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/>.
52 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
53 the check for any unsuccessful recipients did not notice the limit, and
54 erroneously found still-pending ones.
56 JH/08 Pipeline CHUNKING command and data together, on kernels that support
57 MSG_MORE. Only in-clear (not on TLS connections).
59 JH/09 Avoid using a temporary file during transport using dkim. Unless a
60 transport-filter is involved we can buffer the headers in memory for
61 creating the signature, and read the spool data file once for the
62 signature and again for transmission.
68 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
69 than -2003 did; needs libidn2 in addition to libidn.
71 JH/02 The path option on a pipe transport is now expanded before use.
73 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
74 Patch provided by "Björn", documentation fix added too.
76 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
77 missing a wire-to-host endian conversion.
79 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
80 close after a BDAT command line could be taken as a following command,
81 giving a synch failure. Fix by only checking for synch immediately
82 before acknowledging the chunk.
84 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
85 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
86 macro. Patches provided by Josh Soref.
88 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
89 Previously we did not; the RFC seems ambiguous and VRFY is not listed
90 by IANA as a service extension. However, John Klensin suggests that we
93 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
94 the dkim code may be unix-mode line endings rather than smtp wire-format
95 CRLF, so prepend a CR to any bare LF.
97 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
98 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
100 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
101 the first were themselves being wrongly included in the feed into dkim
102 processing; with most chunk sizes in use this resulted in an incorrect
103 body hash calculated value.
105 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
106 DKIM signature block, for verification. Although advised against by
107 standards it is specifically not ruled illegal.
109 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
111 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
112 missing a body hash (the bh= tag).
114 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
115 It seems that HAProxy sends the Proxy Protocol information in clear and
116 only then does a TLS startup, so do the same.
118 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
119 TCP connections (such as for Spamd) unless the daemon successfully set
120 Fast Open mode on its listening sockets. This fixes breakage seen on
121 too-old kernels or those not configured for Fast Open, at the cost of
122 requiring both directions being enabled for TFO, and TFO never being used
123 by non-daemon-related Exim processes.
125 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
126 endings, at least on the first header line. Try to canonify any that get
127 past that check, despite the cost.
129 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
130 now limited to an arbitrary five deep, while parsing addresses with the
131 strip_excess_angle_brackets option enabled.
133 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
134 instead leave the unprompted TLS handshake in socket buffer for the
135 TLS library to consume.
137 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
139 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
141 JH/16 Drop variables when they go out of scope. Memory management drops a whole
142 region in one operation, for speed, and this leaves assigned pointers
143 dangling. Add checks run only under the testsuite which checks all
144 variables at a store-reset and panics on a dangling pointer; add code
145 explicitly nulling out all the variables discovered. Fixes one known
146 bug: a transport crash, where a dangling pointer for $sending_ip_address
147 originally assigned in a verify callout, is re-used.
149 PP/06 Drop '.' from @INC in various Perl scripts.
151 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
153 PP/08 Reduce a number of compilation warnings under clang; building with
154 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
155 should be warning-free.
157 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
159 HS/01 Fix portability problems introduced by PP/08 for platforms where
160 realloc(NULL) is not equivalent to malloc() [SunOS et al].
162 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
163 chunk. This allows us to accept broken chunked messages. We need a more
164 general solution here.
166 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
167 already-broken messages in the queue.
169 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
171 JH/19 Fix reference counting bug in routing-generated-address tracking.
177 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
178 supports it and a size is available (ie. the sending peer gave us one).
180 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
181 years of being deprecated). The replacements are the ACLs
182 acl_smtp_mime and acl_not_smtp_mime.
184 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
185 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
186 or even an in-clear connection were permitted. Now, if the host lookup
187 was dnssec and dane was requested then the host is only used if the
188 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
189 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
190 if one fails this test.
191 This means that a poorly-configured remote DNS will make it incommunicado;
192 but it protects against a DNS-interception attack on it.
194 JH/04 Bug 1810: make continued-use of an open smtp transport connection
195 non-noisy when a race steals the message being considered.
197 JH/05 If main configuration option tls_certificate is unset, generate a
198 self-signed certificate for inbound TLS connections.
200 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
201 in rewrites and routers.
203 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
204 and logged a warning sing 4.83; now they are a configuration file error.
206 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
207 (lacking @domain). Apply the same qualification processing as RCPT.
209 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
211 JH/10 Support ${sha256:} applied to a string (as well as the previous
214 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
215 a cutthrough deliver is pending, as we always want to make a connection.
216 This also avoids re-routing the message when later placing the cutthrough
217 connection after a verify cache hit.
218 Do not update it with the verify result either.
220 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
221 when routing results in more than one destination address.
223 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
224 signing (which inhibits the cutthrough capability). Previously only
225 the presence of an option was tested; now an expansion evaluating as
226 empty is permissible (obviously it should depend only on data available
227 when the cutthrough connection is made).
229 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
230 the relevant preceding SMTP command did not note the pipelining mode.
232 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
233 Previously they were not counted.
235 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
236 as one having no matching records. Previously we deferred the message
237 that needed the lookup.
239 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
240 distinguished as "(=".
242 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
243 for missing MX records. Previously it only worked for missing A records.
245 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
247 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
248 after the data-go-ahead and data-ack. Patch from Jason Betts.
250 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
251 even for a "none" policy. Patch from Tony Meyer.
253 JH/22 Fix continued use of a connection for further deliveries. If a port was
254 specified by a router, it must also match for the delivery to be
257 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
258 When one of the recipients of a message was unsuitable for the connection
259 (has no matching addresses), we lost track of needing to mark it
260 deferred. As a result mail would be lost.
262 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
264 JH/25 Decoding ACL controls is now done using a binary search; the source code
265 takes up less space and should be simpler to maintain. Merge the ACL
266 condition decode tables also, with similar effect.
268 JH/26 Fix problem with one_time used on a redirect router which returned the
269 parent address unchanged. A retry would see the parent address marked as
270 delivered, so not attempt the (identical) child. As a result mail would
273 JH/27 Fix a possible security hole, wherein a process operating with the Exim
274 UID can gain a root shell. Credit to http://www.halfdog.net/ for
275 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
278 JH/28 Enable {spool,log} filesystem space and inode checks as default.
279 Main config options check_{log,spool}_{inodes,space} are now
280 100 inodes, 10MB unless set otherwise in the configuration.
282 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
283 Previously it only applied to the main-section connection policy
286 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
288 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
289 by me. Added RFC7919 DH primes as an alternative.
291 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
292 are not in the system include path.
294 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
295 GnuTLS, when a session startup failed (eg because the client disconnected)
296 Exim did stdio operations after fclose. This was exposed by a recent
297 change which nulled out the file handle after the fclose.
299 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
300 signed directly by the cert-signing cert, rather than an intermediate
301 OCSP-signing cert. This is the model used by LetsEncrypt.
303 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
305 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
306 an incoming connection.
308 HS/02 Bug 1802: Do not half-close the connection after sending a request
311 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
312 fallback to "prime256v1".
314 JH/34 SECURITY: Use proper copy of DATA command in error message.
315 Could leak key material. Remotely exploitable. CVE-2016-9963.
321 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
322 and 3.4.4 - once the server is enabled to respond to an OCSP request
323 it does even when not requested, resulting in a stapling non-aware
324 client dropping the TLS connection.
326 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
327 support variable-length bit vectors. No functional change.
329 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
330 The I= interface field on outgoing lines is now after the H= remote
331 host field, same as incoming lines. There is a separate
332 outgoing_interface log selector which allows you to disable the
335 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
336 If not running log_selector +smtp_connection the mainlog would be held
337 open indefinitely after a "too many connections" event, including to a
338 deleted file after a log rotate. Leave the per net connection logging
339 leaving it open for efficiency as that will be quickly detected by the
340 check on the next write.
342 HS/01 Bug 1671: Fix post transport crash.
343 Processing the wait-<transport> messages could crash the delivery
344 process if the message IDs didn't exist for some reason. When
345 using 'split_spool_directory=yes' the construction of the spool
346 file name failed already, exposing the same netto behaviour.
348 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
349 mime_regex ACL conditions.
351 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
352 to DSN fail messages (bounces): remote IP, remote greeting, remote response
353 to HELO, local diagnostic string.
355 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
356 log line to debug. Even when configured with a tls authenticator many
357 client connections are expected to not authenticate in this way, so
358 an authenticate fail is not an error.
360 HS/02 Add the Exim version string to the process info. This way exiwhat
361 gives some more detail about the running daemon.
363 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
364 matter for fast-change records such as DNSBLs.
366 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
367 retry record, even if constant. There may be multiple transports with
368 different interface settings and the retry behaviour needs to be kept
371 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
373 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
375 JH/10 Bug 840: fix log_defer_output option of pipe transport
377 JH/11 Bug 830: use same host for all RCPTS of a message, even under
378 hosts_randomize. This matters a lot when combined with mua_wrapper.
380 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
381 ${quote_pgsql:<string>} operator.
383 JH/13 Bug 1708: avoid misaligned access in cached lookup.
385 JH/14 Change header file name for freeradius-client. Relevant if compiling
386 with Radius support; from the Gentoo tree and checked under Fedora.
388 JH/15 Bug 1712: Introduce $prdr_requested flag variable
390 JH/16 Bug 1714: Permit an empty string as expansion result for transport
391 option transport_filter, meaning no filtering.
393 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
395 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
396 defaults to "*" (all hosts). The variable is now available when not built
397 with TLS, default unset, mainly to enable keeping the testsuite sane.
398 If a server certificate is not supplied (via tls_certificate) an error is
399 logged, and clients will find TLS connections fail on startup. Presumably
400 they will retry in-clear.
401 Packagers of Exim are strongly encouraged to create a server certificate
402 at installation time.
404 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
405 with the $config_file variable.
407 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
408 in transport context, after the attempt, and per-recipient. The latter type
409 is per host attempted. The event data is the error message, and the errno
410 information encodes the lookup type (A vs. MX) used for the (first) host,
411 and the trailing two digits of the smtp 4xx response.
413 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
414 to write to mainlog (or rejectlog, paniclog) in the window between file
415 creation and permissions/ownership being changed. Particularly affects
416 installations where exicyclog is run as root, rather than exim user;
417 result is that the running daemon panics and dies.
