filter timeouts. For
4170 consistency with the overall process timeout, these timeouts are now
4171 treated as errors, giving rise to delivery failures by default. However,
4172 there is now a new Boolean option for the pipe transport called
4173 timeout_defer, which, if set TRUE, converts the failures into defers for
4174 both kinds of timeout. A transport filter timeout is now identified in
4177 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4178 systems where "make" and "gmake" are different, calling "gmake" at top
4179 level broke things. I've arranged for the value of $(MAKE) to be passed
4180 from the Makefile to this script so that it can call the same version of
4184 A note about Exim versions 4.44 and 4.50
4185 ----------------------------------------
4187 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4188 changes of various kinds. As a consequence, a big documentation update was
4189 needed. This delayed the release for rather longer than seemed good, especially
4190 in the light of a couple of (minor) security issues. Therefore, the changes
4191 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4192 release. So 4.44 and 4.50 are in effect two different branches that both start
4195 I have left the 4.50 change log unchanged; it contains all the changes since
4196 4.43. The change log for 4.44 is below; many of its items are identical to
4197 those for 4.50. This seems to be the most sensible way to preserve the
4198 historical information.
4204 1. Minor wording change to the doc/README.SIEVE file.
4206 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4207 computation of the current number of files was incorrect.
4209 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4210 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4213 4. Give more explanation in the error message when the command for a transport
4214 filter fails to execute.
4216 5. There are several places where Exim runs a non-Exim command in a
4217 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4218 was being done only for the command run by the queryprogram router. It is
4219 now done for all such subprocesses. The other cases are: ${run, transport
4220 filters, and the commands run by the lmtp and pipe transports.
4222 6. Added CONFIGURE_GROUP build-time option.
4224 7. Some older OS have a limit of 256 on the maximum number of file
4225 descriptors. Exim was using setrlimit() to set 1000 as a large value
4226 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4227 systems. I've change it so that if it can't get 1000, it tries for 256.
4229 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4230 was an oversight, and furthermore, ever since the addition of extra
4231 controls (e.g. 4.43/32), the checks on when to allow different forms of
4232 "control" were broken. There should now be diagnostics for all cases when a
4233 control that does not make sense is encountered.
4235 9. Added the /retain_sender option to "control=submission".
4237 10. $recipients is now available in the predata ACL (oversight).
4239 11. Tidy the search cache before the fork to do a delivery from a message
4240 received from the command line. Otherwise the child will trigger a lookup
4241 failure and thereby defer the delivery if it tries to use (for example) a
4242 cached ldap connection that the parent has called unbind on.
4244 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4245 of $address_data from the recipient verification was clobbered by the
4246 sender verification.
4248 13. The value of address_data from a sender verification is now available in
4249 $sender_address_data in subsequent conditions in the ACL statement.
4251 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4253 15. Added a new option "connect=<time>" to callout options, to set a different
4256 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4257 was its contents. (It was OK if the option was not defined at all.)
4259 17. A "Completed" log line is now written for messages that are removed from
4260 the spool by the -Mrm option.
4262 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4263 information about exactly what failed.
4265 19. Added -dd to debug only the daemon process.
4267 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4268 handles timeouts, both on the server side and network timeouts. Renamed the
4269 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4271 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4272 It is now set to "smtps".
4274 22. $host_address is now set to the target address during the checking of
4275 ignore_target_hosts.
4277 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4278 being passed; this would have caused $sender_host_name to have been used if
4279 matching the list had actually called for a host name (not very likely,
4280 since this list is usually IP addresses). A host name is now passed as
4283 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4284 code that specifies a non-restarting handler (typically sigaction() in
4285 modern systems) in an attempt to fix a rare and obscure crash bug.
4287 25. Narrowed the window for a race in the daemon that could cause it to ignore
4288 SIGCHLD signals. This is not a major problem, because they are used only to
4289 wake it up if nothing else does.
4291 26. A malformed maildirsize file could cause Exim to calculate negative values
4292 for the mailbox size or file count. Odd effects could occur as a result.
4293 The maildirsize information is now recalculated if the size or filecount
4296 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4297 support for a long time. Removed HAVE_SYS_VFS_H.
4299 28. Installed the latest version of exipick from John Jetmore.
4301 29. In an address list, if the pattern was not a regular expression, an empty
4302 subject address (from a bounce message) matched only if the pattern was an
4303 empty string. Non-empty patterns were not even tested. This was the wrong
4304 because it is perfectly reasonable to use an empty address as part of a
4305 database query. An empty address is now tested by patterns that are
4306 lookups. However, all the other forms of pattern expect the subject to
4307 contain a local part and a domain, and therefore, for them, an empty
4308 address still always fails if the pattern is not itself empty.
4310 30. Exim went into a mad DNS loop when attempting to do a callout where the
4311 host was specified on an smtp transport, and looking it up yielded more
4312 than one IP address.
4314 31. Re-factored the code for checking spool and log partition space into a
4315 function that finds that data and another that does the check. The former
4316 is then used to implement four new variables: $spool_space, $log_space,
4317 $spool_inodes, and $log_inodes.
4319 32. The RFC2047 encoding function was originally intended for short strings
4320 such as real names; it was not keeping to the 75-character limit for
4321 encoded words that the RFC imposes. It now respects the limit, and
4322 generates multiple encoded words if necessary. To be on the safe side, I
4323 have increased the buffer size for the ${rfc2047: expansion operator from
4326 33. It is now permitted to omit both strings after an "if" condition; if the
4327 condition is true, the result is "true". As before, when the second string
4328 is omitted, a false condition yields an empty string. This makes it less
4329 cumbersome to write custom ACL and router conditions.
4331 34. Failure to deliver a bounce message always caused it to be frozen, even if
4332 there was an errors_to setting on the router. The errors_to setting is now
4335 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4336 canonical form (fully expanded) before being placed in
4337 $sender_host_address.
4339 36. The table in the code that translates DNS record types into text (T_A to
4340 "A" for instance) was missing entries for NS and CNAME. It is just possible
4341 that this could have caused confusion if both these types were looked up
4342 for the same domain, because the text type is used as part of Exim's
4343 per-process caching. But the chance of anyone hitting this buglet seems
4346 37. The dnsdb lookup has been extended in a number of ways.
4348 (1) There is a new type, "zns", which walks up the domain tree until it
4349 finds some nameserver records. It should be used with care.
4351 (2) There is a new type, "mxh", which is like "mx" except that it returns
4352 just the host names, not the priorities.
4354 (3) It is now possible to give a list of domains (or IP addresses) to be
4355 looked up. The behaviour when one of the lookups defers can be
4356 controlled by a keyword.
4358 (4) It is now possible to specify the separator character for use when
4359 multiple records are returned.
4361 38. The dnslists ACL condition has been extended: it is now possible to supply
4362 a list of IP addresses and/or domains to be looked up in a particular DNS
4365 39. Added log_selector=+queue_time_overall.
4367 40. When running the queue in the test harness, wait just a tad after forking a
4368 delivery process, to get repeatability of debugging output.
4370 41. Include certificate and key file names in error message when GnuTLS fails
4371 to set them up, because the GnuTLS error message doesn't include the name
4372 of the failing file when there is a problem reading it.
4374 42. Allow both -bf and -bF in the same test run.
4376 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4378 44. The "Exiscan patch" is now merged into the mainline Exim source.
4380 45. Sometimes the final signoff response after QUIT could fail to get
4381 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4382 < 0 before doing a fflush(). This bug looks as though it goes back to the
4383 introduction of TLS in release 3.20, but "sometimes" must have been rare
4384 because the tests only now provoked it.
4386 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4387 (this can affect the format of dates).
4389 47. exim_tidydb, when checking for the continued existence of a message for
4390 which it has found a message-specific retry record, was not finding
4391 messages that were in split spool directories. Consequently, it was
4392 deleting retry records that should have stayed in existence.
4394 48. Steve fixed some bugs in eximstats.
4396 49. The SPA authentication driver was not abandoning authentication and moving
4397 on to the next authenticator when an expansion was forced to fail,
4398 contradicting the general specification for all authenticators. Instead it
4399 was generating a temporary error. It now behaves as specified.
4401 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4402 (the order specifies the preference for clients). The order is now AES256,
4403 AES128, 3DES, ARCFOUR128.
4405 51. Small patch to Sieve code - explicitly set From: when generating an
4408 52. Exim crashed if a remote delivery caused a very long error message to be
4409 recorded - for instance if somebody sent an entire SpamAssassin report back
4410 as a large number of 550 error lines. This bug was coincidentally fixed by
4411 increasing the size of one of Exim's internal buffers (big_buffer) that
4412 happened as part of the Exiscan merge. However, to be on the safe side, I
4413 have made the code more robust (and fixed the comments that describe what
4416 53. Now that there can be additional text after "Completed" in log lines (if
4417 the queue_time_overall log selector is set), a one-byte patch to exigrep
4418 was needed to allow it to recognize "Completed" as not the last thing in
4421 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4422 patch that reportedly fixes this has been added. I am not expert enough to
4423 create a test for it. This is what the patch creator wrote:
4425 "I found a little strange behaviour of ldap code when working with
4426 Windows 2003 AD Domain, where users was placed in more than one
4427 Organization Units. When I tried to give exim partial DN, the exit code
4428 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4429 But simultaneously result of request was absolutely normal ldap result,
4430 so I produce this patch..."
4432 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4433 REFERENCE, so I have modified the code to exclude the patch when that macro
4436 55. Some experimental protocols are using DNS PTR records for new purposes. The
4437 keys for these records are domain names, not reversed IP addresses. The
4438 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4439 leaves it alone. Component reversal etc. now happens only for IP addresses.
4442 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4444 57. Double the size of the debug message buffer (to 2048) so that more of very
4445 long debug lines gets shown.
4447 58. The exicyclog utility now does better if the number of log files to keep
4448 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4450 59. Two changes related to the smtp_active_hostname option:
4452 (1) $smtp_active_hostname is now available as a variable.
4453 (2) The default for smtp_banner uses $smtp_active_hostname instead
4454 of $primary_hostname.
4456 60. The host_aton() function is supposed to be passed a string that is known
4457 to be a valid IP address. However, in the case of IPv6 addresses, it was
4458 not checking this. This is a hostage to fortune. Exim now panics and dies
4459 if the condition is not met. A case was found where this could be provoked
4460 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4461 components; fortuitously, this particular loophole had already been fixed
4462 by change 4.50/55 above.
4464 If there are any other similar loopholes, the new check in host_aton()
4465 itself should stop them being exploited. The report I received stated that
4466 data on the command line could provoke the exploit when Exim was running as
4467 exim, but did not say which command line option was involved. All I could
4468 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4469 running as the user.
4472 61. There was a buffer overflow vulnerability in the SPA authentication code
4473 (which came originally from the Samba project). I have added a test to the
4474 spa_base64_to_bits() function which I hope fixes it.
4477 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4478 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4480 63. The daemon start-up calls getloadavg() while still root for those OS that
4481 need the first call to be done as root, but it missed one case: when
4482 deliver_queue_load_max is set with deliver_drop_privilege. This is
4483 necessary for the benefit of the queue runner, because there is no re-exec
4484 when deliver_drop_privilege is set.
4486 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4487 This has been fixed.
4489 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4490 was in use, was not putting the data itself into the right store pool;
4491 consequently, it could be overwritten for a subsequent message in the same
4492 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4495 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4497 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4498 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4499 the function were treating the return as a boolean value, which happened to
4500 work because 0=false and not-0=true, but is not correct code.