419 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
421 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
422 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
423 "pri" and "weight". Note that the previous implicit priority given by the
424 list order is no longer honoured.
426 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
429 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
430 by defining SUPPORT_SOCKS.
432 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
433 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
434 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
435 variables are renamed to proxy_{local,external}_{address,port}.
437 JH/27 Move Internationalisation support from Experimental to mainline, enabled
438 for a build by defining SUPPORT_I18N
440 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
441 of the query string, and make ${quote_redis:} do that quoting.
443 JH/29 Move Events support from Experimental to mainline, enabled by default
444 and removable for a build by defining DISABLE_EVENT.
446 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
448 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
449 cached by the daemon.
451 JH/32 Move Redis support from Experimental to mainline, enabled for a build
452 by defining LOOKUP_REDIS. The libhiredis library is required.
454 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
455 keys are given for lookup.
457 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
458 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
459 only supported when built with TLS support. The PolarSSL SHA routines
460 are still used when the TLS library is too old for convenient support.
462 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
463 openssl_options), for security. OpenSSL forces this from version 1.1.0
464 server-side so match that on older versions.
466 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
467 allocation for $value could be released as the expansion processing
468 concluded, but leaving the global pointer active for it.
470 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
471 and to use the domains and local_parts ACL conditions.
473 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
474 incorrectly not doubled on cutthrough transmission, hence seen as a
475 body-termination at the receiving system - resulting in truncated mails.
476 Commonly the sender saw a TCP-level error, and retransmitted the message
477 via the normal store-and-forward channel. This could result in duplicates
478 received - but deduplicating mailstores were liable to retain only the
479 initial truncated version.
481 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
483 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
485 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
486 we're in there, support oversigning also; bug 1309.
488 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
490 HS/04 Add support for keep_environment and add_environment options.
492 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
493 either intentional arithmetic overflow during PRNG, or testing config-
496 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
497 delivery resulted in actual delivery. Cancel cutthrough before DATA
500 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
501 rejects a recipient: pass the reject to the originator.
503 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
504 Many were false-positives and ignorable, but it's worth fixing the
507 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
508 for the new environment-manipulation done at startup. Move the routines
509 from being local to tls.c to being global via the os.c file.
511 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
512 an extract embedded as result-arg for a map, the first arg for extract
513 is unavailable so we cannot tell if this is a numbered or keyed
514 extraction. Accept either.
520 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
523 JH/02 The smtp transport option "multi_domain" is now expanded.
525 JH/03 The smtp transport now requests PRDR by default, if the server offers
528 JH/04 Certificate name checking on server certificates, when exim is a client,
529 is now done by default. The transport option tls_verify_cert_hostnames
530 can be used to disable this per-host. The build option
531 EXPERIMENTAL_CERTNAMES is withdrawn.
533 JH/05 The value of the tls_verify_certificates smtp transport and main options
534 default to the word "system" to access the system default CA bundle.
535 For GnuTLS, only version 3.0.20 or later.
537 JH/06 Verification of the server certificate for a TLS connection is now tried
538 (but not required) by default. The verification status is now logged by
539 default, for both outbound TLS and client-certificate supplying inbound
542 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
545 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
546 Status Notification (bounce) messages are now MIME format per RFC 3464.
547 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
548 under the control of the dsn_advertise_hosts option, and routers may
549 have a dsn_lasthop option.
551 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
552 default, modifiable by a malware= option. The list separator for
553 the options can now be changed in the usual way. Bug 68.
555 JH/10 The smtp_receive_timeout main option is now expanded before use.
557 JH/11 The incoming_interface log option now also enables logging of the
558 local interface on delivery outgoing connections.
560 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
561 if the interface and destination host and port all match.
563 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
566 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
567 Patch from Andrew Lewis.
569 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
570 now supports optional time-restrictions, weighting, and priority
571 modifiers per server. Patch originally by <rommer@active.by>.
573 JH/16 The spamd_address main option now supports a mixed list of local
574 and remote servers. Remote servers can be IPv6 addresses, and
575 specify a port-range.
577 JH/17 Bug 68: The spamd_address main option now supports an optional
578 timeout value per server.
580 JH/18 Bug 1581: Router and transport options headers_add/remove can
581 now have the list separator specified.
583 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
586 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
589 JH/21 Support for the A6 type of dns record is withdrawn.
591 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
592 rather than the verbs used.
594 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
595 from 255 to 1024 chars.
597 JH/24 Verification callouts now attempt to use TLS by default.
599 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
600 are generic router options now. The defaults didn't change.
602 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
603 Original patch from Alexander Shikoff, worked over by JH.
605 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
606 files if ZCAT_COMMAND is not executable.
608 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
610 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
612 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
613 Normally benign, it bites when the pair was led to by a CNAME;
614 modern usage is to not canonicalize the domain to a CNAME target
615 (and we were inconsistent anyway for A-only vs AAAA+A).
617 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
619 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
620 when evaluating $sender_host_dnssec.
622 JH/31 Check the HELO verification lookup for DNSSEC, adding new
623 $sender_helo_dnssec variable.
625 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
627 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
629 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
631 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
632 documented as working, but never had. Support all but $spam_report.
634 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
635 added for tls authenticator.
637 HS/03 Add perl_taintmode main config option
643 TL/01 When running the test suite, the README says that variables such as
644 no_msglog_check are global and can be placed anywhere in a specific
645 test's script, however it was observed that placement needed to be near
646 the beginning for it to behave that way. Changed the runtest perl
647 script to read through the entire script once to detect and set these
648 variables, reset to the beginning of the script, and then run through
649 the script parsing/test process like normal.
651 TL/02 The BSD's have an arc4random API. One of the functions to induce
652 adding randomness was arc4random_stir(), but it has been removed in
653 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
654 function when detected.
656 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
657 cause callback expansion.
659 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
660 syntax errors in an expansion can be treated as a string instead of
661 logging or causing an error, due to the internal use of bool_lax
662 instead of bool when processing it.
664 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
665 server certificates when making smtp deliveries.
667 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
669 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
671 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
673 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
674 Merged patch from Sebastian Wiedenroth.
676 JH/05 Fix results-pipe from transport process. Several recipients, combined
677 with certificate use, exposed issues where response data items split
678 over buffer boundaries were not parsed properly. This eventually
679 resulted in duplicates being sent. This issue only became common enough
680 to notice due to the introduction of connection certificate information,
681 the item size being so much larger. Found and fixed by Wolfgang Breyha.
683 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
684 size buffer was used, resulting in syntax errors when an expansion
687 JH/07 Add support for directories of certificates when compiled with a GnuTLS
688 version 3.3.6 or later.
690 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
691 is EXPERIMENTAL_EVENT, the main-configuration and transport options
692 both become "event_action", the variables become $event_name, $event_data
693 and $event_defer_errno. There is a new variable $verify_mode, usable in
694 routers, transports and related events. The tls:cert event is now also
695 raised for inbound connections, if the main configuration event_action
698 TL/06 In test suite, disable OCSP for old versions of openssl which contained
699 early OCSP support, but no stapling (appears to be less than 1.0.0).
701 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
702 server certificate names available under the smtp transport option
703 "tls_verify_cert_hostname" now do not permit multi-component wildcard
706 JH/10 Time-related extraction expansions from certificates now use the main
707 option "timezone" setting for output formatting, and are consistent
708 between OpenSSL and GnuTLS compilations. Bug 1541.
710 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
711 encoded parameter in the incoming message. Bug 1558.
713 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
714 include certificate info, eximon was claiming there were spoolfile
717 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
719 JH/14 Log delivery-related information more consistently, using the sequence
720 "H=<name> [<ip>]" wherever possible.
722 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
723 are problematic for Debian distribution, omit them from the release
726 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
728 JH/16 Fix string representation of time values on 64bit time_t architectures.
731 JH/17 Fix a null-indirection in certextract expansions when a nondefault
732 output list separator was used.
737 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
738 checkers that were complaining about end of non-void function with no
741 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
742 This was a regression introduced in 4.83 by another bugfix.
744 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
746 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
747 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
753 TF/01 Correctly close the server side of TLS when forking for delivery.
755 When a message was received over SMTP with TLS, Exim failed to clear up
756 the incoming connection properly after forking off the child process to
757 deliver the message. In some situations the subsequent outgoing
758 delivery connection happened to have the same fd number as the incoming
759 connection previously had. Exim would try to use TLS and fail, logging
760 a "Bad file descriptor" error.
762 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
763 utilities have not been installed.
765 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
766 temporary space as the ACL may create new global variables.
768 TL/01 LDAP support uses per connection or global context settings, depending
769 upon the detected version of the libraries at build time.
771 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
772 to extract and use the src ip:port in logging and expansions as if it
773 were a direct connection from the outside internet. PPv2 support was
774 updated based on HAProxy spec change in May 2014.
776 JH/02 Add ${listextract {number}{list}{success}{fail}}.
778 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
779 Properly escape header and check for NULL return.
781 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
784 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
786 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
787 characters in header names, implemented as a verify condition.
788 Contributed by Michael Fischer v. Mollard.
790 TL/05 Rename SPF condition results err_perm and err_temp to standardized
791 results permerror and temperror. Previous values are deprecated but
792 still accepted. In a future release, err_perm and err_temp will be
793 completely removed, which will be a backward incompatibility if the
794 ACL tests for either of these two old results. Patch contributed by
795 user bes-internal on the mailing list.
797 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
799 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
800 selectors, in both main and reject logs.
802 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
805 JH/07 Add malware type "sock" for talking to simple daemon.
807 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
809 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
810 routers/transports under cutthrough routing.
812 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
813 numbers. Touch up "bool" conditional to keep the same definition.
815 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
817 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
819 JH/12 Expand items in router/transport headers_add or headers_remove lists
820 individually rather than the list as a whole. Bug 1452.
822 Required for reasonable handling of multiple headers_ options when
823 they may be empty; requires that headers_remove items with embedded
824 colons must have them doubled (or the list-separator changed).
826 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
827 view the policy declared in the DMARC record. Currently, $dmarc_status
828 is a combined value of both the record presence and the result of the
831 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
833 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
834 dnslookup router and the smtp transport (applying to the forward
837 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
838 of ldap servers used for a specific lookup. Patch provided by Heiko
841 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
842 New variable $lookup_dnssec_authenticated for observability.
844 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
845 Patch submitted by Lars Timman.
847 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
849 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
850 Requires trusted mode and valid format message id, aborts otherwise.
851 Patch contributed by Heiko Schlichting.
853 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
854 certextract with support for various fields. Bug 1358.
856 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
857 is requested by default, modifiable by smtp transport option
860 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
861 operate on certificate variables to give certificate fingerprints
862 Also new ${sha256:cert_variable}.
864 JH/23 The PRDR feature is moved from being Experimental into the mainline.
866 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
867 Christian Aistleitner.
869 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
871 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
872 file. Patch from Wolfgang Breyha.
874 JH/25 Expand the coverage of the delivery $host and $host_address to
875 client authenticators run in verify callout. Bug 1476.
877 JH/26 Port service names are now accepted for tls_on_connect_ports, to
878 align with daemon_smtp_ports. Bug 72.
880 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
881 support and error reporting did not work properly.
883 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
884 and is readable. Patch from Andrew Colin Kissa.
886 TL/14 Enhance documentation of ${run expansion and how it parses the
887 commandline after expansion, particularly in the case when an
888 unquoted variable expansion results in an empty value.
890 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
892 PP/02 Fix internal collision of T_APL on systems which support RFC3123
893 by renaming away from it. Addresses GH issue 15, reported by
896 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
898 TL/15 SECURITY: prevent double expansion in math comparison functions
899 (can expand unsanitized data). Not remotely exploitable.
906 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
908 PP/02 Make -n do something, by making it not do something.
909 When combined with -bP, the name of an option is not output.
911 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
914 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
915 $sender_host_name and config options to manage this, and basic check
918 PP/05 DSCP support for outbound connections and control modifier for inbound.
920 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
921 (Only plugin which currently uses this is kerberos4, which nobody should
922 be using, but we should make it available and other future plugins might
923 conceivably use it, even though it would break NAT; stuff *should* be
924 using channel bindings instead).
926 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
927 name; added for Sendmail compatibility; requires admin caller.
928 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
929 just ignore it); requires trusted caller.
930 Also parse but ignore: -Ac -Am -X<logfile>
933 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
935 TL/02 Add +smtp_confirmation as a default logging option.
937 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
938 Patch by Magnus Holmgren from 2007-02-20.
940 TL/04 Bugzilla 1281 - Spec typo.
941 Bugzilla 1283 - Spec typo.
942 Bugzilla 1290 - Spec grammar fixes.
944 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
946 TL/06 Add Experimental DMARC support using libopendmarc libraries.
948 TL/07 Fix an out of order global option causing a segfault. Reported to dev
949 mailing list by by Dmitry Isaikin.
951 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
953 JH/02 Support "G" suffix to numbers in ${if comparisons.
955 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
957 NM/01 Bugzilla 1197 - Spec typo
958 Bugzilla 1196 - Spec examples corrections
960 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
962 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
963 gnutls_enable_pkcs11, but renamed to more accurately indicate its
966 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
967 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
969 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
970 "acl {{name}{arg}...}", and optional args on acl condition
973 JH/05 Permit multiple router/transport headers_add/remove lines.
975 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
977 JH/07 Avoid using a waiting database for a single-message-only transport.
978 Performance patch from Paul Fisher. Bugzilla 1262.
980 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
983 JH/09 Add $headers_added variable, with content from use of ACL modifier
984 add_header (but not yet added to the message). Bugzilla 199.
986 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
987 Pulled from Bugzilla 817 by Wolfgang Breyha.
989 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
991 (nb: this is the same fix as in Exim 4.80.1)
993 JH/11 Add A= logging on delivery lines, and a client_set_id option on
996 JH/12 Add optional authenticated_sender logging to A= and a log_selector
999 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1001 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1002 advertise SMTP AUTH mechanism to us, instead of a generic
1003 protocol violation error. Also, make Exim more robust to bad
1004 data from the Dovecot auth socket.
1006 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1008 When a queue runner is handling a message, Exim first routes the
1009 recipient addresses, during which it prunes them based on the retry
1010 hints database. After that it attempts to deliver the message to
1011 any remaining recipients. It then updates the hints database using
1014 So if a recipient address works intermittently, it can get repeatedly
1015 deferred at routing time. The retry hints record remains fresh so the
1016 address never reaches the final cutoff time.
1018 This is a fairly common occurrence when a user is bumping up against
1019 their storage quota. Exim had some logic in its local delivery code
1020 to deal with this. However it did not apply to per-recipient defers
1021 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1023 This change adds a proper retry rule check during routing so that the
1024 final cutoff time is checked against the message's age. We only do
1025 this check if there is an address retry record and there is not a
1026 domain retry record; this implies that previous attempts to handle
1027 the address had the retry_use_local_parts option turned on. We use
1028 this as an approximation for the destination being like a local
1029 delivery, as in LMTP.
1031 I suspect this new check makes the old local delivery cutoff check
1032 redundant, but I have not verified this so I left the code in place.
1034 TF/02 Correct gecos expansion when From: is a prefix of the username.
1036 Test 0254 submits a message to Exim with the header
1040 When I ran the test suite under the user fanf2, Exim expanded
1041 the header to contain my full name, whereas it should have added
1042 a Resent-Sender: header. It erroneously treats any prefix of the
1043 username as equal to the username.
1045 This change corrects that bug.
1047 GF/01 DCC debug and logging tidyup
1048 Error conditions log to paniclog rather than rejectlog.
1049 Debug lines prefixed by "DCC: " to remove any ambiguity.
1051 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1053 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1054 Bug spotted by Jeremy Harris; was flawed since initial commit.
1055 Would have resulted in OCSP responses post-SNI triggering an Exim
1056 NULL dereference and crash.
1058 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1060 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1061 Bug detection, analysis and fix by Samuel Thibault.
1062 Bugzilla 1331, Debian bug #698092.
1064 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1066 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1067 Server implementation by Todd Lyons, client by JH.
1068 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1069 config variable "prdr_enable" controls whether the server
1070 advertises the facility. If the client requests PRDR a new
1071 acl_data_smtp_prdr ACL is called once for each recipient, after
1072 the body content is received and before the acl_smtp_data ACL.
1073 The client is controlled by both of: a hosts_try_prdr option
1074 on the smtp transport, and the server advertisement.
1075 Default client logging of deliveries and rejections involving
1076 PRDR are flagged with the string "PRDR".
1078 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1079 fclose(). Diagnosis by Todd Lyons.
1081 PP/17 Update configure.default to handle IPv6 localhost better.
1082 Patch by Alain Williams (plus minor tweaks).
1085 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1086 This is now consistent with GnuTLS, and is now documented: the
1087 previous undocumented portable approach to treating the option as
1088 unset was to force an expansion failure. That still works, and
1089 an empty string is now equivalent.
1091 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1092 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1093 not performing validation itself.
1095 PP/20 Added force_command boolean option to pipe transport.
1096 Patch from Nick Koston, of cPanel Inc.
1098 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1101 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1103 PP/21 Fix eximon continuous updating with timestamped log-files.
1104 Broken in a format-string cleanup in 4.80, missed when I repaired the
1105 other false fix of the same issue.
1106 Report and fix from Heiko Schlichting.
1109 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1110 Report from Prashanth Katuri.
1112 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1113 It's SecureTransport, so affects any MacOS clients which use the
1114 system-integrated TLS libraries, including email clients.
1116 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1117 using a MIME ACL for non-SMTP local injection.
1118 Report and assistance in diagnosis by Warren Baker.
1120 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1122 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1124 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1125 last id that failed so it may be referenced in subsequent ACL's.
1127 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1130 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1131 ldap library initialization, allowing self-signed CA's to be
1132 used. Also properly sets require_cert option later in code by
1133 using NULL (global ldap config) instead of ldap handle (per
1134 session). Bug diagnosis and testing by alxgomz.
1136 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1137 the src/util/ subdirectory.
1139 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1140 renamed to Transport Post Delivery Action by Jeremy Harris, as
1143 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1144 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1145 redis_servers = needs to be configured which will be used by the redis
1146 lookup. Patch from Warren Baker, of The Packet Hub.
1148 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1150 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1151 hostname or reverse DNS when processing a host list. Used suggestions
1152 from multiple comments on this bug.
1154 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1156 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1157 Missed a few lines, added it to make the runtest require no keyboard
1160 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1161 contains upper case chars. Make router use caseful_local_part.
1163 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1164 support when GnuTLS has been built with p11-kit.
1170 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1172 This, or similar/improved, will also be change PP/11 of 4.82.
1178 PP/01 Handle short writes when writing local log-files.
1179 In practice, only affects FreeBSD (8 onwards).
1180 Bugzilla 1053, with thanks to Dmitry Isaikin.
1182 NM/01 Bugzilla 949 - Documentation tweak
1184 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1187 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1189 PP/02 Implemented gsasl authenticator.
1191 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1193 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1194 `pkg-config foo` for cflags/libs.
1196 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1197 with rest of GSASL and with heimdal_gssapi.
1199 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1200 `pkg-config foo` for cflags/libs for the TLS implementation.
1202 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1203 properties get this fed in as external SSF. A number of robustness
1204 and debugging improvements to the cyrus_sasl authenticator.
1206 PP/08 cyrus_sasl server now expands the server_realm option.
1208 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1209 Patch by Jeremy Harris.
1211 PP/10 Added dbmjz lookup type.
1213 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1215 PP/12 MAIL args handles TAB as well as SP, for better interop with
1216 non-compliant senders.
1217 Analysis and variant patch by Todd Lyons.
1219 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1220 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1221 Patch from Dirk Mueller <dmueller@suse.com>
1223 PP/13 tls_peerdn now print-escaped for spool files.
1224 Observed some $tls_peerdn in wild which contained \n, which resulted
1225 in spool file corruption.
1227 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1228 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1229 or write after TLS renegotiation, which otherwise led to messages
1232 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1233 as a tracking header (ie: a signed header comes before the signature).
1234 Patch from Wolfgang Breyha.
1236 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1237 comma-sep list; embedded commas doubled.
1239 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1241 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1243 Report and patch from Dmitry Banschikov.
1245 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1246 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1247 failures. We appear to now support TLS1.1+ with Exim.
1249 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1250 lets Exim select keys and certificates based upon TLS SNI from client.
1251 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1252 before an outbound SMTP session. New log_selector, +tls_sni.
1254 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1255 NULL dereference. Report and patch from Alun Jones.