4502 68. The host_aton() function was not handling scoped IPv6 addresses (those
4503 with, for example, "%eth0" on the end) correctly.
4505 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4506 negated items (that is, ~something) in unsigned ints. Some compilers
4507 apparently mutter when there is no cast.
4509 70. If an address verification called from an ACL failed, and did not produce a
4510 user-specific message (i.e. there was only a "system" message), nothing was
4511 put in $acl_verify_message. In this situation, it now puts the system
4514 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4515 session; change 4.31/43 added the unwanted input to the log line - except
4516 that it did not do this in the start of session case. It now does.
4518 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4519 This is wrong and can cause the other end to generate a synchronization
4520 error if it is another Exim or anything else that does the synchronization
4521 check. A QUIT command is no longer sent after a timeout.
4523 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4524 during host lookups.
4526 74. The defer_ok option of callout verification was not working if it was used
4527 when verifying addresses in header lines, that is, for this case:
4529 verify = header_sender/callout=defer_ok
4531 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4532 those file descriptors could be used for SMTP connections. If anything
4533 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4534 could be sent to the SMTP client, causing chaos. The daemon now opens
4535 stdin, stdout, and stderr to /dev/null when it puts itself into the
4538 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4539 log by default. The user can override this with suitable Perl magic.
4541 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4542 the log message when discard triggers, was not working for the DATA ACL or
4543 for the non-SMTP ACL.
4545 78. Error message wording change in sieve.c.
4547 79. If smtp_accept_max_per_host was set, the number of connections could be
4548 restricted to fewer than expected, because the daemon was trying to set up
4549 a new connection before checking whether the processes handling previous
4550 connections had finished. The check for completed processes is now done
4551 earlier. On busy systems, this bug wouldn't be noticed because something
4552 else would have woken the daemon, and it would have reaped the completed
4555 80. If a message was submitted locally by a user whose login name contained one
4556 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4557 It caused a spool format error. I have fixed the spool reading code. A
4558 related problem was that the "from" clause in the Received: line became
4559 illegal because of the space(s). It is now covered by ${quote_local_part.
4561 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4564 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4566 83. Patch from Sieve maintainer to fix unterminated string problem in
4567 "vacation" handling.
4569 84. Some minor changes to the Linux configuration files to help with other
4570 OS variants using glibc.
4572 85. One more patch for Sieve to update vacation handling to latest spec.
4575 ----------------------------------------------------
4576 See the note above about the 4.44 and 4.50 releases.
4577 ----------------------------------------------------
4583 1. Change 4.43/35 introduced a bug that caused file counts to be
4584 incorrectly computed when quota_filecount was set in an appendfile
4587 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4588 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4591 3. Give more explanation in the error message when the command for a transport
4592 filter fails to execute.
4594 4. There are several places where Exim runs a non-Exim command in a
4595 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4596 was being done only for the command run by the queryprogram router. It is
4597 now done for all such subprocesses. The other cases are: ${run, transport
4598 filters, and the commands run by the lmtp and pipe transports.
4600 5. Some older OS have a limit of 256 on the maximum number of file
4601 descriptors. Exim was using setrlimit() to set 1000 as a large value
4602 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4603 systems. I've change it so that if it can't get 1000, it tries for 256.
4605 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4606 was an oversight, and furthermore, ever since the addition of extra
4607 controls (e.g. 4.43/32), the checks on when to allow different forms of
4608 "control" were broken. There should now be diagnostics for all cases when a
4609 control that does not make sense is encountered.
4611 7. $recipients is now available in the predata ACL (oversight).
4613 8. Tidy the search cache before the fork to do a delivery from a message
4614 received from the command line. Otherwise the child will trigger a lookup
4615 failure and thereby defer the delivery if it tries to use (for example) a
4616 cached ldap connection that the parent has called unbind on.
4618 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4619 of $address_data from the recipient verification was clobbered by the
4620 sender verification.
4622 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4623 was its contents. (It was OK if the option was not defined at all.)
4625 11. A "Completed" log line is now written for messages that are removed from
4626 the spool by the -Mrm option.
4628 12. $host_address is now set to the target address during the checking of
4629 ignore_target_hosts.
4631 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4632 being passed; this would have caused $sender_host_name to have been used if
4633 matching the list had actually called for a host name (not very likely,
4634 since this list is usually IP addresses). A host name is now passed as
4637 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4638 code that specifies a non-restarting handler (typically sigaction() in
4639 modern systems) in an attempt to fix a rare and obscure crash bug.
4641 15. Narrowed the window for a race in the daemon that could cause it to ignore
4642 SIGCHLD signals. This is not a major problem, because they are used only to
4643 wake it up if nothing else does.
4645 16. A malformed maildirsize file could cause Exim to calculate negative values
4646 for the mailbox size or file count. Odd effects could occur as a result.
4647 The maildirsize information is now recalculated if the size or filecount
4650 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4651 support for a long time. Removed HAVE_SYS_VFS_H.
4653 18. Updated exipick to current release from John Jetmore.
4655 19. Allow an empty sender to be matched against a lookup in an address list.
4656 Previously the only cases considered were a regular expression, or an
4659 20. Exim went into a mad DNS lookup loop when doing a callout where the
4660 host was specified on the transport, if the DNS lookup yielded more than
4663 21. The RFC2047 encoding function was originally intended for short strings
4664 such as real names; it was not keeping to the 75-character limit for
4665 encoded words that the RFC imposes. It now respects the limit, and
4666 generates multiple encoded words if necessary. To be on the safe side, I
4667 have increased the buffer size for the ${rfc2047: expansion operator from
4670 22. Failure to deliver a bounce message always caused it to be frozen, even if
4671 there was an errors_to setting on the router. The errors_to setting is now
4674 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4675 canonical form (fully expanded) before being placed in
4676 $sender_host_address.
4678 24. Updated eximstats to version 1.33
4680 25. Include certificate and key file names in error message when GnuTLS fails
4681 to set them up, because the GnuTLS error message doesn't include the name
4682 of the failing file when there is a problem reading it.
4684 26. Expand error message when OpenSSL has problems setting up cert/key files.
4687 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4688 (this can affect the format of dates).
4690 28. exim_tidydb, when checking for the continued existence of a message for
4691 which it has found a message-specific retry record, was not finding
4692 messages that were in split spool directories. Consequently, it was
4693 deleting retry records that should have stayed in existence.
4695 29. eximstats updated to version 1.35
4696 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4697 1.35 - bugfix such that pie charts by volume are generated correctly
4699 30. The SPA authentication driver was not abandoning authentication and moving
4700 on to the next authenticator when an expansion was forced to fail,
4701 contradicting the general specification for all authenticators. Instead it
4702 was generating a temporary error. It now behaves as specified.
4704 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4705 (the order specifies the preference for clients). The order is now AES256,
4706 AES128, 3DES, ARCFOUR128.
4708 31. Small patch to Sieve code - explicitly set From: when generating an
4711 32. Exim crashed if a remote delivery caused a very long error message to be
4712 recorded - for instance if somebody sent an entire SpamAssassin report back
4713 as a large number of 550 error lines. This bug was coincidentally fixed by
4714 increasing the size of one of Exim's internal buffers (big_buffer) that
4715 happened as part of the Exiscan merge. However, to be on the safe side, I
4716 have made the code more robust (and fixed the comments that describe what
4719 33. Some experimental protocols are using DNS PTR records for new purposes. The
4720 keys for these records are domain names, not reversed IP addresses. The
4721 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4722 leaves it alone. Component reversal etc. now happens only for IP addresses.
4725 34. The host_aton() function is supposed to be passed a string that is known
4726 to be a valid IP address. However, in the case of IPv6 addresses, it was
4727 not checking this. This is a hostage to fortune. Exim now panics and dies
4728 if the condition is not met. A case was found where this could be provoked
4729 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4730 components; fortuitously, this particular loophole had already been fixed
4731 by change 4.50/55 or 4.44/33 above.
4733 If there are any other similar loopholes, the new check in host_aton()
4734 itself should stop them being exploited. The report I received stated that
4735 data on the command line could provoke the exploit when Exim was running as
4736 exim, but did not say which command line option was involved. All I could
4737 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4738 running as the user.
4741 35. There was a buffer overflow vulnerability in the SPA authentication code
4742 (which came originally from the Samba project). I have added a test to the
4743 spa_base64_to_bits() function which I hope fixes it.
4746 36. The daemon start-up calls getloadavg() while still root for those OS that
4747 need the first call to be done as root, but it missed one case: when
4748 deliver_queue_load_max is set with deliver_drop_privilege. This is
4749 necessary for the benefit of the queue runner, because there is no re-exec
4750 when deliver_drop_privilege is set.
4752 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4753 was in use, was not putting the data itself into the right store pool;
4754 consequently, it could be overwritten for a subsequent message in the same
4755 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4758 38. Sometimes the final signoff response after QUIT could fail to get
4759 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4760 < 0 before doing a fflush(). This bug looks as though it goes back to the
4761 introduction of TLS in release 3.20, but "sometimes" must have been rare
4762 because the tests only now provoked it.
4768 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4769 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4770 Now it returns an integer. A number of calls were still expecting a T/F
4771 return. Fortuitously, in all cases, the tests worked in OK situations,
4772 which is the norm. However, things would have gone wrong on any write
4773 failures on the smtp file descriptor. This function is used when sending
4774 messages over SMTP and also when doing verify callouts.
4776 2. When Exim is called to do synchronous delivery of a locally submitted
4777 message (the -odf or -odi options), it no longer closes stderr before doing
4780 3. Implemented the mua_wrapper option.
4782 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4784 5. Implemented the functions header_remove(), header_testname(),
4785 header_add_at_position(), and receive_remove_recipient(), and exported them
4788 6. If an ACL "warn" statement specified the addition of headers, Exim already
4789 inserted X-ACL-Warn: at the start if there was no header name. However, it
4790 was not making this test for the second and subsequent header lines if
4791 there were newlines in the string. This meant that an invalid header could
4792 be inserted if Exim was badly configured.
4794 7. Allow an ACL "warn" statement to add header lines at the start or after all
4795 the Received: headers, as well as at the end.
4797 8. Added the rcpt_4xx retry error code.
4799 9. Added postmaster_mailfrom=xxx to callout verification option.
4801 10. Added mailfrom=xxxx to the callout verification option, for verify=
4804 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4805 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4807 12. Inserted some casts to stop certain compilers warning when using pointer
4808 differences as field lengths or precisions in printf-type calls (mostly
4809 affecting debugging statements).
4811 13. Added optional readline() support for -be (dynamically loaded).
4813 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4814 same clock tick as a message's arrival, so that its received time was the
4815 same as the "first fail" time on the retry record, and that message
4816 remained on the queue past the ultimate address timeout, every queue runner
4817 would try a delivery (because it was past the ultimate address timeout) but
4818 after another failure, the ultimate address timeout, which should have then
4819 bounced the address, did not kick in. This was a "< instead of <=" error;
4820 in most cases the first failure would have been in the next clock tick
4821 after the received time, and all would be well.
4823 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4824 being recognized when the domain list was tested by the match_domain
4825 condition in an expansion string.
4827 16. Added the ${str2b64: operator.
4829 17. Exim was always calling setrlimit() to set a large limit for the number of
4830 processes, without checking whether the existing limit was already
4831 adequate. (It did check for the limit on file descriptors.) Furthermore,
4832 errors from getrlimit() and setrlimit() were being ignored. Now they are
4833 logged to the main and panic logs, but Exim does carry on, to try to do its
4834 job under whatever limits there are.
4836 18. Imported PCRE 5.0.
4838 19. Trivial typo in log message " temporarily refused connection" (the leading
4841 20. If the log selector return_path_on_delivery was set and an address was
4842 redirected to /dev/null, the delivery process crashed because it assumed
4843 that a return path would always be set for a "successful" delivery. In this
4844 case, the whole delivery is bypassed as an optimization, and therefore no
4847 21. Internal re-arrangement: the function for sending a challenge and reading
4848 a response while authentication was assuming a zero-terminated challenge
4849 string. It's now changed to take a pointer and a length, to allow for
4850 binary data in such strings.
4852 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4854 23. Exim was not respecting finduser_retries when seeking the login of the
4855 uid under which it was called; it was always trying 10 times. (The default
4856 setting of finduser_retries is zero.) Also, it was sleeping after the final
4857 failure, which is pointless.
4859 24. Implemented tls_on_connect_ports.
4861 25. Implemented acl_smtp_predata.
4863 26. If the domain in control=submission is set empty, Exim assumes that the
4864 authenticated id is a complete email address when it generates From: or
4865 Sender: header lines.
4867 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4868 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4869 chown and chgrp in /bin and hostname in /usr/bin.
4871 28. Exim was keeping the "process log" file open after each use, just as it
4872 does for the main log. This opens the possibility of it remaining open for
4873 long periods when the USR1 signal hits a daemon. Occasional processlog
4874 errors were reported, that could have been caused by this. Anyway, it seems
4875 much more sensible not to leave this file open at all, so that is what now
4878 29. The long-running daemon process does not normally write to the log once it
4879 has entered its main loop, and it closes the log before doing so. This is
4880 so that log files can straightforwardly be renamed and moved. However,
4881 there are a couple of unusual error situations where the daemon does write
4882 log entries, and I had neglected to close the log afterwards.
4884 30. The text of an SMTP error response that was received during a remote
4885 delivery was being truncated at 512 bytes. This is too short for some of
4886 the long messages that one sometimes sees. I've increased the limit to
4889 31. It is now possible to make retry rules that apply only when a message has a
4890 specific sender, in particular, an empty sender.
4892 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4893 it possible to be selective about when SMTP synchronization is enforced.
4895 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4897 32. Implemented hosts_connection_nolog.
4899 33. Added an ACL for QUIT.
4901 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4904 35. Added mailbox_size and mailbox_filecount to appendfile.
4906 36. Added control = no_multiline_responses to ACLs.
4908 37. There was a bug in the logic of the code that waits for the clock to tick
4909 in the case where the clock went backwards by a substantial amount such
4910 that the microsecond fraction of "now" was more than the microsecond
4911 fraction of "then" (but the whole seconds number was less).
4913 38. Added support for the libradius Radius client library this is found on
4914 FreeBSD (previously only the radiusclient library was supported).
4920 1. When certain lookups returned multiple values in the form name=value, the
4921 quoting of the values was not always being done properly. Specifically:
4922 (a) If the value started with a double quote, but contained no whitespace,
4924 (b) If the value contained whitespace other than a space character (i.e.
4925 tabs or newlines or carriage returns) it was not quoted.
4926 This fix has been applied to the mysql and pgsql lookups by writing a
4927 separate quoting function and calling it from the lookup code. The fix
4928 should probably also be applied to nisplus, ibase and oracle lookups, but
4929 since I cannot test any of those, I have not disturbed their existing code.
4931 2. A hit in the callout cache for a specific address caused a log line with no
4932 reason for rejecting RCPT. Now it says "Previous (cached) callout
4933 verification failure".
4935 3. There was an off-by-one bug in the queryprogram router. An over-long
4936 return line was truncated at 256 instead of 255 characters, thereby
4937 overflowing its buffer with the terminating zero. As well as fixing this, I
4938 have increased the buffer size to 1024 (and made a note to document this).
4940 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4941 when Exim is waiting for an SMTP response from a remote server, Exim
4942 restarts its select() call on the socket, thereby resetting its timeout.
4943 This is not a problem when such interrupts are rare. Somebody set up a cron
4944 job to run exiwhat every 2 minutes, which is less than the normal select()
4945 timeout (5 or 10 minutes). This meant that the select() timeout never
4946 kicked in because it was always reset. I have fixed this by comparing the
4947 time when an interrupt arrives with the time at the start of the first call
4948 to select(). If more time than the timeout has elapsed, the interrupt is
4949 treated as a timeout.
4951 5. Some internal re-factoring in preparation for the addition of Sieve
4952 extensions (by MH). In particular, the "personal" test is moved to a
4953 separate function, and given an option for scanning Cc: and Bcc: (which is
4954 not set for Exim filters).
4956 6. When Exim created an email address using the login of the caller as the
4957 local part (e.g. when creating a From: or Sender: header line), it was not
4958 quoting the local part when it contained special characters such as @.
4960 7. Installed new OpenBSD configuration files.
4962 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4963 try to make them clearer.
4965 9. Callout options, other than the timeout value, were being ignored when
4966 verifying sender addresses in header lines. For example, when using
4968 verify = header_sender/callout=no_cache
4970 the cache was (incorrectly) being used.
4972 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4973 only the Cygwin environment.
4975 11. When return_path_on_delivery was set as a log selector, if different remote
4976 addresses in the same message used different return paths and parallel
4977 remote delivery occurred, the wrong values would sometimes be logged.
4978 (Whenever a remote delivery process finished, the return path value from
4979 the most recently started remote delivery process was logged.)
4981 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4982 lines when AUTH and/or TLS are in use. This is the "received protocol"
4983 field. Exim used to use "asmtp" for authenticated SMTP, without any
4984 indication (in the protocol name) for TLS use. Now it follows the RFC and
4985 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4986 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4987 names appear in log lines as well as in Received: header lines.
4989 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4990 extensions, and comparison tests, and to fix some bugs.
4992 14. Changes to the "personal" filter test:
4994 (1) The test was buggy in that it was just doing the equivalent of
4995 "contains" tests on header lines. For example, if a user's address was
4996 anne@some.where, the "personal" test would incorrectly be true for
4998 To: susanne@some.where
5000 This test is now done by extracting each address from the header in turn,
5001 and checking the entire address. Other tests that are part of "personal"
5002 are now done using regular expressions (for example, to check local parts
5003 of addresses in From: header lines).
5005 (2) The list of non-personal local parts in From: addresses has been
5006 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5007 taken from the Sieve specification recommendations.
5009 (3) If the message contains any header line starting with "List-" it is
5010 treated as non-personal.
5012 (4) The test for "circular" in the Subject: header line has been removed
5013 because it now seems ill-conceived.
5015 15. Minor typos in src/EDITME comments corrected.
5017 16. Installed latest exipick from John Jetmore.
5019 17. If headers_add on a router specified a text string that was too long for
5020 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5021 of string_sprintf() is now avoided.
5023 18. $message_body_size was not set (it was always zero) when running the DATA
5024 ACL and the local_scan() function.
5026 19. For the "mail" command in an Exim filter, no default was being set for
5027 the once_repeat time, causing a random time value to be used if "once" was
5028 specified. (If the value happened to be <= 0, no repeat happened.) The
5029 default is now 0s, meaning "never repeat". The "vacation" command was OK
5030 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5031 (I found it when inspecting the code).
5033 20. There is now an overall timeout for performing a callout verification. It
5034 defaults to 4 times the callout timeout, which applies to individual SMTP
5035 commands during the callout. The overall timeout applies when there is more
5036 than one host that can be tried. The timeout is checked before trying the
5037 next host. This prevents very long delays if there are a large number of
5038 hosts and all are timing out (e.g. when the network connections are timing
5039 out). The value of the overall timeout can be changed by specifying an
5040 additional sub-option for "callout", called "maxwait". For example:
5042 verify = sender/callout=5s,maxwait=20s
5044 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5045 to the end before writing, but this should make it even safer).
5047 22. Exim was forgetting that it had advertised PIPELINING for the second and
5048 subsequent messages on an SMTP connection. It was also not resetting its
5049 memory on STARTTLS and an internal HELO.
5051 23. When Exim logs an SMTP synchronization error within a session, it now
5052 records whether PIPELINING has been advertised or not.
5054 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5055 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5056 rather than long int.
5058 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5060 26. Added the never_mail option to autoreply.
5066 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5067 crash if the getsockname() call failed; this can happen if a connection is
5068 closed very soon after it is established. The problem was simply in the
5069 order in which certain operations were done, causing Exim to try to write
5070 to the SMTP stream before it had set up the file descriptor. The bug has
5071 been fixed by making things happen in the correct order.
5077 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5078 before the connection was closed, thus losing the rejection response.
5080 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5081 some early Solaris releases, but causes trouble in current releases where
5082 socklen_t is defined.
5084 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5087 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5090 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5091 that releasing the top of it at the end releases what was used for sub-
5092 expansions (unless the block got too big). However, discard this block if
5093 the first thing is a variable or header, so that we can use its block when
5094 it is dynamic (useful for very large $message_headers, for example).
5096 6. Lookups now cache *every* query, not just the most recent. A new, separate
5097 store pool is used for this. It can be recovered when all lookup caches are
5098 flushed. Lookups now release memory at the end of their result strings.
5099 This has involved some general refactoring of the lookup sources.
5101 7. Some code has been added to the store_xxx() functions to reduce the amount
5102 of flapping under certain conditions.
5104 8. log_incoming_interface used to affect only the <= reception log lines. Now
5105 it causes the local interface and port to be added to several more SMTP log
5106 lines, for example "SMTP connection from", and rejection lines.
5108 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5110 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5112 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5113 could be overwritten at the end of the current message (or the start of a
5114 new message if it was set in a HELO ACL). The value is now preserved for
5115 the duration of the SMTP connection.
5117 12. If a transport had a headers_rewrite setting, and a matching header line
5118 contained an unqualified address, that address was qualified, even if it
5119 did not match any rewriting rules. The underlying bug was that the values
5120 of the flags that permit the existence of unqualified sender and recipient
5121 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5122 non-local messages, and by -bnq for local messages) were not being
5123 preserved with the message after it was received.
5125 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5126 "next input=" as part of the text comprising the host identity instead of
5127 the correct text. The code was using the same buffer for two different
5128 strings. However, depending on which order the printing function evaluated
5129 its arguments, the bug did not always show up. Under Linux, for example, my
5130 test suite worked just fine.
5132 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5133 doesn't work with some older versions of Perl. It has been changed to "my",
5134 which in any case is probably the better facility to use.
5136 15. A really picky compiler found some instances of statements for creating
5137 error messages that either had too many or two few arguments for the format
5140 16. The size of the buffer for calls to the DNS resolver has been increased
5141 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5142 for addresses that have a lot of PTR records. This alleviates a problem; it
5143 does not fully solve it.
5145 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5146 buffer now truncates the list and logs the incident, which is the same
5147 action as happens when Exim is looking up a host name and its aliases.
5148 Previously in this situation something unpredictable would happen;
5149 sometimes it was "internal error: store_reset failed".
5151 18. If a server dropped the connection unexpectedly when an Exim client was
5152 using GnuTLS and trying to read a response, the client delivery process
5153 crashed while trying to generate an error log message.
5155 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5156 string, for example:
5158 warn message = H1: something\nH2: something
5160 the text was added as a single header line from Exim's point of view
5161 though it ended up OK in the delivered message. However, searching for the
5162 second and subsequent header lines using $h_h2: did not work. This has been
5163 fixed. Similarly, if a system filter added multiple headers in this way,
5164 the routers could not see them.