1257 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1258 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1259 Not seeing resolver debug output on NetBSD, but suspect this is a
1260 resolver implementation change.
1262 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1263 Left warnings. Added "eximon gdb" invocation mode.
1265 PP/21 Defaulting "accept_8bitmime" to true, not false.
1267 PP/22 Added -bw for inetd wait mode support.
1269 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1270 locate the relevant includes and libraries. Made this the default.
1272 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1273 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1275 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1276 This may cause build issues on older platforms.
1278 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1279 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1280 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1281 Added SNI support via GnuTLS too.
1282 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1284 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1286 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1287 Applied second patch from Janne, implementing suggestion to default
1288 multiple-strings-in-record handling to match SPF spec.
1290 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1292 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1293 read-only, out of scope).
1294 Patch from Wolfgang Breyha, report from Stuart Northfield.
1296 PP/29 Fix three issues highlighted by clang analyser static analysis.
1297 Only crash-plausible issue would require the Cambridge-specific
1298 iplookup router and a misconfiguration.
1299 Report from Marcin Mirosław.
1301 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1303 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1304 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1305 As part of this, removing so much warning spew let me fix some minor
1306 real issues in debug logging.
1308 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1309 assignment on my part. Fixed.
1311 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1312 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1313 Janne Snabb (who went above and beyond: thank you).
1315 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1316 string otherwise requires a connection and a bunch more work and it's
1317 relatively easy to get wrong. Should also expose TLS library linkage
1320 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1321 64-bit ${eval} (JH/03).
1323 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1324 GNU libc to support some of the 64-bit stuff, should not lead to
1325 conflicts. Defined before os.h is pulled in, so if a given platform
1326 needs to override this, it can.
1328 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1329 protection layer was required, which is not implemented.
1330 Bugzilla 1254, patch from Wolfgang Breyha.
1332 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1333 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1334 tls_dhparam take prime identifiers. Also unbreak combination of
1335 OpenSSL+DH_params+TLSSNI.
1337 PP/39 Disable SSLv2 by default in OpenSSL support.
1343 PP/01 Solaris build fix for Oracle's LDAP libraries.
1344 Bugzilla 1109, patch from Stephen Usher.
1346 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1348 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1351 TF/02 Fix a couple more cases where we did not log the error message
1352 when unlink() failed. See also change 4.74-TF/03.
1354 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1355 lock up or crash if it happened to be inside a call to libc when it
1356 got a SIGUSR1 from exiwhat.
1358 The SIGUSR1 handler appends the current process status to the process
1359 log which is later printed by exiwhat. It used to use the general
1360 purpose logging code to do this, but several functions it calls are
1361 not safe for signals.
1363 The new output code in the SIGUSR1 handler is specific to the process
1364 log, and simple enough that it's easy to inspect for signal safety.
1365 Removing some special cases also simplifies the general logging code.
1366 Removing the spurious timestamps from the process log simplifies
1369 TF/04 Improved ratelimit ACL condition.
1371 The /noupdate option has been deprecated in favour of /readonly which
1372 has clearer semantics. The /leaky, /strict, and /readonly update modes
1373 are mutually exclusive. The update mode is no longer included in the
1374 database key; it just determines when the database is updated. (This
1375 means that when you upgrade Exim will forget old rate measurements.)
1377 Exim now checks that the per_* options are used with an update mode that
1378 makes sense for the current ACL. For example, when Exim is processing a
1379 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1380 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1381 must specify per_mail/readonly. If you omit the update mode it defaults to
1382 /leaky where that makes sense (as before) or /readonly where required.
1384 The /noupdate option is now undocumented but still supported for
1385 backwards compatibility. It is equivalent to /readonly except that in
1386 ACLs where /readonly is required you may specify /leaky/noupdate or
1387 /strict/noupdate which are treated the same as /readonly.
1389 A useful new feature is the /count= option. This is a generalization
1390 of the per_byte option, so that you can measure the throughput of other
1391 aggregate values. For example, the per_byte option is now equivalent
1392 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1394 The per_rcpt option has been generalized using the /count= mechanism
1395 (though it's more complicated than the per_byte equivalence). When it is
1396 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1397 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1398 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1399 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1400 Note that using per_rcpt with a non-readonly update mode in more than
1401 one ACL will cause the recipients to be double-counted. (The per_mail
1402 and per_byte options don't have this problem.)
1404 The handling of very low rates has changed slightly. If the computed rate
1405 is less than the event's count (usually one) then this event is the first
1406 after a long gap. In this case the rate is set to the same as this event's
1407 count, so that the first message of a spam run is counted properly.
1409 The major new feature is a mechanism for counting the rate of unique
1410 events. The new per_addr option counts the number of different
1411 recipients that someone has sent messages to in the last time period. It
1412 behaves like per_rcpt if all the recipient addresses are different, but
1413 duplicate recipient addresses do not increase the measured rate. Like
1414 the /count= option this is a general mechanism, so the per_addr option
1415 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1416 example, measure the rate that a client uses different sender addresses
1417 with the options per_mail/unique=$sender_address. There are further
1418 details in the main documentation.
1420 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1422 TF/06 Removed a few PCRE remnants.
1424 TF/07 Automatically extract Exim's version number from tags in the git
1425 repository when doing development or release builds.
1427 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1428 Bugzilla 879. Patch from Paul Fisher.
1430 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1431 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1434 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1436 PP/05 Variable $av_failed, true if the AV scanner deferred.
1437 Bugzilla 1078. Patch from John Horne.
1439 PP/06 Stop make process more reliably on build failure.
1440 Bugzilla 1087. Patch from Heiko Schlittermann.
1442 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1443 Bugzilla 1089. Patch from Heiko Schlittermann.
1445 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1446 Bugzilla 1131. Patch from Holger Weiß.
1448 PP/09 Handle IPv6 addresses with SPF.
1449 Bugzilla 860. Patch from Wolfgang Breyha.
1451 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1453 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1456 PP/11 match_* no longer expand right-hand-side by default.
1457 New compile-time build option, EXPAND_LISTMATCH_RHS.
1458 New expansion conditions, "inlist", "inlisti".
1460 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1462 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1464 PP/14 fix log_write() format string regression from TF/03.
1465 Bugzilla 1152. Patch from Dmitry Isaikin.
1471 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1473 PP/02 Harmonised TLS library version reporting; only show if debugging.
1474 Layout now matches that introduced for other libraries in 4.74 PP/03.
1476 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1478 PP/04 New "dns_use_edns0" global option.
1480 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1483 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1484 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1486 TK/01 Updated PolarSSL code to 0.14.2.
1487 Bugzilla 1097. Patch from Andreas Metzler.
1489 PP/07 Catch divide-by-zero in ${eval:...}.
1490 Fixes bugzilla 1102.
1492 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1495 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1496 format-string attack -- SECURITY: remote arbitrary code execution.
1498 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1499 time unintentionally subject to list matching rules, letting the header
1500 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1501 arbitrary string expansion). This allowed for information disclosure.
1503 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1504 INT_MIN/-1 -- value coerced to INT_MAX.
1510 NM/01 Workaround for PCRE version dependency in version reporting
1513 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1514 This fixes portability to compilers other than gcc, notably
1515 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1517 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1518 makefiles for portability to HP-UX and POSIX correctness.
1520 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1521 Also via indented variable definition in the Makefile.
1522 (Debugging by Oliver Heesakkers).
1524 PP/02 Restore caching of spamd results with expanded spamd_address.
1525 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1527 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1528 Improves build reliability. Fix from: Frank Elsner
1530 NM/02 Fix wide character breakage in the rfc2047 coding
1531 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1533 NM/03 Allow underscore in dnslist lookups
1534 Fixes bug 1026. Patch from Graeme Fowler
1536 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1537 Code patches from Adam Ciarcinski of NetBSD.
1539 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1542 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1543 is logged, to avoid truncation. Patch from John Horne.
1545 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1546 Patch from Jakob Hirsch.
1548 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1549 SQL string expansion failure details.
1550 Patch from Andrey Oktyabrski.
1552 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1553 Patch from Simon Arlott.
1555 PP/09 New lookups functionality failed to compile on old gcc which rejects
1556 extern declarations in function scope.
1557 Patch from Oliver Fleischmann
1559 PP/10 Use sig_atomic_t for flags set from signal handlers.
1560 Check getgroups() return and improve debugging.
1561 Fixed developed for diagnosis in bug 927 (which turned out to be
1564 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1565 Patch from Mark Zealey.
1567 PP/12 Bugzilla 1056: Improved spamd server selection.
1568 Patch from Mark Zealey.
1570 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1571 Based on patch from Heiko Schlittermann.
1573 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1574 Patch from Uwe Doering, sign-off by Michael Haardt.
1576 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1577 variable declaration deep within a block. Bug and patch from
1580 PP/15 lookups-Makefile IRIX compatibility coercion.
1582 PP/16 Make DISABLE_DKIM build knob functional.
1584 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1585 Patch by Simon Arlott
1587 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1588 variable argument macros. Our copy now differs from upstream.
1594 TF/01 Failure to get a lock on a hints database can have serious
1595 consequences so log it to the panic log.
1597 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1598 controlled using the smtp_confirmation log selector.
1600 TF/03 Include the error message when we fail to unlink a spool file.
1602 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1603 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1604 for maintaining out-of-tree patches for some time.
1606 PP/01 Bugzilla 139: Documentation and portability issues.
1607 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1608 Handle per-OS dynamic-module compilation flags.
1610 PP/02 Let /dev/null have normal permissions.
1611 The 4.73 fixes were a little too stringent and complained about the
1612 permissions on /dev/null. Exempt it from some checks.
1613 Reported by Andreas M. Kirchwitz.
1615 PP/03 Report version information for many libraries, including
1616 Exim version information for dynamically loaded libraries. Created
1617 version.h, now support a version extension string for distributors
1618 who patch heavily. Dynamic module ABI change.
1620 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1621 privilege escalation vulnerability whereby the Exim run-time user
1622 can cause root to append content of the attacker's choosing to
1625 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1628 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1629 If dropping privileges for untrusted macros, we disabled normal logging
1630 on the basis that it would fail; for the Exim run-time user, this is not
1631 the case, and it resulted in successful deliveries going unlogged.