5166 20. Expanded the error message when iplsearch is called with an invalid key to
5167 suggest using net-iplsearch in a host list.
5169 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5170 ACLs are no longer actually imposed (and a message to that effect is
5173 22. If a "gecos" field in a passwd entry contained escaped characters, in
5174 particular, if it contained a \" sequence, Exim got it wrong when building
5175 a From: or a Sender: header from that name. A second bug also caused
5176 incorrect handling when an unquoted " was present following a character
5177 that needed quoting.
5179 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5180 was not being matched caselessly.
5182 24. Arranged for all hyphens in the exim.8 source to be escaped with
5185 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5186 properly. Recipient callout cache records were still being keyed to include
5187 the sender, even when use_sender was set false. This led to far more
5188 callouts that were necessary. The sender is no longer included in the key
5189 when use_sender is false.
5191 26. Added "control = submission" modifier to ACLs.
5193 27. Added the ${base62d: operator to decode base 62 numbers.
5195 28. dnsdb lookups can now access SRV records.
5197 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5198 the configuration file.
5200 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5201 (-v) mode. This makes the output for a verbose queue run more intelligible.
5203 31. Added a use_postmaster feature to recipient callouts.
5205 32. Added the $body_zerocount variable, containing the number of binary zero
5206 bytes in the message body.
5208 33. The time of last modification of the "new" subdirectory is now used as the
5209 "mailbox time last read" when there is a quota error for a maildir
5212 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5214 35. Added +ignore_unknown as a special item in host lists.
5216 36. Code for decoding IPv6 addresses in host lists is now included, even if
5217 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5218 address was recognized as an IP address, but was then not correctly decoded
5219 into binary, causing unexpected and incorrect effects when compared with
5226 1. Very minor rewording of debugging text in manualroute to say "list of
5227 hosts" instead of "hostlist".
5229 2. If verify=header_syntax was set, and a header line with an unqualified
5230 address (no domain) and a large number of spaces between the end of the
5231 name and the colon was received, the reception process suffered a buffer
5232 overflow, and (when I tested it) crashed. This was caused by some obsolete
5233 code that should have been removed. The fix is to remove it!
5235 3. When running in the test harness, delay a bit after writing a bounce
5236 message to get a bit more predictability in the log output.
5238 4. Added a call to search_tidyup() just before forking a reception process. In
5239 theory, someone could use a lookup in the expansion of smtp_accept_max_
5240 per_host which, without the tidyup, could leave open a database connection.
5242 5. Added the variables $recipient_data and $sender_data which get set from a
5243 lookup success in an ACL "recipients" or "senders" condition, or a router
5244 "senders" option, similar to $domain_data and $local_part_data.
5246 6. Moved the writing of debug_print from before to after the "senders" test
5249 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5250 problems for message scanning, either using a data ACL, or using
5251 local_scan() because the Received: header was not generated till after they
5252 were called (in order to set the time as the time of reception completion).
5253 I have revised the way this works. The header is now generated after the
5254 body is received, but before the ACL or local_scan() are called. After they
5255 are run, the timestamp in the header is updated.
5261 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5262 before starting a queue runner without re-exec. This happened only when
5263 deliver_drop_privilege was set or when the Exim user was set to root. The
5264 effect of the bug was that timeouts during subsequent deliveries caused
5265 crashes instead of being properly handled. The handler is now left at its
5266 default (and expected) setting.
5268 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5269 message, again when deliver_drop_privilege is set or Exim is run as root.
5270 The bug described in (1) was not present in this case, but the tidying up
5271 of the other signals was missing. I have made the two cases consistent.
5273 3. The ignore_target_hosts setting on a manualroute router was being ignored
5274 for hosts that were looked up using the /MX notation.
5276 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5279 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5280 operated on the sender address. After changing the $sender_address to <>
5281 for the sender address verify, Exim was re-instated it as the original
5282 (before rewriting) address, but remembering that it had rewritten it, so it
5283 wasn't rewriting it again. This bug also had the effect of breaking the
5284 sender address verification caching when the sender address was rewritten.
5286 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5287 This has been changed so that if the ip literal address matches
5288 ignore_target_hosts, the router declines.
5290 7. Added expansion conditions match_domain, match_address, and match_local_
5291 part (NOT match_host).
5293 8. The placeholder for the Received: header didn't have a length field set.
5295 9. Added code to Exim itself and to exim_lock to test for a specific race
5296 condition that could lead to file corruption when using MBX delivery. The
5297 issue is with the lockfile that is created in /tmp. If this file is removed
5298 after a process has opened it but before that process has acquired a lock,
5299 there is the potential for a second process to recreate the file and also
5300 acquire a lock. This could lead to two Exim processes writing to the file
5301 at the same time. The added code performs the same test as UW imapd; it
5302 checks after acquiring the lock that its file descriptor still refers to
5303 the same named file.
5305 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5306 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5309 11. Added the smtp_active_hostname option. If used, this will typically be made
5310 to depend on the incoming interface address. Because $interface_address is
5311 not set up until the daemon has forked a reception process, error responses
5312 that can happen earlier (such as "too many connections") no longer contain
5315 12. If an expansion in a condition on a "warn" statement fails because a lookup
5316 defers, the "warn" statement is abandoned, and the next ACL statement is
5317 processed. Previously this caused the whole ACL to be aborted.
5319 13. Added the iplsearch lookup type.
5321 14. Added ident_timeout as a log selector.
5323 15. Added tls_certificate_verified as a log selector.
5325 16. Added a global option tls_require_ciphers (compare the smtp transport
5326 option of the same name). This controls incoming TLS connections.
5328 17. I finally figured out how to make tls_require_ciphers do a similar thing
5329 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5330 before starting the TLS session.
5332 18. Tabs are now shown as \t in -bP output.
5334 19. If the log selector return_path_on_delivery was set, Exim crashed when
5335 bouncing a message because it had too many Received: header lines.
5337 20. If two routers both had headers_remove settings, and the first one included
5338 a superfluous trailing colon, the final name in the first list and the
5339 first name in the second list were incorrectly joined into one item (with a
5340 colon in the middle).
5346 1. Added -C and -D options to the exinext utility, mainly to make it easier
5347 to include in the automated testing, but these could be helpful when
5348 multiple configurations are in use.
5350 2. The exinext utility was not formatting the output nicely when there was
5351 an alternate port involved in the retry record key, nor when there was a
5352 message id as well (for retries that were specific to a specific message
5353 and a specific host). It was also confused by IPv6 addresses, because of
5354 the additional colons they contain. I have fixed the IPv4 problem, and
5355 patched it up to do a reasonable job for IPv6.
5357 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5358 delivery, the log line now contains "pipelined" if PIPELINING was used.
5360 4. An SMTP transport process used to panic and die if the bind() call to set
5361 an explicit outgoing interface failed. This has been changed; it is now
5362 treated in the same way as a connect() failure.
5364 5. A reference to $sender_host_name in the part of a conditional expansion
5365 that was being skipped was still causing a DNS lookup. This no longer
5368 6. The def: expansion condition was not recognizing references to header lines
5369 that used bh_ and bheader_.
5371 7. Added the _cache feature to named lists.
5373 8. The code for checking quota_filecount in the appendfile transport was
5374 allowing one more file than it should have been.
5376 9. For compatibility with Sendmail, the command line option
5384 and sets the incoming protocol and host name (for trusted callers). The
5385 host name and its colon can be omitted when only the protocol is to be set.
5386 Note the Exim already has two private options, -pd and -ps, that refer to
5387 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5388 "s", but I don't think that's a major issue.
5390 10. A number of refactoring changes to the code, none of which should affect
5393 (a) The number of logging options was getting close to filling up the
5394 32-bit word that was used as a bit map. I have split them into two classes:
5395 those that are passed in the argument to log_write(), and those that are
5396 only ever tested independently outside of that function. These are now in
5397 separate 32-bit words, so there is plenty of room for expansion again.
5398 There is no change in the user interface or the logging behaviour.
5400 (b) When building, for example, log lines, the code previously used a
5401 macro that called string_cat() twice, in order to add two strings. This is
5402 not really sufficiently general. Furthermore, there was one instance where
5403 it was actually wrong because one of the argument was used twice, and in
5404 one call a function was used. (As it happened, calling the function twice
5405 did not affect the overall behaviour.) The macro has been replaced by a
5406 function that can join an arbitrary number of extra strings onto a growing
5409 (c) The code for expansion conditions now uses a table and a binary chop
5410 instead of a serial search (which was left over from when there were very
5411 few conditions). Also, it now recognizes conditions like "pam" even when
5412 the relevant support is not compiled in: a suitably worded error message is
5413 given if an attempt is made to use such a condition.
5415 11. Added ${time_interval:xxxxx}.
5417 12. A bug was causing one of the ddress fields not to be passed back correctly
5418 from remote delivery subprocesses. The field in question was not being
5419 subsequently used, so this caused to problems in practice.
5421 13. Added new log selectors queue_time and deliver_time.
5423 14. Might have fixed a bug in maildirsizefile handling that threw up
5424 "unexpected character" debug warnings, and recalculated the data
5425 unnecessarily. In any case, I expanded the warning message to give more
5428 15. Added the message "Restricted characters in address" to the statements in
5429 the default ACL that block characters like @ and % in local parts.
5431 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5432 Three changes have been made:
5434 (a) There was a serious bug; a negative response to MAIL caused the whole
5435 recipient domain to be cached as invalid, thereby blocking all messages
5436 to all local parts at the same domain, from all senders. This bug has
5437 been fixed. The domain is no longer cached after a negative response to
5438 MAIL if the sender used is not empty.
5440 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5443 (c) A new callout option, "use_sender" has been added for people who want
5444 the modified behaviour.
5450 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5453 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5454 indeed breaks things for older releases.
5456 3. Added additional logging to the case where there is a problem reading data
5457 from a filter that is running in a subprocess using a pipe, in order to
5458 try to track down a specific problem.
5460 4. Testing facility fudge: when running in the test harness and attempting
5461 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5462 getting "No route to host". Convert this to a timeout.
5464 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5467 6. Some OS don't have socklen_t but use size_t instead. This affects the
5468 fifth argument of getsockopt() amongst other things. This is now
5469 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5470 can be set for individual OS. I have set it for SunOS5, OSF1, and
5471 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5472 some earlier ones do not.
5474 7. Change 4.30/15 was not doing the test caselessly.
5476 8. The standard form for an IPv6 address literal was being rejected by address
5477 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5478 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5479 this, as well as the form without the "IPv6" on the front (but only when
5480 address literals are enabled, of course).
5482 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5484 10. Exim crashed if a message with an empty sender address specified by -f
5485 encountered a router with an errors_to setting. This could be provoked only
5486 by a command such as
5490 where an empty string was supplied; "<>" did not hit this bug.
5492 11. Installed PCRE release 4.5.
5494 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5495 remained set. It is now erased.
5497 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5498 times from message ids (which are base 36 rather than the normal 62).
5500 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5501 were being counted as actual protocol errors, and logged if the log
5502 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5503 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5504 and DATA following a set of rejected RCPTs do not count as protocol errors.
5505 In other words, Exim assumes they were pipelined, though this may not
5506 actually be the case. Of course, in all cases the client gets an
5507 appropriate error code.