1632 Fixed. Reported by Andreas Metzler.
1638 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1639 only prepend for the Resent-* case. Fixes regression introduced in
1640 Exim 4.70 by NM/22 for Bugzilla 607.
1642 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1643 increasing numbers of administrators be bitten by this.
1645 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1647 PP/03 Bugzilla 994: added openssl_options main configuration option.
1649 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1651 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1653 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1655 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1656 without a peer certificate, leading to a segfault because of an
1657 assumption that peers always have certificates. Be a little more
1658 paranoid. Problem reported by Martin Tscholak.
1660 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1661 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1662 NB: ClamAV planning to remove STREAM in "middle of 2010".
1663 CL also introduces -bmalware, various -d+acl logging additions and
1664 more caution in buffer sizes.
1666 PP/09 Implemented reverse_ip expansion operator.
1668 PP/10 Bugzilla 937: provide a "debug" ACL control.
1670 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1672 PP/12 Bugzilla 973: Implement --version.
1674 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1676 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1678 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1680 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1681 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1682 ignore trailing whitespace.
1684 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1686 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1689 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1690 Notification from Dr Andrew Aitchison.
1692 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1693 ExtendedDetectionInfo response format.
1694 Notification from John Horne.
1696 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1699 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1700 XSL and documented dependency on system catalogs, with examples of how
1703 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1706 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1707 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1708 configuration file which is writeable by the Exim user or group.
1710 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1711 of configuration files to cover files specified with the -C option if
1712 they are going to be used with root privileges, not just the default
1715 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1716 option (effectively making it always true).
1718 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1719 files to be used while preserving root privileges.
1721 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1722 that rogue child processes cannot use them.
1724 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1725 run-time user, instead of root.
1727 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1728 Exim run-time user without dropping privileges.
1730 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1731 result string, instead of calling string_vformat() twice with the same
1734 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1735 for other users. Others should always drop root privileges if they use
1736 -C on the command line, even for a whitelisted configure file.
1738 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1740 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1746 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1747 $data_path, and $header_path variables; fixed documentation bugs and
1750 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1751 exipick to access non-standard spools, including the "frozen" queue
1754 NM/01 Bugzilla 965: Support mysql stored procedures.
1755 Patch from Alain Williams
1757 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1759 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1760 Patch from Andreas Metzler
1762 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1763 Patch from Kirill Miazine
1765 NM/05 Bugzilla 671: Added umask to procmail example.
1767 JJ/03 installed exipick 20100323.0, fixing doc bug
1769 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1770 directory. Notification and patch from Dan Rosenberg.
1772 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1774 TK/02 Improve log output when DKIM signing operation fails.
1776 MH/01 Treat the transport option dkim_domain as a colon separated
1777 list, not as a single string, and sign the message with each element,
1778 omitting multiple occurences of the same signer.
1780 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1781 Bugzilla 985, 986. Patch by Simon Arlott
1783 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1784 Patch by Simon Arlott
1786 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1787 MBX locking. Notification from Dan Rosenberg.
1793 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1795 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1797 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1799 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1801 NM/04 Bugzilla 915: Flag broken perl installation during build.
1807 TK/01 Added patch by Johannes Berg that expands the main option
1808 "spamd_address" if it starts with a dollar sign.
1810 TK/02 Write list of recipients to X-Envelope-Sender header when building
1811 the mbox-format spool file for content scanning (suggested by Jakob
1814 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1815 (http://www.dcc-servers.net/) support via dccifd. Activated by
1816 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1818 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1819 by Mark Daniel Reidel <mr@df.eu>.
1821 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1822 When building exim an external PCRE library is now needed -
1823 PCRE is a system library on the majority of modern systems.
1824 See entry on PCRE_LIBS in EDITME file.
1826 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1827 conversation. Added nologin parameter to request.
1828 Patch contributed by Kirill Miazine.
1830 TF/01 Do not log submission mode rewrites if they do not change the address.
1832 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1834 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1835 log files in place. Contributed by Roberto Lima.
1837 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1839 TF/03 Bugzilla 615: When checking the local_parts router precondition
1840 after a local_part_suffix or local_part_prefix option, Exim now
1841 does not use the address's named list lookup cache, since this
1842 contains cached lookups for the whole local part.
1844 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1845 Robert Millan. Documentation is in experimental-spec.txt.
1847 TF/04 Bugzilla 668: Fix parallel build (make -j).
1849 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1851 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1852 Patch provided by Jan Srzednicki.
1854 TF/05 Leading white space used to be stripped from $spam_report which
1855 wrecked the formatting. Now it is preserved.
1857 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1858 that they are available at delivery time.
1860 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1862 TF/08 TLS error reporting now respects the incoming_interface and
1863 incoming_port log selectors.
1865 TF/09 Produce a more useful error message if an SMTP transport's hosts
1866 setting expands to an empty string.
1868 NM/06 Bugzilla 744: EXPN did not work under TLS.
1869 Patch provided by Phil Pennock.
1871 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1872 Patch provided by Richard Godbee.
1874 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1875 acl_smtp_notquit, added index entry.
1877 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1878 Patch provided by Eugene Bujak.
1880 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1881 accept(). Patch provided by Maxim Dounin.
1883 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1884 Patch provided by Phil Pennock.
1886 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1888 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1889 Patch provided by Brad "anomie" Jorsch.
1891 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1892 Patch provided by Dean Brooks.
1894 TK/05 Add native DKIM support (does not depend on external libraries).
1896 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1897 Patch provided by Graeme Fowler.
1899 NM/16 Bugzilla 851: Documentation example syntax fix.
1901 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1903 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1906 NM/19 Bugzilla 745: TLS version reporting.
1907 Patch provided by Phil Pennock.
1909 NM/20 Bugzilla 167: bool: condition support.
1910 Patch provided by Phil Pennock.
1912 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1913 clients. Patch provided by Phil Pennock.
1915 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1916 Patch provided by Brad "anomie" Jorsch.
1918 NM/23 Bugzilla 687: Fix misparses in eximstats.
1919 Patch provided by Heiko Schlittermann.
1921 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1922 Patch provided by Heiko Schlittermann.
1924 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1925 plus update to original patch.
1927 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1929 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1930 Patch provided by David Brownlee.
1932 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1934 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1936 NM/30 Bugzilla 888: TLS documentation bugfixes.
1938 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1940 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1941 Unlike the original bugzilla I have changed all shell scripts in src tree.
1943 NM/33 Bugzilla 898: Transport filter timeout fix.
1944 Patch by Todd Rinaldo.
1946 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1947 Patch by Serge Demonchaux.
1949 NM/35 Bugzilla 39: Base64 decode bug fixes.
1950 Patch by Jakob Hirsch.
1952 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1954 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1956 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1958 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1964 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1965 ALT-N's libdkim that I have put here:
1966 http://duncanthrax.net/exim-experimental/
1968 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1969 were called 'true' and it seems that C99 defines that as a reserved
1970 keyword to be used with 'bool' variable types. That means you could
1971 not include C99-style headers which use bools without triggering
1972 build errors in sieve.c.
1974 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1975 as mailq or other aliases. Changed the --help handling significantly
1976 to do whats expected. exim_usage() emits usage/help information.
1978 SC/01 Added the -bylocaldomain option to eximstats.
1980 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1982 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1984 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1990 PH/01 Another patch from the Sieve maintainer.
1992 PH/02 When an IPv6 address is converted to a string for single-key lookup
1993 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1994 dots are used instead of colons so that keys in lsearch files need not
1995 contain colons. This was done some time before quoting was made available
1996 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1997 (notated using the quote facility) so as to distinguish them from IPv4
1998 keys. This meant that lookups for IP addresses in host lists did not work
1999 for iplsearch lookups.
2001 This has been fixed by arranging for IPv6 addresses to be expressed with
2002 colons if the lookup type is iplsearch. This is not incompatible, because
2003 previously such lookups could never work.
2005 The situation is now rather anomalous, since one *can* have colons in
2006 ordinary lsearch keys. However, making the change in all cases is
2007 incompatible and would probably break a number of configurations.
2009 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2012 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2013 conversion specification without a maximum field width, thereby enabling
2014 a rogue spamd server to cause a buffer overflow. While nobody in their
2015 right mind would setup Exim to query an untrusted spamd server, an
2016 attacker that gains access to a server running spamd could potentially
2017 exploit this vulnerability to run arbitrary code as the Exim user.
2019 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2020 $primary_hostname instead of what libspf2 thinks the hosts name is.
2022 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2023 a directory entry by the name of the lookup key. Previously, if a
2024 symlink pointed to a non-existing file or a file in a directory that
2025 Exim lacked permissions to read, a lookup for a key matching that
2026 symlink would fail. Now it is enough that a matching directory entry
2027 exists, symlink or not. (Bugzilla 503.)
2029 PH/03 The body_linecount and body_zerocount variables are now exported in the
2032 PH/04 Added the $dnslist_matched variable.
2034 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2035 This means they are set thereafter only if the connection becomes
2038 PH/06 Added the client_condition to authenticators so that some can be skipped
2039 by clients under certain conditions.
2041 PH/07 The error message for a badly-placed control=no_multiline_responses left
2042 "_responses" off the end of the name.
2044 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2046 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2047 (without spaces) instead of just copying the configuration text.
2049 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2051 PH/11 Added $max_received_linelength.
2053 PH/12 Added +ignore_defer and +include_defer to host lists.
2055 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2056 way in which PCRE > 7.0 is built.
2058 PH/14 Implemented queue_only_load_latch.
2060 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2061 MAIL command. The effect was to mangle the value on 64-bit systems.
2063 PH/16 Another patch from the Sieve maintainer.
2065 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2067 PH/18 If a system quota error occurred while trying to create the file for
2068 a maildir delivery, the message "Mailbox is full" was not appended to the
2069 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2070 only to a quota excession during the actual writing of the file.
2072 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2073 characters?) which causes problems in log lines. The DN values are now
2074 passed through string_printing() before being added to log lines.
2076 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2077 and InterBase are left for another time.)
2079 PH/21 Added message_body_newlines option.
2081 PH/22 Guard against possible overflow in moan_check_errorcopy().
2083 PH/23 POSIX allows open() to be a macro; guard against that.
2085 PH/24 If the recipient of an error message contained an @ in the local part
2086 (suitably quoted, of course), incorrect values were put in $domain and
2087 $local_part during the evaluation of errors_copy.
2093 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2094 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2097 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2098 issue a MAIL command.
2100 PH/02 In an ACL statement such as
2102 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2104 if a client was not listed at all, or was listed with a value other than
2105 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2106 the condition was not true (as it should be), so access was not denied.
2107 The bug was that the ! inversion was incorrectly passed on to the second
2108 item. This has been fixed.
2110 PH/03 Added additional dnslists conditions == and =& which are different from
2111 = and & when the dns lookup returns more than one IP address.
2113 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2114 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2116 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2117 FSYNC, which compiles an option called disable_fsync that allows for
2118 bypassing fsync(). The documentation is heavily laced with warnings.
2120 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2122 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2123 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2124 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2125 including adding "make clean"; (3) Added -fPIC when compiling the test
2126 dynamically loaded module, to get rid of a warning.
2128 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2129 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2130 = 0s. The bug is otherwise harmless.
2132 PH/07 There was a bug in the dovecot authenticator such that the value of
2133 $auth1 could be overwritten, and so not correctly preserved, after a
2134 successful authentication. This usually meant that the value preserved by
2135 the server_setid option was incorrect.
2137 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2139 PH/09 Installed PCRE release 7.0.
2141 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2142 run for batched SMTP input. It is now run at the start of every message
2143 in the batch. While fixing this I discovered that the process information
2144 (output by running exiwhat) was not always getting set for -bs and -bS
2145 input. This is fixed, and it now also says "batched" for BSMTP.
2147 PH/11 Added control=no_pipelining.
2149 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2150 patch, slightly modified), and move the expansion of helo_data till after
2151 the connection is made in the smtp transport (so it can use these
2154 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2156 PH/14 Added log_selector = +pid.
2158 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2160 PH/16 Add ${if forany and ${if forall.
2162 PH/17 Added dsn_from option to vary the From: line in DSNs.
2164 PH/18 Flush SMTP output before performing a callout, unless control =
2165 no_callout_flush is set.
2167 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2168 was true (the default) a successful delivery failed to delete the retry
2169 item, thus causing premature timeout of the address. The bug is now
2172 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2174 PH/21 Long custom messages for fakedefer and fakereject are now split up
2175 into multiline responses in the same way that messages for "deny" and
2176 other ACL rejections are.
2178 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2179 with slight modification.
2181 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2182 draft, changing the syntax and factoring some duplicate code".
2184 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2185 for deliveries of the second and subsequent messages over the same SMTP
2188 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2189 ${reduce, with only minor "tidies".
2191 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2193 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2194 expansion side effects.
2196 PH/27 When a message times out after an over-quota error from an Exim-imposed
2197 quota, the bounce message says "mailbox is full". This message was not
2198 being given when it was a system quota that was exceeded. It now should
2201 MH/03 Made $recipients available in local_scan(). local_scan() already has
2202 better access to the recipient list through recipients_list[], but
2203 $recipients can be useful in postmaster-provided expansion strings.
2205 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2206 in the case of a MAIL command with additional options following the
2207 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2208 were accidentally chopped off.
2210 PH/29 SMTP synchronization checks are implemented when a command is read -
2211 there is a check that no more input is waiting when there shouldn't be
2212 any. However, for some commands, a delay in an ACL can mean that it is
2213 some time before the response is written. In this time, more input might
2214 arrive, invalidly. So now there are extra checks after an ACL has run for
2215 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2216 pipelining has not been advertised.
2218 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2220 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2221 returned in the SMTP response when smtp_return_error_details was false.
2222 This has been fixed.
2224 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2225 instead of the C I/O that was originally supplied, because problems were
2226 reported on Solaris.
2228 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2229 Exim which did not show up earlier: it was assuming that a call to
2230 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2231 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2232 was a macro that became an assignment, so it seemed to work. This has
2233 changed to a proper function call with a void return, hence the compile
2234 error. Exim's code has been fixed.
2236 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2239 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2241 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2243 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2244 because it turns out to be too big or there is a timeout) there is no
2245 "Completed" line in the log. When some messages of this type were
2246 selected by exigrep, they were listed as "not completed". Others were
2247 picked up by some special patterns. I have improved the selection
2248 criteria to be more general.
2250 PH/38 The host_find_failed option in the manualroute router can now be set
2251 to "ignore", to completely ignore a host whose IP address cannot be
2252 found. If all hosts are ignored, the behaviour is controlled by the new
2253 host_all_ignored option.
2255 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2256 homing or because of multiple MX records with /mx) generated more than
2257 one IP address, and the following item turned out to be the local host,
2258 all the secondary addresses of the first item were incorrectly removed
2259 from the list, along with the local host and any following hosts (which
2260 is what is supposed to happen).
2262 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2263 whoever called Exim into the -H file. In the case of the daemon it was
2264 behaving confusingly. When first started, it used values for whoever
2265 started the daemon, but after a SIGHUP it used the Exim user (because it
2266 calls itself on a restart). I have changed the code so that it now always
2269 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2270 message are rejected with the same error (e.g. no authentication or bad
2271 sender address), and a DATA command is nevertheless sent (as can happen
2272 with PIPELINING or a stupid MUA), the error message that was given to the
2273 RCPT commands is included in the rejection of the DATA command. This is
2274 intended to be helpful for MUAs that show only the final error to their
2277 PH/42 Another patch from the Sieve maintainer.
2279 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2280 Eximstats - Fixed some broken HTML links and added missing column headers
2282 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2283 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2285 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2287 PH/43 Yet another patch from the Sieve maintainer.
2289 PH/44 I found a way to check for a TCP/IP connection going away before sending
2290 the response to the final '.' that terminates a message, but only in the
2291 case where the client has not sent further data following the '.'
2292 (unfortunately, this is allowed). However, in many cases there won't be
2293 any further data because there won't be any more messages to send. A call
2294 to select() can be used: if it shows that the input is "ready", there is
2295 either input waiting, or the socket has been closed. An attempt to read
2296 the next input character can distinguish the two cases. Previously, Exim
2297 would have sent an OK response which the client would never have see.
2298 This could lead to message repetition. This fix should cure that, at
2299 least in a lot of common cases.
2301 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2302 advertised in response to EHLO.
2308 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2309 fixed by 4.65/MH/01 (is this a record?) are fixed:
2311 (i) An empty string was always treated as zero by the numeric comparison
2312 operators. This behaviour has been restored.
2314 (ii) It is documented that the numeric comparison operators always treat
2315 their arguments as decimal numbers. This was broken in that numbers
2316 starting with 0 were being interpreted as octal.
2318 While fixing these problems I realized that there was another issue that
2319 hadn't been noticed. Values of message_size_limit (both the global option
2320 and the transport option) were treated as octal if they started with 0.
2321 The documentation was vague. These values are now always treated as
2322 decimal, and I will make that clear in the documentation.
2328 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2329 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2332 MH/01 Don't check that the operands of numeric comparison operators are
2333 integers when their expansion is in "skipping" mode (fixes bug
2334 introduced by 4.64-PH/07).
2336 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2337 child addresses, Exim now panics and dies. Previously, because the count
2338 is held in a short int, deliveries were likely to be lost. As such a
2339 large number of recipients for a single message is ridiculous
2340 (performance will be very, very poor), I have chosen to impose a limit
2341 rather than extend the field.
2347 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2348 leftover -K file (the existence of which was triggered by #402).
2349 While we were at it, introduced process PID as part of the -K
2350 filename. This should rule out race conditions when creating
2353 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2354 processing considerably. Previous code took too long for large mails,
2355 triggering a timeout which in turn triggers #401.
2357 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2358 in the DK code in transports.c. sendfile() is not really portable,
2359 hence the _LINUX specificness.
2361 TF/01 In the add_headers option to the mail command in an Exim filter,
2362 there was a bug that Exim would claim a syntax error in any
2363 header after the first one which had an odd number of characters
2366 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2367 callout verification, Exim cached a "reject" for the entire domain. This
2368 is correct for most verifications, but it is not correct for a recipient
2369 verification with use_sender or use_postmaster set, because in that case
2370 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2371 case of MAIL FROM:<> rejection from other early rejections (e.g.
2372 rejection of HELO). When verifying a recipient using a non-null MAIL
2373 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2374 Whatever the result of the callout, the value of the domain cache is
2375 left unchanged (for any other kind of callout, getting as far as trying
2376 RCPT means that the domain itself is ok).
2378 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2381 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2382 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2383 session, the daemon ignores ECONNECT errors and logs others; it now
2384 ignores EPIPE as well.
2386 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2387 (quoted-printable decoding).
2389 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2390 later the small subsequent patch to fix an introduced bug.
2392 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2394 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2396 PH/08 An error is now given if message_size_limit is specified negative.
2398 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2399 to be given (somewhat) arbitrary names.
2401 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2404 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2405 miscellaneous code fixes
2407 PH/10 Added the log_reject_target ACL modifier to specify where to log
2410 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2411 hostname. This is wrong, because it relates to the incoming message (and
2412 probably the interface on which it is arriving) and not to the outgoing
2413 callout (which could be using a different interface). This has been
2414 changed to use the value of the helo_data option from the smtp transport
2415 instead - this is what is used when a message is actually being sent. If
2416 there is no remote transport (possible with a router that sets up host
2417 addresses), $smtp_active_hostname is used.
2419 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2420 tweaks were necessary in order to get it to work (see also 21 below):
2421 (a) The code assumed that strncpy() returns a negative number on buffer
2422 overflow, which isn't the case. Replaced with Exim's string_format()
2424 (b) There were several signed/unsigned issues. I just did the minimum
2425 hacking in of casts. There is scope for a larger refactoring.
2426 (c) The code used strcasecmp() which is not a standard C function.
2427 Replaced with Exim's strcmpic() function.
2428 (d) The code set only $1; it now sets $auth1 as well.
2429 (e) A simple test gave the error "authentication client didn't specify
2430 service in request". It would seem that Dovecot has changed its
2431 interface. Fortunately there's a specification; I followed it and
2432 changed what the client sends and it appears to be working now.
2434 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2437 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2438 address A is aliased to B and C, where B exists and C does not. Without
2439 -v the output is "A verified" because verification stops after a
2440 successful redirection if more than one address is generated. However,
2441 with -v the child addresses are also verified. Exim was outputting "A
2442 failed to verify" and then showing the successful verification for C,
2443 with its parentage. It now outputs "B failed to verify", showing B's
2444 parentage before showing the successful verification of C.