5509 15. If a lookup fails in an ACL condition, a message about the failure may
5510 be available; it is used if testing the ACL cannot continue, because most
5511 such messages specify what the cause of the deferral is. However, some
5512 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5513 that caused an old message to be retained and used if a later statement
5514 caused a defer, replacing the real cause of the deferral.
5516 16. If an IP address had so many PTR records that the DNS lookup buffer
5517 was not large enough to hold them, Exim could crash while trying to process
5518 the truncated data. It now detects and logs this case.
5520 17. Further to 4.21/58, another change has been made: if (and only if) the
5521 first line of a message (the first header line) ends with CRLF, a bare LF
5522 in a subsequent header line has a space inserted after it, so as not to
5523 terminate the header.
5525 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5526 unnecessarily, used atoi() instead of strtol(), and didn't check the
5527 termination when getting file sizes from file names by regex.
5529 19. Completely re-implemented the support for maildirsize files, in the light
5530 of a number of problems with the previous contributed implementation
5531 (4.30/29). In particular:
5533 . If the quota is zero, the maildirsize file is maintained, but no quota is
5536 . If the maildir directory does not exist, it is created before any attempt
5537 to write a maildirsize file.
5539 . The quota value in the file is just a cache; if the quota is changed in
5540 the transport, the new value overrides.
5542 . A regular expression is available for excluding directories from the
5545 20. The autoreply transport checks the characters in options that define the
5546 message's headers; it allows continued headers, but it was checking with
5547 isspace() after an embedded newline instead of explicitly looking for a
5550 21. If all the "regular" hosts to which an address was routed had passed their
5551 expiry times, and had not reached their retry times, the address was
5552 bounced, even if fallback hosts were defined. Now Exim should go on to try
5555 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5556 equivalent code in the SMTP transport. Some hosts send humungous responses
5557 to HELO/EHLO, more than 1024 it seems.
5559 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5560 gives compiler warnings in some environments. I've now done it "properly",
5563 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5564 (because of problems with the built-in one) was declared to return uschar *
5565 instead of char *, causing compiler failure.
5567 25. Fixed a file descriptor leak when processing alias/forward files.
5569 26. Fixed a minor format string issue in dbfn.c.
5571 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5573 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5574 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5577 29. When a sender address is verified, it is cached, to save repeating the test
5578 when there is more than one recipient in a message. However, when the
5579 verification involves a callout, it is possible for different callout
5580 options to be set for different recipients. It is too complicated to keep
5581 track of this in the cache, so now Exim always runs a verification when a
5582 callout is required, relying on the callout cache for the optimization.
5583 The overhead is duplication of the address routing, but this should not be
5586 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5587 to be verified with callout=postmaster, and the main callout worked but the
5588 postmaster check failed, the verification correctly failed. However, if a
5589 subsequent RCPT command asked for sender verification *without* the
5590 postmaster check, incorrect caching caused this verification also to fail,
5593 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5594 it was not caching the DNS options (qualify_single, search_parents) that
5595 were used when the lookup failed. A subsequent lookup with different
5596 options therefore always gave the same answer, though there were cases
5597 where it should not have. (Example: a "domains = !$mx_any" option on a
5598 dnslookup router: the "domains" option is always processed without any
5599 widening, but the router might have qualify_single set.) Now Exim uses the
5600 cached value only when the same options are set.
5602 32. Added John Jetmore's "exipick" utility to the distribution.
5604 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5605 than a timeout (e.g. a certificate is required, and is not provided), an
5606 Exim server now closes the connection immediately. Previously it waited for
5607 the client to close - but if the client is SSL, it seems that they each
5608 wait for each other, leading to a delay before one of them times out.
5610 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5611 maintained 0.8.x compatibility because I don't think many are using it, and
5612 it is clearly obsolete.
5614 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5617 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5618 client certificate was expired. A simple patch fixes this, though I don't
5619 understand the full logic of why the verify callback is called multiple
5622 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5623 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5624 which causes problems with some clients (such as the Certicom SSL Plus
5625 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5626 disables the coutermeasure allowing Eudora to connect."
5628 38. Exim was not checking that a write() to a log file succeeded. This could
5629 lead to Bad Things if a log got too big, in particular if it hit a file
5630 size limit. Exim now panics and dies if it cannot write to a log file, just
5631 as it does if it cannot open a log file.
5633 39. Modified OS/Makefile-Linux so that it now contains
5635 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5637 The two -D definitions ensure that Exim is compiled with large file
5638 support, which makes it possible to handle log files that are bigger than
5641 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5642 instance) a domain was checked against a named list that involved a lookup,
5643 causing $domain_data to be set, then another domain was checked against the
5644 same list, then the first domain was re-checked, the value of $domain_data
5645 after the final check could be wrong. In particular, if the second check
5646 failed, it could be set empty. This bug probably also applied to
5649 41. The strip_trailing_dot option was not being applied to the address given
5650 with the -f command-line option.
5652 42. The code for reading a message's header from the spool was incrementing
5653 $received_count, but never initializing it. This meant that the value was
5654 incorrect (doubled) while delivering a message in the same process in which
5655 it was received. In the most common configuration of Exim, this never
5656 happens - a fresh exec is done - but it can happen when
5657 deliver_drop_privilege is set.
5659 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5660 it now includes up to 150 characters of the unexpected data in the log
5663 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5664 and building data strings. The size of both of these buffers was 10 000
5665 bytes - far larger than anybody would *ever* want, thought I. Needless to
5666 say, somebody hit the limit. I have increased the maximum line length to
5667 20 000 and the maximum data length of concatenated lines to 100 000. I have
5668 also fixed two bugs, because there was no checking on these buffers. Tsk,
5669 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5670 buffer is too small.
5672 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5673 lsearch lookups. Now it does.
5675 46. When parsing a route_list item in a manualroute router, a fixed-length
5676 buffer was used for the list of hosts. I made this 1024 bytes long,
5677 thinking that nobody would ever have a list of hosts that long. Wrong.
5678 Somebody had a whole pile of complicated expansion conditions, and the
5679 string was silently truncated, leading to an expansion error. It turns out
5680 that it is easier to change to an unlimited length (owing to other changes
5681 that have happened since this code was originally written) than to build
5682 structure for giving a limitation error. The length of the item that
5683 expands into the list of hosts is now unlimited.
5685 47. The lsearch lookup could not handle data where the length of text line was
5686 more than 4095 characters. Such lines were truncated, leading to shortened
5687 data being returned. It should now handle lines of any length.
5689 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5690 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5693 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5694 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5695 no longer happens in comment lines. A list of replacements is now placed
5696 at the head of all of the source files, except those whose only change is
5697 to replace PERL_COMMAND in the very first #! line.
5699 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5700 messages on the queue, with a bottom-up merge sort, using code contributed
5701 by Michael Haardt. This should make operations like -bp somewhat faster on
5702 large queues. It won't affect queue runners, except when queue_run_in_order
5705 51. Installed eximstats 1.31 in the distribution.
5707 52. Added support for SRV lookups to the dnslookup router.
5709 53. If an ACL referred to $message_body or $message_body_end, the value was not
5710 reset for any messages that followed in the same SMTP session.
5712 54. The store-handling optimization for building very long strings was not
5713 differentiating between the different store pools. I don't think this
5714 actually made any difference in practice, but I've tidied it.
5716 55. While running the routers to verify a sender address, $sender_address
5717 was still set to the sender address. This is wrong, because when routing to
5718 send a bounce to the sender, it would be empty. Therefore, I have changed
5719 it so that, while verifying a sender address, $sender_address is set to <>.
5720 (There is no change to what happens when verifying a recipient address.)
5722 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5723 target A or AAAA records (if not already returned) without resetting the
5724 qualify_single or search_parents options of the DNS resolver. These are
5725 inappropriate in this case because the targets of MX and SRV records must
5726 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5727 target that, when qualified, matched something in the local domain. These
5728 two options are now turned off when doing these lookups.
5730 57. It seems that at least some releases of Reiserfs (which does not have the
5731 concept of a fixed number of inodes) returns zero and not -1 for the
5732 number of available inodes. This interacted badly with check_spool_inodes,
5733 which assumed that -1 was the "no such thing" setting. What I have done is
5734 to check that the total number of inodes is greater than zero before doing
5735 the test of how many are available.
5737 58. When a "warn" ACL statement has a log_message modifier, the message is
5738 remembered, and not repeated. This is to avoid a lot of repetition when a
5739 message has many recipients that cause the same warning to be written.
5740 However, Exim was preserving the list of already written lines for an
5741 entire SMTP session, which doesn't seem right. The memory is now reset if a
5742 new message is started.
5744 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5745 debugging output unless log_rewrite was also set.
5747 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5748 of (int)(handle) when we know that handle contains (void *)(-1).
5750 61. The Exim daemon panic-logs an error return when it closes the incoming
5751 connection. However "connection reset by peer" seems to be common, and
5752 isn't really an error worthy of noting specially, so that particular error
5755 62. When Exim is trying to find all the local interfaces, it used to panic and
5756 die if the ioctl to get the interface flags failed. However, it seems that
5757 on at least one OS (Solaris 9) it is possible to have an interface that is
5758 included in the list of interfaces, but for which you get a failure error
5759 for this call. This happens when the interface is not "plumbed" into a
5760 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5761 failure of the "get flags" call assumes that the interface is down.
5763 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5764 makes life easier for people who are doing arithmetic on fields extracted
5765 from dates, where you often get leading zeros that should not be
5766 interpreted as octal.
5768 64. Added qualify_domain to the redirect router, to override the global
5771 65. If a pathologically long header line contained very many addresses (the
5772 report of this problem mentioned 10 000) and each of them was rewritten,
5773 Exim could use up a very large amount of memory. (It kept on making new
5774 copies of the header line as it rewrote, and never released the old ones.)
5775 At the expense of a bit more processing, the header rewriting function has
5776 been changed so that it no longer eats memory in this way.
5778 66. The generation of the Received: header has been moved from the time that a
5779 message starts to be received, to the time that it finishes. The timestamp
5780 in the Received: header should now be very close to that of the <= log
5781 line. There are two side-effects of this change:
5783 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5784 logged header lines no longer include the local Received: line, because
5785 it has not yet been created. The same applies to a copy of the message
5786 that is returned to a non-SMTP sender when a message is rejected.
5788 (b) When a filter file is tested using -bf, no additional Received: header
5789 is added to the test message. After some thought, I decided that this
5792 This change does not affect the value of $received_for. It is still set
5793 after address rewriting, but before local_scan() is called.
5795 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5797 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5798 gave an unhelpful panic error message, and a defer error. I have managed to
5799 change this behaviour so that it now rejects any supplied certificate,
5800 which seems right, as the list of acceptable certificates is empty.
5802 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5803 gave an unhelpful defer error. I have not managed to make this reject any
5804 supplied certificates, but the error message it gives is "no certificate
5805 supplied", which is not helpful.
5807 70. exigrep's output now also includes lines that are not associated with any
5808 message, but which match the given pattern. Implemented by a patch from
5809 Martin Sluka, which also tidied up the Perl a bit.
5811 71. Recipient callout verification, like sender verification, was using <> in
5812 the MAIL FROM command. This isn't really the right thing, since the actual
5813 sender may affect whether the remote host accepts the recipient or not. I
5814 have changed it to use the actual sender in the callout; this means that
5815 the cache record is now keyed on a recipient/sender pair, not just the
5816 recipient address. There doesn't seem to be a real danger of callout loops,
5817 since a callout by the remote host to check the sender would use <>.
5818 [SEE ABOVE: changed after hitting problems.]