2446 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2447 look up a TXT record in a specific list after matching in a combined
2450 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2451 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2452 they consult the DNS. I had assumed they would set it the way they
2453 wanted; and indeed my experiments on Linux seem to show that in some
2454 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2455 To be on the safe side, however, I have now made the interface to
2456 host_find_byname() similar to host_find_bydns(), with an argument
2457 containing the DNS resolver options. The host_find_byname() function now
2458 sets these options at its start, just as host_find_bydns() does. The smtp
2459 transport options dns_qualify_single and dns_search_parents are passed to
2460 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2461 of host_find_byname() use the default settings of RES_DEFNAMES
2462 (qualify_single) but not RES_DNSRCH (search_parents).
2464 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2465 spool_read_header() do less string testing, by means of a preliminary
2466 switch on the second character of optional "-foo" lines. (This is
2467 overdue, caused by the large number of possibilities that now exist.
2468 Originally there were few.) While I was there, I also converted the
2469 str(n)cmp tests so they don't re-test the leading "-" and the first
2470 character, in the hope this might squeeze out yet more improvement.
2472 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2473 flag allowing group syntax was set by the header_syntax check but not
2474 turned off, possible causing trouble later; (2) The flag was not being
2475 set at all for the header_verify test, causing "group"-style headers to
2476 be rejected. I have now set it in this case, and also caused header_
2477 verify to ignore an empty address taken from a group. While doing this, I
2478 came across some other cases where the code for allowing group syntax
2479 while scanning a header line wasn't quite right (mostly, not resetting
2480 the flag correctly in the right place). These bugs could have caused
2481 trouble for malformed header lines. I hope it is now all correct.
2483 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2484 with the "reply" argument non-NULL. The code, however (which originally
2485 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2486 but it didn't always do it. This confused somebody who was copying the
2487 code for some other use. I have removed all the tests.
2489 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2490 feature that was used to support insecure browsers during the U.S. crypto
2491 embargo. It requires special client support, and Exim is probably the
2492 only MTA that supported it -- and would never use it because real RSA is
2493 always available. This code has been removed, because it had the bad
2494 effect of slowing Exim down by computing (never used) parameters for the
2495 RSA_EXPORT functionality.
2497 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2498 authenticator to fail if there's a tab character in the incoming data
2499 (there should never be unless someone is messing about, as it's supposed
2500 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2501 if the connection is using TLS or if the remote IP is the same as the
2502 local IP, and the "valid-client-cert option" if a client certificate has
2505 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2506 authenticators. This can be used for authorization after authentication
2507 succeeds. (In the case of plaintext, it servers for both authentication
2510 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2511 if any retry times were supplied.
2513 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2514 connection, where there is no client IP address to check. In this
2515 situation, the verify now always succeeds.
2517 PH/25 Applied John Jetmore's -Mset patch.
2519 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2521 PH/27 In a string expansion for a processed (not raw) header when multiple
2522 headers of the same name were present, leading whitespace was being
2523 removed from all of them, but trailing whitespace was being removed only
2524 from the last one. Now trailing whitespace is removed from each header
2525 before concatenation. Completely empty headers in a concatenation (as
2526 before) are ignored.
2528 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2529 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2531 PH/29 [Removed. This was a change that I later backed out, and forgot to
2532 correct the ChangeLog entry (that I had efficiently created) before
2533 committing the later change.]
2535 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2536 address errors (4xx response to RCPT) over the same connection as other
2537 messages routed to the same hosts. Such deliveries are always "forced",
2538 so retry times are not inspected. This resulted in far too many retries
2539 for the affected addresses. The effect occurred only when there were more
2540 hosts than the hosts_max_try setting in the smtp transport when it had
2541 the 4xx errors. Those hosts that it had tried were not added to the list
2542 of hosts for which the message was waiting, so if all were tried, there
2543 was no problem. Two fixes have been applied:
2545 (i) If there are any address or message errors in an SMTP delivery, none
2546 of the hosts (tried or untried) are now added to the list of hosts
2547 for which the message is waiting, so the message should not be a
2548 candidate for sending over the same connection that was used for a
2549 successful delivery of some other message. This seems entirely
2550 reasonable: after all the message is NOT "waiting for some host".
2551 This is so "obvious" that I'm not sure why it wasn't done
2552 previously. Hope I haven't missed anything, but it can't do any
2553 harm, as the worst effect is to miss an optimization.
2555 (ii) If, despite (i), such a delivery is accidentally attempted, the
2556 routing retry time is respected, so at least it doesn't keep
2557 hammering the server.
2559 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2560 in ${readsocket because some servers need this prod.
2562 PH/32 Added some extra debug output when updating a wait-xxx database.
2564 PH/33 The hint "could be header name not terminated by colon", which has been
2565 given for certain expansion errors for a long time, was not being given
2566 for the ${if def:h_colon_omitted{... case.
2568 PH/34 The spec says: "With one important exception, whenever a domain list is
2569 being scanned, $domain contains the subject domain." There was at least
2570 one case where this was not true.
2572 PH/35 The error "getsockname() failed: connection reset by peer" was being
2573 written to the panic log as well as the main log, but it isn't really
2574 panic-worthy as it just means the connection died rather early on. I have
2575 removed the panic log writing for the ECONNRESET error when getsockname()
2578 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2579 runs only) independently of the message's sender address. This meant
2580 that, if the 4xx error was in fact related to the sender, a different
2581 message to the same recipient with a different sender could confuse
2582 things. In particular, this can happen when sending to a greylisting
2583 server, but other circumstances could also provoke similar problems.
2584 I have changed the default so that the retry time for these errors is now
2585 based a combination of the sender and recipient addresses. This change
2586 can be overridden by setting address_retry_include_sender=false in the
2589 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2590 remote server are returned as part of bounce messages. This was not
2591 happening for LMTP over a pipe (the lmtp transport), but now it is the
2592 same for both kinds of LMTP.
2594 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2595 in header lines that were in fact CNAMEs. This is no longer the case.
2597 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2598 and queue runs started by the daemon processed all messages. This has
2599 been fixed so that -R and -S can now usefully be given with -q<time>.
2601 PH/40 Import PCRE release 6.7 (fixes some bugs).
2603 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2605 PH/42 Give an error if -q is specified more than once.
2607 PH/43 Renamed the variables $interface_address and $interface_port as
2608 $received_ip_address and $received_port, to make it clear that these
2609 values apply to message reception, and not to the outgoing interface when
2610 a message is delivered. (The old names remain recognized, of course.)
2612 PH/44 There was no timeout on the connect() call when using a Unix domain
2613 socket in the ${readsocket expansion. There now is.
2615 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2616 be meaningful with "accept".
2618 SC/01 Eximstats V1.43
2619 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2621 SC/02 Eximstats V1.44
2622 Use a glob alias rather than an array ref in the generated
2623 parser. This improves both readability and performance.
2625 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2626 Collect SpamAssassin and rejection statistics.
2627 Don't display local sender or destination tables unless
2628 there is data to show.
2629 Added average volumes into the top table text output.
2631 SC/04 Eximstats V1.46
2632 Collect data on the number of addresses (recipients)
2633 as well as the number of messages.
2635 SC/05 Eximstats V1.47
2636 Added 'Message too big' to the list of mail rejection
2637 reasons (thanks to Marco Gaiarin).
2639 SC/06 Eximstats V1.48
2640 Mainlog lines which have GMT offsets and are too short to
2641 have a flag are now skipped.
2643 SC/07 Eximstats V1.49 (Alain Williams)
2644 Added the -emptyok flag.
2646 SC/08 Eximstats V1.50
2647 Fixes for obtaining the IP address from reject messages.
2649 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2650 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2651 whitespace changes from 4.64-PH/27
2653 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2656 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2657 are found, allow negative numbers in numeric criteria)
2659 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2661 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2662 to match changes made in 4.64-PH/43
2664 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2666 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2667 "domain" is actually an IP address, and force a failure. This locks out
2668 those revolvers/nameservers that support "A-for-A" lookups, in
2669 contravention of the specifications.
2671 PH/48 When a host name was looked up from an IP address, and the subsequent
2672 forward lookup of the name timed out, the host name was left in
2673 $sender_host_name, contrary to the specification.
2675 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2676 restricted to single-key lookups, Exim was not diagnosing an error if
2677 * or *@ was used with a query-style lookup.
2679 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2681 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2682 long ago, but noone interested enough thought of it. Let's just say that
2683 the "1.1" means that there are some new functions that weren't there at
2684 some point in the past.
2686 PH/51 Error processing for expansion failure of helo_data from an smtp
2687 transport during callout processing was broken.
2689 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2690 tested/used via the -bh/-bhc/-bs options.
2692 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2693 bug, fixed in subsequent PCRE releases).
2695 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2696 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2698 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2704 SC/01 Use a glob alias rather than an array ref in eximstats generated
2705 parser. This improves both readability and performance.
2707 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2708 Don't display local sender or destination tables in eximstats unless
2709 there is data to show.
2710 Added average volumes into the eximstats top table text output.
2712 SC/03 Collect data on the number of addresses (recipients) as well
2713 as the number of messages in eximstats.
2715 TF/01 Correct an error in the documentation for the redirect router. Exim
2716 does (usually) call initgroups() when daemonizing.
2718 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2719 with consistent privilege compared to when running as a daemon.
2721 TF/03 Note in the spec that $authenticated_id is not set for local
2722 submissions from trusted users.
2724 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2725 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2727 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2728 by adding some example configuration directives to the default
2729 configuration file. A little bit of work is required to uncomment the
2730 directives and define how usernames and passwords are checked, but
2731 there is now a framework to start from.
2733 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2734 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2735 without this. I don't know how relevant this is to other LDAP libraries.
2737 PH/02 Add the verb name to the "unknown ACL verb" error.
2739 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2741 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2743 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2744 directory not expanded when it should be if an expanded home directory
2745 was set for the address (which is overridden by the transport).
2747 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2750 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2751 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2752 because it is too late at that time, and has no effect.
2754 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2755 security issue with \' (bugzilla #107). I could not use the
2756 PQescapeStringConn() function, because it needs a PGconn value as one of
2759 PH/08 When testing addresses using -bt, indicate those final addresses that
2760 are duplicates that would not cause an additional delivery. At least one
2761 person was confused, thinking that -bt output corresponded to deliveries.