5820 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5821 temporary errors. However, in the case of such a code being given after
5822 the end of a data transmission (i.e. after ".") Exim was failing to write
5823 a retry record for the message. (Yes, there was some broken host that was
5824 actually sending 8xx at this point.)
5826 73. An unknown lookup type in a host list could cause Exim to panic-die when
5827 the list was checked. (An example that provoked this was putting <; in the
5828 middle of a list instead of at the start.) If this happened during a DATA
5829 ACL check, a -D file could be left lying around. This kind of configuration
5830 error no longer causes Exim to die; instead it causes a defer error. The
5831 incident is still logged to the main and panic logs.
5833 74. Buglet left over from Exim 3 conversion. The message "too many messages
5834 in one connection" was written to the rejectlog but not the mainlog, except
5835 when address rewriting (yes!) was being logged.
5837 75. Added write_rejectlog option.
5839 76. When a system filter was run not as root (that is, when system_filter_user
5840 was set), the values of the $n variables were not being returned to the
5841 main process; thus, they were not subsequently available in the $sn
5844 77. Added +return_path_on_delivery log selector.
5846 78. A connection timeout was being treated differently from recipients deferred
5847 when testing hosts_max_try with a message that was older than the host's
5848 retry timeout. (The host should not be counted, thus allowing all hosts to
5849 be tried at least once before bouncing.) This may have been the cause of an
5850 occasionally reported bug whereby a message would remain on the queue
5851 longer than the retry timeout, but would be bounced if a delivery was
5852 forced. I say "may" because I never totally pinned down the problem;
5853 setting up timeout/retry tests is difficult. See also the next item.
5855 79. The ultimate address timeout was not being applied to errors that involved
5856 a combination of host plus message (for example, a timeout on a MAIL
5857 command). When an address resolved to a number of possible hosts, and they
5858 were not all tried for each delivery (e.g. because of hosts_max_try), a
5859 message could remain on the queue longer than the retry timeout.
5861 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5864 81. Fixed an obscure SMTP outgoing bug which required at least the following
5865 conditions: (a) there was another message waiting for the same server;
5866 (b) the server returned 5xx to all RCPT commands in the first message so
5867 that the message was not completed; (c) the server dropped the connection
5868 or gave a negative response to the RSET that Exim sends to abort the
5869 transaction. The observed case was a dropped connection after DATA that had
5870 been sent in pipelining mode. That is, the server had advertised PIPELINING
5871 but was not implementing it correctly. The effect of the bug was incorrect
5872 behaviour, such as trying another host, and this could lead to a crash.
5878 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5879 and daemon.c were passed as pointers to ints; they should have been
5880 pointers to socklen_t variables (which are typically unsigned ints).
5882 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5885 3. Fixed a really odd bug that affected only the testing scheme; patching a
5886 certain fixed string in the binary changed the value of another string that
5887 happened to be identical to the end of the original first string.
5889 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5890 name", it returns that address as the IP address. On some operating
5891 systems (e.g. Solaris), it also passes back the IP address string as the
5892 "host name". However, on others (e.g. Linux), it passes back an empty
5893 string. Exim wasn't checking for this, and was changing the host name to an
5894 empty string, assuming it had been canonicalized.
5896 5. Although rare, it is permitted to have more than one PTR record for a given
5897 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5898 all the names associated with an address, because they do in Solaris.
5899 However, it seems that they do not in Linux for data that comes from the
5900 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5901 I found this out when I moved to a new Linux workstation and tried to run
5902 the Exim test suite.
5904 To get round this problem I have changed the code so that it now does its
5905 own call to the DNS to look up PTR records when searching for a host name.
5906 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5907 addresses that are only in /etc/hosts are still found.
5909 This behaviour is, however, controlled by an option called host_lookup_
5910 order, which defaults to "bydns:byaddr". If people want to use the other
5911 order, or indeed, just use one or the other means of lookup, they can
5912 specify it in this variable.
5914 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5915 some operating systems, this comes back from gethostbyaddr() as an empty
5916 string, and this is what Exim used to test for. However, it seems that in
5917 other systems, "." is yielded. Exim now tests for this case too.
5919 7. The values of check_spool_space and check_log_space are now held internally
5920 as a number of kilobytes instead of an absolute number of bytes. If a
5921 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5922 kilobyte. This means that much larger values can be stored.
5924 8. Exim monitor: an attempt to get the action menu when not actually pointing
5925 at a message produces an empty menu entitled "No message selected". This
5926 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5927 no entries in it ("Shell widget menu has zero width and/or height"). So I
5928 have added a single, blank menu entry in this case.
5930 9. Added ${quote_local_part.
5932 10. MIME decoding is now applied to the contents of Subject: header lines when
5935 11. Now that a reference to $sender_host_address automatically causes a reverse
5936 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5937 host lookup before query-style lookups in lists that might use this
5938 variable. This has therefore been abolished, and the "net-" prefix is no
5939 longer necessary for query-style lookups.
5941 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5942 have been a typo for LFLAGS, so it has been changed.
5944 13. The install script calls Exim with "-C /dev/null" in order to find the
5945 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5946 to be output. However, since Exim outputs its version number before the
5947 error, it didn't break the script. It just looked ugly. I fixed this by
5948 always allowing "-C /dev/null" if the caller is root.
5950 14. Ignore overlarge ACL variable number when reading spool file - insurance
5951 against a later release with more variables having written the file.
5953 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5954 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5955 this, as well as the form without the "IPv6" on the front.
5957 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5958 OS/Makefile-Darwin file.
5960 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5961 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5963 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5964 variable. However, some OS use ints for inodes. I've added cast to long int
5965 to get rid of the compiler warning.
5967 19. I had forgotten to lock out "/../" in configuration file names when
5968 ALT_CONFIG_PREFIX was set.
5970 20. Routers used for verification do not need to specify transports. However,
5971 if such a router generated a host list, and callout was configured, Exim
5972 crashed, because it could not find a port number from the (non-existent)
5973 transport. It now assumes port 25 in this circumstance.
5975 21. Added the -t option to exigrep.
5977 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5978 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5979 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5980 them is set, LOOKUP_LSEARCH is forced.
5982 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5984 24. Added sender and host information to the "rejected by local_scan()" log
5985 line; previously there was no indication of these.
5987 25. Added .include_if_exists.
5989 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5990 Linux. It turns out that not all file systems support this. Apparently some
5991 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5992 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5993 not supported on the file descriptor, is now ignored when Exim is trying to
5994 sync a directory. This applies only to Linux.
5996 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5998 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5999 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6002 29. Added support for maildirsize files from supplied patch (modified a bit).
6004 30. The use of :fail: followed by an empty string could lead Exim to respond to
6005 sender verification failures with (e.g.):
6007 550 Verification failed for <xxx>
6008 550 Sender verify failed
6010 where the first response line was missing the '-' that indicates it is not
6011 the final line of the response.
6013 31. The loop for finding the name of the user that called Exim had a hardwired
6014 limit of 10; it now uses the value of finduser_retries, which is used for
6015 all other user lookups.
6017 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6020 33. Exim was neglecting to zero errno before one call of strtol() when
6021 expanding a string and expecting an integer value. On some systems this
6022 resulted in spurious "integer overflow" errors. Also, it was casting the
6023 result into an int without checking.
6025 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6026 did not work. The code looks as if it has *never* worked, though it appears
6027 to have been documented since at least release 1.62. I have made it work.
6029 35. The "timeout_DNS" error in retry rules, also documented since at least
6030 1.62, also never worked. As it isn't clear exactly what this means, and
6031 clearly it isn't a major issue, I have abolished the feature by treating it
6032 as "timeout", and writing a warning to the main and panic logs.
6034 36. The display of retry rules for -brt wasn't always showing the error code
6037 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6038 timeout_connect_A, timeout_connect_MX.
6040 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6041 to the empty sender.
6043 39. The daemon was not analysing the content of -oX till after it had closed
6044 stderr and disconnected from the controlling terminal. This meant that any
6045 syntax errors were only noted on the panic log, and the return code from
6046 the command was 0. By re-arranging the code a little, I've made the
6047 decoding happen first, so such errors now appear on stderr, and the return
6048 code is 1. However, the actual setting up of the sockets still happens in
6049 the disconnected process, so errors there are still only recorded on the
6052 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6053 connections (as happens on some IP stacks) was logged at start up time as
6054 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6055 it from "IPv6 and IPv4", which means that two separate sockets are being
6058 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6059 says whether AF_INET or AF_INET6 was passed as an argument.
6061 42. Exiwhat output was messed up when time zones were included in log
6064 43. Exiwhat now gives more information about the daemon's listening ports,
6065 and whether -tls-on-connect was used.
6067 44. The "port" option of the smtp transport is now expanded.
6069 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6070 silently ignored. Now an error message is written to the main and panic
6073 46. There's a new ACL modifier called "logwrite" which writes to a log file
6074 as soon as it is encountered.
6076 47. Added $local_user_uid and $local_user_gid at routing time.
6078 48. Exim crashed when trying to verify a sender address that was being
6081 49. Exim was recognizing only a space character after ".include". It now also
6082 recognizes a tab character.
6084 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6085 extracting the relevant information from the specification. The man page no
6086 longer contains scrambled data for the -d option, and I've added a section
6087 at the front about calling Exim under different names.
6089 51. Added "extra_headers" argument to the "mail" command in filter files.
6091 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6094 53. Installed eximstats 1.29.
6096 54. Added transport_filter_timeout as a generic transport option.
6098 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6099 Cc: header lines. This was required by RFC 822, but it not required by RFC
6102 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6103 incoming messages that did not have them. Now it does so only if the
6104 message originates locally, that is, if there is no associated remote host
6105 address. When Resent- header lines are present, this applies to the Resent-
6106 lines rather than the non-Resent- lines.
6108 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6109 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6111 58. Messages for configuration errors now include the name of the main
6112 configuration file - useful now that there may be more than one file in a
6113 list (.included file names were always shown).
6115 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6116 for those rare installations that do not start the daemon as root or run it
6117 setuid root. I've cut out the call to initgroups() if the daemon is not
6120 60. The Exim user and group can now be bound into the binary as text strings
6121 that are looked up at the start of Exim's processing.
6123 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6125 62. Added $mailstore_basename variable.
6127 63. Installed patch to sieve.c from Michael Haardt.
6129 64. When Exim failed to open the panic log after failing to open the main log,
6130 the original message it was trying to log was written to stderr and debug
6131 output, but if they were not available (the usual case in production), it
6132 was lost. Now it is written to syslog before the two lines that record the
6133 failures to open the logs.
6135 65. Users' Exim filters run in subprocesses under the user's uid. It is
6136 possible for a "deliver" command or an alias in a "personal" command to
6137 provoke an address rewrite. If logging of address rewriting is configured,
6138 this fails because the process is not running as root or exim. There may be
6139 a better way of dealing with this, but for the moment (because 4.30 needs
6140 to be released), I have disabled address rewrite logging when running a
6141 filter in a non-root, non-exim process.
6147 1. The buildconfig auxiliary program wasn't quoting the value set for
6148 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6149 not defined. This bug was masked in 4.22 by the effect that was fixed in
6152 2. Some messages that were rejected after a message id was allocated were
6153 shown as "incomplete" by exigrep. It no longer does this for messages that
6154 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6156 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6157 have allow_domain_literals set, the ID did not get logged in the <= line.
6158 Domain literals are now always recognized in Message-ID: header lines.
6160 4. The first argument for a ${extract expansion item is the key name or field
6161 number. Leading and trailing spaces in this item were not being ignored,
6162 causing some misleading effects.
6164 5. When deliver_drop_privilege was set, single queue runner processes started
6165 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6166 same command in the process it spins off) were not dropping privilege.
6168 6. When the daemon running as "exim" started a queue runner, it always
6169 re-executed Exim in the spun-off process. This is a waste of effort when
6170 deliver_drop_privilege is set. The new process now just calls the
6171 queue-runner function directly.
6177 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6180 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6181 ignored. Though the use of -oP was forcing the writing of a pid file, it
6182 was always written to the default place.
6184 3. If the message "no IP address found for host xxxx" is generated during
6185 incoming verification, it is now followed by identification of the incoming
6186 connection (so you can more easily find what provoked it).
6188 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6190 5. Added some features to "harden" Exim a bit more against certain attacks:
6192 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6193 be put in Local/Makefile. This is like the never_users runtime option,
6194 but it cannot be overridden. The default setting is "root".
6196 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6197 prefix string with which any file named in a -C command line option
6200 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6201 is retained for -C and -D only if the caller of Exim is root. Without
6202 it, the exim user may also use -C and -D and retain privilege.
6204 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6205 command line option is disabled.
6207 6. Macro names set by the -D option must start with an upper case letter, just
6208 like macro names defined in the configuration file.
6210 7. Added "dereference=" facility to LDAP.
6212 8. Two instances of the typo "uknown" in the source files are fixed.
6214 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6215 the Configure-Makefile script screwed up while processing it.
6217 10. Incorporated PCRE 4.4.
6219 11. The SMTP synchronization check was not operating right at the start of an
6220 SMTP session. For example, it could not catch a HELO sent before the client
6221 waited for the greeting. There is now a check for outstanding input at the
6222 point when the greeting is written. Because of the duplex, asynchronous
6223 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6224 way, but not yet received, when the check is performed.
6226 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6227 on TCP/IP sockets, because this apparently causes some broken clients to
6230 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6231 unchanged) from the Cygwin maintainer.
6233 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6234 was supported instead of testing for mbx. Effectively a typo.
6236 15. The spa authenticator server code was not checking that the input it
6237 received was valid base64.
6239 16. The debug output line for the "set" modifier in ACLs was not showing the
6240 name of the variable that was being set.
6242 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6244 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6245 lookup on its own. Something else was needed to trigger the lookup. For
6246 example, a match in host_lookup or the need for a host name in a host list.
6247 Now, if $sender_host_name is referenced and the host name has not yet been
6248 looked up, a lookup is performed. If the lookup fails, the variable remains
6249 empty, and $host_lookup_failed is set to "1".
6251 19. Added "eqi" as a case-independent comparison operator.
6253 20. The saslauthd authentication condition could segfault if neither service
6254 nor realm was specified.
6256 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6257 error message that was logged was misleading, and incoming SMTP
6258 connections were dropped. The message is now more accurate, and temporary
6259 errors are given to SMTP connections.
6261 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6262 (except RSET) with a 421 error, until QUIT is received. However, it was
6263 failing to send a response to QUIT.
6265 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6266 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6267 ACL is now run after the helo_try_verify_hosts code.
6269 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6270 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6271 case-independent, so other case variants are also recognized). Apparently
6272 some systems use these upper case variants.
6274 25. If more than two messages were waiting for the same host, and a transport
6275 filter was specified for the transport, Exim sent two messages over the
6276 same TCP/IP connection, and then failed with "socket operation on non-
6277 socket" when it tried to send the third.
6279 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6281 27. The extern definition of crypt16() in expand.c was not being excluded when
6282 the OS had its own crypt16() function.
6284 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6285 as a preferred synonym for return_size_limit, both as an option and as an
6288 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6290 30. Changed the default configuration ACL to relax the local part checking rule
6291 for addresses that are not in any local domains. For these addresses,
6292 slashes and pipe symbols are allowed within local parts, but the sequence
6293 /../ is explicitly forbidden.
6295 31. SPA server authentication was not clearing the challenge buffer before
6298 32. log_message in a "warn" ACL statement was writing to the reject log as
6299 well as to the main log, which contradicts the documentation and doesn't
6300 seem right (because no rejection is happening). So I have stopped it.
6302 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6303 However, I am unable to do any testing of this.
6305 34. Fixed an infelicity in the appendfile transport. When checking directories
6306 for a mailbox, to see if any needed to be created, it was accidentally
6307 using path names with one or more superfluous leading slashes; tracing
6308 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6310 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6311 contents are added to the log line that is written for every discarded
6312 recipient. (Previously a log_message setting was ignored.)
6314 36. The ${quote: operator now quotes the string if it is empty.
6316 37. The install script runs exim in order to find its version number. If for
6317 some reason other than non-existence or emptiness, which it checks, it
6318 could not run './exim', it was installing it with an empty version number,
6319 i.e. as "exim-". This error state is now caught, and the installation is
6322 38. An argument was missing from the function that creates an error message
6323 when Exim fails to connect to the socket for saslauthd authentication.
6324 This could cause Exim to crash, or give a corrupted message.
6326 39. Added isip, isip4, and isip6 to ${if conditions.
6328 40. The ACL variables $acl_xx are now saved with the message, and can be
6329 accessed later in routers, transports, and filters.
6331 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6332 strings in the file are not string-expanded.
6334 42. If a MAIL command specified a SIZE value that was too large to fit into an
6335 int variable, the check against message_size_limit failed. Such values are
6336 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6337 day this will have to be increased, but I don't think I want to be around
6338 when emails are that large.
6345 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6346 iconv() is not standard in FreeBSD.
6348 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6349 IPv6 enabled. The observed symptom was a segmentation fault on return from
6350 the function os_common_find_running_interfaces() in src/os.c.
6352 3. In the check_special_case() function in daemon.c I had used "errno" as an
6353 argument name, which causes warnings on some systems. This was basically a
6354 typo, since it was named "eno" in the comments!
6356 4. The code that waits for the clock to tick (at a resolution of some fraction
6357 of a second) so as to ensure message-id uniqueness was always waiting for
6358 at least one whole tick, when it could have waited for less. [This is
6359 almost certainly not relevant at current processor speeds, where it is
6360 unlikely to ever wait at all. But we try to future-proof.]
6362 5. The function that sleeps for a time interval that includes fractions of a
6363 second contained a race. It did not block SIGALRM between setting the
6364 timer, and suspending (a couple of lines later). If the interval was short
6365 and the sigsuspend() was delayed until after it had expired, the suspension
6366 never ended. On busy systems this could lead to processes getting stuck for
6369 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6370 process, before it forks any delivery processes. The open lookup caching
6371 mechanism meant that the open file or database connection was passed into
6372 the delivery process. The problem was that delivery processes always tidy
6373 up cached lookup data. This could cause a problem for the next delivery
6374 process started by the queue runner, because the external queue runner
6375 process does not know about the closure. So the next delivery process
6376 still has data in the lookup cache. In the case of a file lookup, there was
6377 no problem because closing a file descriptor in a subprocess doesn't affect
6378 the parent. However, if the lookup was caching a connection to a database,
6379 the connection was closed, and the second delivery process was likely to
6380 see errors such as "PGSQL: query failed: server closed the connection
6381 unexpectedly". The problem has been fixed by closing all cached lookups
6382 in a queue runner before running a delivery process.
6384 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6385 seem to have the "const" qualifier which it has on other OS. I've
6388 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6389 *of the same type* (client or server) with the same public name that an
6390 error should be diagnosed.
6392 9. When Exim looked up a host name for an IP address, but failed to find the
6393 original IP address when looking up the host name (a safety check), it
6394 output the message "<ip address> does not match any IP for NULL", which was
6395 confusing, to say the least. The bug was that the host name should have
6396 appeared instead of "NULL".
6398 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6399 user that is built into the binary, and the -C or -D options is used, root
6400 privilege is dropped before the configuration file is read. In addition,
6401 logging is switched to stderr instead of the normal log files. If the
6402 configuration then re-defines the Exim user, the unprivileged environment
6403 is probably not what is expected, so Exim logs a panic warning message (but
6406 However, if deliver_drop_privilege is set, the unprivileged state may well
6407 be exactly what is intended, so the warning has been cut out in that case,
6408 and Exim is allowed to try to write to its normal log files.
6414 1. smtp_return_error_details was not giving details for temporary sender
6415 or receiver verification errors.
6417 2. Diagnose a configuration error if two authenticators have the same public
6420 3. Exim used not to create the message log file for a message until the first
6421 delivery attempt. This could be confusing when incoming messages were held
6422 for policy or load reasons. The message log file is now created at the time
6423 the message is received, and an initial "Received" line is written to it.
6425 4. The automatically generated man page for command line options had a minor
6426 bug that caused no ill effects; however, a more serious problem was that
6427 the procedure for building the man page automatically didn't always
6428 operate. Consequently, release 4.20 contains an out-of-date version. This
6429 shouldn't happen again.
6431 5. When building Exim with embedded Perl support, the script that builds the
6432 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6433 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6435 6. The freeze_tell option was not being used for messages that were frozen on
6436 arrival, either by an ACL or by local_scan().
6438 7. Added the smtp_incomplete_transaction log selector.
6440 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6441 was accepting AUTH without a new EHLO.
6443 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6444 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6447 10. Logging of TCP/IP connections (when configured) now happens in the main
6448 daemon process instead of the child process, so that the TCP/IP connection
6449 count is more accurate (but it can never be perfect).
6451 11. The use of "drop" in a nested ACL was not being handled correctly in the
6452 outer ACL. Now, if condition failure induced by the nested "drop" causes
6453 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6454 or "require"), the connection is dropped.
6456 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6457 that yield "discard" can now be used with an "accept" or a "discard" verb,
6458 but an error is generated for any others (because I can't see a useful way
6459 to define what should happen).
6461 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6462 are now processed in the same way as lines in the Exim configuration file.
6463 In particular, continuation lines are supported.
6465 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6467 15. Added -ti meaning -t -i.
6469 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6470 domains, and warn by logging if others are found.
6472 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6473 returned from GIFCONF when Exim is trying to find the list of interfaces on
6474 a host. The code in os.c has been modified to copy each ifreq to an aligned
6475 structure in all cases.
6477 Also, in some cases, the returned ifreq's were being copied to a 'struct
6478 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6479 means the last couple of bytes of an IPv6 address could be chopped if the
6480 ifreq contained only a normal sockaddr (14 bytes storage).
6482 18. Named domain lists were not supported in the hosts_treat_as_local option.
6483 An entry such as +xxxx was not recognized, and was treated as a literal
6486 19. Ensure that header lines added by a DATA ACL are included in the reject log
6487 if the ACL subsequently rejects the message.
6489 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6490 MD5 (which is deprecated).
6492 21. When testing a filter file using -bf, Exim was writing a message when it
6493 took the sender from a "From " line in the message, but it was not doing so
6494 when it took $return_path from a Return-Path: header line. It now does.
6496 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6497 with a valid header line field name (a series of printing characters
6498 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6500 23. Changed "disc" in the source to "disk" to conform to the documentation and
6501 the book and for uniformity.
6503 24. Ignore Sendmail's -Ooption=value command line item.
6505 25. When execve() failed while trying to run a command in a pipe transport,
6506 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6507 could be confused with a return value of 69 from the command itself. This
6508 has been changed to 127, the value the shell returns if it is asked to run
6509 a non-existent command. The wording for the related log line suggests a
6510 non-existent command as the problem.