2762 (Suppressing duplicates isn't a good idea as you lose the information
2763 about possibly different redirections that led to the duplicates.)
2765 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2766 systems where poll() doesn't work, in particular OS X.
2768 PH/10 Added more information to debugging output for retry time not reached.
2770 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2771 operations in malware.c.
2773 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2776 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2777 syslog_duplication set false, log lines that would normally be written
2778 both the the main log and to the reject log were not written to syslog at
2781 PH/14 In the default configuration, change the use of "message" in ACL warn
2782 statements to "add_header".
2784 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2785 not followed by a command (e.g. "seen endif").
2787 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2788 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2791 PH/17 Added extra conditions to the default value of delay_warning_condition
2795 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2796 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2797 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2800 The Auto-Submitted: and various List- headers are standardised, whereas I
2801 don't think Precedence: ever was.
2803 PH/18 Refactored debugging code in route_finduser() to show more information,
2804 in particular, the error code if getpwnam() issues one.
2806 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2807 This is apparently needed in addition to the PH/07 change above to avoid
2808 any possible encoding problems.
2810 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2811 but not after initializing Perl.
2813 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2814 output them only if debugging. By default they are written stderr,
2815 apparently, which is not desirable.
2817 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2820 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2823 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2825 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2826 authenticated or an ident call has been made. Suppress the default
2827 values for $authenticated_id and $authenticated_sender (but permit -oMai
2828 and -oMas) when testing with -bh.
2830 PH/24 Re-jigged the order of the tests in the default configuration so that the
2831 tests for valid domains and recipients precede the DNS black list and CSA
2832 tests, on the grounds that those ones are more expensive.
2834 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2835 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2836 This bug exists in every version of Exim that I still have, right back to
2839 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2840 However, an attempt to turn on case-sensitivity in a regex key by
2841 including (?-i) didn't work because the subject string was already
2842 lowercased, and the effects were non-intuitive. It turns out that a
2843 one-line patch can be used to allow (?-i) to work as expected.
2849 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2850 other effects) broke the use of negated acl sub-conditions.
2852 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2855 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2856 "Deny" causes Exim to reject the incoming connection with a 554 error.
2857 Unfortunately, if there is a major crisis, such as a disk failure,
2858 tcp-wrappers gives "deny", whereas what one would like would be some
2859 kind of temporary error. A kludge has been added to help with this.
2860 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2861 554 error is used if errno is still zero or contains ENOENT (which occurs
2862 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2865 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2867 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2868 errors. Otherwise a message that provokes a temporary error (when other
2869 messages do not) can cause a whole host to time out.
2871 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2872 addresses were routed directly to files or pipes from a redirect router.
2873 File deliveries just didn't batch; pipe deliveries might have suffered
2876 PH/06 A failure to get a lock for a hints database would erroneously always say
2877 "Failed to get write lock", even when it was really a read lock.
2879 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2880 of 0600. This has been changed to use the value of the lockfile_mode
2881 option (which defaults to 0600).
2883 PH/08 Applied small patch from the Sieve maintainer.
2885 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2886 folder from quota calculations, a direct delivery into this folder messed
2887 up the contents of the maildirsize file. This was because the regex was
2888 used only to exclude .Trash (or whatever) when the size of the mailbox
2889 was calculated. There was no check that a delivery was happening into an
2890 excluded directory. This bug has been fixed by ignoring all quota
2891 processing for deliveries into excluded directories.
2893 PH/10 Added the maildirfolder_create_regex option to appendfile.
2899 PH/01 The code for finding all the local interface addresses on a FreeBSD
2900 system running IPv6 was broken. This may well have applied to all BSD
2901 systems, as well as to others that have similar system calls. The broken
2902 code found IPv4 interfaces correctly, but gave incorrect values for the
2903 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2904 that it would not match correctly against @[] and not recognize the IPv6
2907 PH/02 The ipliteral router was not recognizing addresses of the form user@
2908 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2910 PH/03 Added disable_ipv6.
2912 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2913 decoded form, because it is most often used to construct To: headers
2914 lines in autoreplies, and the decoded form may well be syntactically
2915 invalid. However, $reply_address has leading white space removed, and all
2916 newlines turned into spaces so that the autoreply transport does not
2919 PH/05 If group was specified without a user on a router, and no group or user
2920 was specified on a transport, the group from the router was ignored.
2922 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2923 for visible compile-time settings that can be used to change these
2924 numbers, for those that want even more. Backwards compatibility with old
2925 spool files has been maintained. However, going back to a previous Exim
2926 release will lost any variables that are in spool files.
2928 PH/07 Two small changes when running in the test harness: increase delay when
2929 passing a TCP/IP connection to a new process, in case the original
2930 process has to generate a bounce, and remove special handling of
2931 127.0.0.2 (sic), which is no longer necessary.
2933 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2934 be the same on different OS.
2936 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2939 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2940 whether --show-vars was specified or not
2942 JJ/02 exipick: Added support for new ACL variable spool format introduced
2945 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2946 syntactically invalid From: or Reply-to: line, and a filter used this to
2947 generate an autoreply, and therefore failed to obtain an address for the
2948 autoreply, Exim could try to deliver to a non-existent relative file
2949 name, causing unrelated and misleading errors. What now happens is that
2950 it logs this as a hard delivery error, but does not attempt to create a
2953 PH/11 The exinext utility has a -C option for testing purposes, but although
2954 the given file was scanned by exinext itself; it wasn't being passed on
2955 when Exim was called.
2957 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2958 an end-of-file indication when reading a command response.
2960 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2961 compiled. In many other places in Exim, IPv6 addresses are always
2962 recognized, so I have changed this. It also means that IPv4 domain
2963 literals of the form [IPV4:n.n.n.n] are now always recognized.
2965 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2966 used if the router is not running as root, for example, when verifying at
2967 ACL time, or when using -bh. The debugging output from this situation was
2968 non-existent - all you got was a failure to exec. I have made two
2971 (a) Failures to set uid/gid, the current directory, or a process leader
2972 in a subprocess such as that created by queryprogram now generate
2973 suitable debugging output when -d is set.
2975 (b) The queryprogram router detects when it is not running as root,
2976 outputs suitable debugging information if -d is set, and then runs
2977 the subprocess without attempting to change uid/gid.
2979 PH/15 Minor change to Makefile for building test_host (undocumented testing
2982 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2983 additional section of a DNS packet that returns MX or SRV records.
2984 Instead, it always explicitly searches for A/AAAA records. This avoids
2985 major problems that occur when a DNS server includes only records of one
2986 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2987 fixed another bug: if SRV records were looked up and the corresponding
2988 address records were *not* found in the additional section, the port
2989 values from the SRV records were lost.
2991 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2992 using the correct key (the original address) when searching the retry
2993 rules in order to find which one to use for generating the retry hint.
2995 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2996 adding the default one. Similarly, if it contains a Reply-To: header, the
2997 errors_reply_to option, if set, is not used.
2999 PH/19 When calculating a retry time, Exim used to measure the "time since
3000 failure" by looking at the "first failed" field in the retry record. Now
3001 it does not use this if it is later than than the arrival time of the
3002 message. Instead it uses the arrival time. This makes for better
3003 behaviour in cases where some deliveries succeed, thus re-setting the
3004 "first failed" field. An example is a quota failure for a huge message
3005 when small messages continue to be delivered. Without this change, the
3006 "time since failure" will always be short, possible causing more frequent
3007 delivery attempts for the huge message than are intended.
3008 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3010 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3011 $1, $2, $3) because the numerical variables can be reset during some
3012 expansion items (e.g. "match"), thereby losing the authentication data.
3014 PH/21 Make -bV show the size of off_t variables so that the test suite can
3015 decide whether to run tests for quotas > 2G.
3017 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3018 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3019 filecount value is greater than 2G or if a quota value is greater than 2G
3020 on a system where the size of off_t is not greater than 4, a panic error
3023 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3024 never match. The debug and -bh output now contains an explicit error
3025 message indicating a malformed IPv4 address or mask.
3027 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3028 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3029 PH/23 above applies.
3031 PH/25 Do not write to syslog when running in the test harness. The only
3032 occasion when this arises is a failure to open the main or panic logs
3033 (for which there is an explicit test).
3035 PH/26 Added the /no_tell option to "control=freeze".
3037 PH/27 If a host name lookup failed very early in a connection, for example, if
3038 the IP address matched host_lookup and the reverse lookup yielded a name
3039 that did not have a forward lookup, an error message of the form "no IP
3040 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3041 could be logged. Now it outputs the IP address instead of "NULL".
3043 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3044 allows the sender and the authenticated sender to be set when
3045 submitting a message from within Exim. Since child_open_exim() is
3046 documented for local_scan(), the new function should be too.
3048 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3049 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3050 results in an empty string is now treated as unset.
3052 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3054 PH/31 Added sender_verify_fail logging option.
3056 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3057 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3058 tidied the source and removed it altogether.
3060 PH/33 When a queue run was abandoned because the load average was too high, a
3061 log line was always written; now it is written only if the queue_run log
3062 selector is set. In addition, the log line for abandonment now contains
3063 information about the queue run such as the pid. This is always present
3064 in "start" and "stop" lines but was omitted from the "abandon" line.
3066 PH/34 Omit spaces between a header name and the colon in the error message that
3067 is given when verify = headers_syntax fails (if there are lots of them,
3068 the message gets confusing).
3070 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3071 names, as there are now some PTR records that contain slashes. This check
3072 is only to protect against broken name servers that fall over on strange
3073 characters, so the fact that it applies to all lookups doesn't matter.
3075 PH/36 Now that the new test suite is complete, we can remove some of the
3076 special code in Exim that was needed for the old test suite. For example,
3077 sorting DNS records because real resolvers return them in an arbitrary
3078 order. The new test suite's fake resolver always returns records in the
3081 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3082 bounces) except when queue_only is set, to avoid logging races between
3083 the different processes.
3085 PH/38 Panic-die if .include specifies a non-absolute path.
3087 PH/39 A tweak to the "H" retry rule from its user.
3089 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3090 a label. They prevented compilation on older perls.
3092 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3093 a warning to be raised on newish perls.
3095 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3096 on queue. Changes to match documented behaviour of showing count of
3097 messages matching specified criteria.
3099 PH/40 Changed the default ident timeout from 30s to 5