6512 26. If received_header_text expands to an empty string, do not add a Received:
6513 header line to the message. (Well, it adds a token one on the spool, but
6514 marks it "old" so that it doesn't get used or transmitted.)
6516 27. Installed eximstats 1.28 (addition of -nt option).
6518 28. There was no check for failure on the call to getsockname() in the daemon
6519 code. This can fail if there is a shortage of resources on the system, with
6520 ENOMEM, for example. A temporary error is now given on failure.
6522 29. Contrary to the C standard, it seems that in some environments, the
6523 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6524 program. Exim now does this explicitly; it affects the formatting of
6525 timestamps using strftime().
6527 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6528 complaints. I've now initialized all the variables, to avoid this.
6530 32. Header lines added by a system filter were not being "seen" during
6531 transport-time rewrites.
6533 33. The info_callback() function passed to OpenSSL is set up with type void
6534 (*)(SSL *, int, int), as described somewhere. However, when calling the
6535 function (actually a macro) that sets it up, the type void(*)() is
6536 expected. I've put in a cast to prevent warnings from picky compilers.
6538 34. If a DNS black list lookup found a CNAME record, but there were no A
6539 records associated with the domain it pointed at, Exim crashed.
6541 35. If a DNS black list lookup returned more than one A record, Exim ignored
6542 all but the first. It now scans all returned addresses if a particular IP
6543 value is being sought. In this situation, the contents of the
6544 $dnslist_value variable are a list of all the addresses, separated by a
6547 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6548 to accept a host name and all its aliases if the forward lookup for any of
6549 them yielded the IP address of the incoming connection. Now it accepts only
6550 those names whose forward lookup yields the correct IP address. Any other
6551 names are discarded. This closes a loophole whereby a rogue DNS
6552 administrator could create reverse DNS records to break through a
6553 wildcarded host restriction in an ACL.
6555 37. If a user filter or a system filter that ran in a subprocess used any of
6556 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6557 the wrong values were passed to the pipe command ($thisaddress had the
6558 value of $0, $0 had the value of $1, etc). This bug was introduced by
6559 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6561 38. Improved the line breaking for long SMTP error messages from ACLs.
6562 Previously, if there was no break point between 40 and 75 characters, Exim
6563 left the rest of the message alone. Two changes have been made: (a) I've
6564 reduced the minimum length to 35 characters; (b) if it can't find a break
6565 point between 35 and 75 characters, it looks ahead and uses the first one
6566 that it finds. This may give the occasional overlong line, but at least the
6567 remaining text gets split now.
6569 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6570 file descriptors might be low, and that setting 1000 would always raise it.
6571 It turns out that in some environments, the limit is already over 1000 and
6572 that lowering it causes trouble. So now Exim takes care not to decrease it.
6574 40. When delivering a message, the value of $return_path is set to $sender_
6575 address at the start of routing (routers may change the value). By an
6576 oversight, this default was not being set up when an address was tested by
6577 -bt or -bv, which affected the outcome if any router or filter referred to
6580 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6581 log only when "message" or "log_message" are set. However, if one of the
6582 conditions was an address verification, or a call to a nested ACL, the
6583 messages generated by the underlying test were being passed through. This
6584 no longer happens. The underlying message is available in $acl_verify_
6585 message for both "message" and "log_message" expansions, so it can be
6586 passed through if needed.
6588 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6589 new expansion $bh_ to give the encoded byte string without charset
6590 translation. Translation happens only if iconv() is available; HAVE_ICONV
6591 indicates this at build time. HEADERS_CHARSET gives the charset to
6592 translate to; headers_charset can change it in the configuration, and
6593 "headers charset" can change it in an individual filter file.
6595 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6596 that creates RFC 2047 encoded "words" labels them as that charset instead
6597 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6598 expansion operator; (ii) when Exim creates a From: line for a local
6599 message; (iii) when a header line is rewritten to include a "phrase" part.
6601 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6602 buggy, causing it to skip the first lines of messages whose message ID
6603 ended in 'D'. This would not have bitten before Exim release 4.14, because
6604 message IDs were unlikely to end in 'D' before then. The effect was to have
6605 incorrect size information for certain domains.
6607 45. #include "config.h" was missing at the start of the crypt16.c module. This
6608 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6611 46. If there was a timeout during a "random" callout check, Exim treated it as
6612 a failure of the random address, and carried on sending RSET and the real
6613 address. If the delay was just some slowness somewhere, the response to the
6614 original RCPT would be taken as a response to RSET and so on, causing
6615 mayhem of various kinds.
6617 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6618 when I implemented it. It didn't allow for the fact that some option values
6619 may legitimately be negative (e.g. size_addition), and it didn't even do
6620 the right test for positive values.
6622 48. Domain names in DNS records are case-independent. Exim always looks them up
6623 in lower case. Some resolvers return domain names in exactly the case they
6624 appear in the zone file, that is, they may contain uppercase letters. Not
6625 all resolvers do this - some return always lower case. Exim was treating a
6626 change of case by a resolver as a change of domain, similar to a widening
6627 of a domain abbreviation. This triggered its re-routing code and so it was
6628 trying to route what was effectively the same domain again. This normally
6629 caused routing to fail (because the router wouldn't handle the domain
6630 twice). Now Exim checks for this case specially, and just changes the
6631 casing of the domain that it ultimately uses when it transmits the message
6634 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6637 50. If a filter generated a file delivery with a non-absolute name (possible if
6638 no home directory exists for the router), the forbid_file option was not
6641 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6642 the existing equality matching.
6644 52. Exim was using ints instead of ino_t variables in some places where it was
6645 dealing with inode numbers.
6647 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6648 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6649 called TMPDIR, and if it finds it is different, it changes its value.
6651 54. The smtp_printf() function is now made available to local_scan() so
6652 additional output lines can be written before returning. There is also an
6653 smtp_fflush() function to enable the detection of a dropped connection.
6654 The variables smtp_input and smtp_batched_input are exported to
6657 55. Changed the default runtime configuration: the message "Unknown user"
6658 has been removed from the ACL, and instead placed on the localuser router,
6659 using the cannot_route_message feature. This means that any verification
6660 failures that generate their own messages won't get overridden. Similarly,
6661 the "Unrouteable address" message that was in the ACL for unverifiable
6662 relay addresses has also been removed.
6664 56. Added hosts_avoid_esmtp to the smtp transport.
6666 57. The exicyclog script was not checking for the esoteric option
6667 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6668 will work only if exicyclog is run under the appropriate euid.
6670 58. Following a discussion on the list, the rules by which Exim recognises line
6671 endings on incoming messages have been changed. The -dropcr and drop_cr
6672 options are now no-ops, retained only for backwards compatibility. The
6673 following line terminators are recognized: LF CRLF CR. However, special
6674 processing applies to CR:
6676 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6677 nor a local message in the state where . is a terminator.
6679 (ii) If a bare CR is encountered in a header line, an extra space is added
6680 after the line terminator so as not to end the header. The reasoning
6681 behind this is that bare CRs in header lines are most likely either
6682 to be mistakes, or people trying to play silly games.
6684 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6685 was being incorrectly given as 18 bytes larger than it should have been.
6686 This is a VOB (very old bug).
6688 60. This may never have affected anything current, but just in case it has:
6689 When the local host is found other than at the start of a list of hosts,
6690 the local host, those with the same MX, and any that follow, are discarded.
6691 When the list in question was part of a longer list of hosts, the following
6692 hosts (not currently being processed) were also being discarded. This no
6693 longer happens. I'm not sure if this situation could ever has previously
6696 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6699 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6700 header. This is something that is recommended in a new Internet Draft, and
6701 is something that is documented as being done by Sendmail. There are two
6702 possible values. For messages generated by the autoreply transport, Exim
6705 Auto-Submitted: auto-replied
6707 whereas for all other generated messages (e.g. bounces) it adds
6709 Auto-Submitted: auto-generated
6711 63. The "personal" condition in filters now includes a test for the
6712 Auto-Submitted: header. If it contains the string "auto-" the message it
6713 not considered personal.
6715 64. Added rcpt_include_affixes as a generic transport option.
6717 65. Added queue_only_override (default true).
6719 66. Added the syslog_duplication option.
6721 67. If what should have been the first header line of a message consisted of
6722 a space followed by a colon, Exim was mis-interpreting it as a header line.
6723 It isn't of course - it is syntactically invalid and should therefore be
6724 treated as the start of the message body. The misbehaviour could have
6725 caused a number of strange effects, including loss of data in subsequent
6726 header lines, and spool format errors.
6728 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6729 client host had authenticated. This control can now be exercised by an ACL
6730 for more flexibility.
6732 69. By default, callouts do not happen when testing with -bh. There is now a
6733 variant, -bhc, which does actually run the callout code, including
6734 consulting and updating the callout cache.
6736 70. Added support for saslauthd authentication, courtesy of Alexander
6739 71. If statvfs() failed on the spool or log directories while checking their
6740 size for availability, Exim confusingly gave the error "space shortage".
6741 Furthermore, in debugging mode it crashed with a floating point exception.
6742 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6743 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6744 serious problem, Exim now writes to the main and panic logs when this
6745 happens, with details of the failure. It then refuses to accept the
6746 incoming message, giving the message "spool directory problem" or "log
6747 directory problem" with a 421 code for SMTP messages.
6749 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6750 0, 1, and 2 exist, because some OS complain for execs without them (see
6751 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6752 descriptors. However, the code omitted to check that the open succeeded,
6753 causing mysterious errors if for some reason the permissions on /dev/null
6754 got screwed. Now Exim writes a message to the main and panic logs, and
6755 bombs out if it can't open /dev/null.
6757 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6758 interact so that it is all more flexible. It is supposed to remain
6759 backwards compatible. Also added extra_local_interfaces.
6761 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6762 to bomb out with an assertion failure - to the client this appears as a
6763 connection drop. This problem occurs in the part of the code that was taken
6764 from the Samba project. Fortunately, the assertion is in a very simple
6765 function, so I have fixed this by reproducing the function inline in the
6766 one place where it is called, and arranging for authentication to fail
6767 instead of killing the process with assert().
6769 75. The SPA client code was not working when the server requested OEM rather
6770 than Unicode encoding.
6772 76. Added code to make require_files with a specific uid setting more usable in
6773 the case where statting the file as root fails - usually a non-root-mounted
6774 NFS file system. When this happens and the failure is EACCES, Exim now
6775 forks a subprocess and does the per-uid checking as the relevant uid.
6777 77. Added process_log_path.
6779 78. If log_file_path was not explicitly set, a setting of check_log_space or
6780 check_log_inodes was ignored.
6782 79. If a space check for the spool or log partitions fails, the incident is now
6783 logged. Of course, in the latter case the data may get lost...
6785 80. Added the %p formatting code to string_format() so that it can be used to
6786 print addresses in debug_print(). Adjusted all the address printing in the
6787 debugging in store.c to use %p rather than %d.
6789 81. There was a concern that a line of code in smtp_in.c could overflow a
6790 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6791 initially expressed, the concern was not well-founded, because trailing
6792 spaces are removed early. However, if the trailing spaces were followed by
6793 a NULL, they did not get removed, so the overflow was possible. Two fixes
6796 (a) I re-wrote the offending code in a cleaner fashion.
6797 (b) If an incoming SMTP command contains a NULL character, it is rejected
6800 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6801 runs initgroups(), so that if the Exim user is in any additional groups,
6802 they will be used during message reception.
6808 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob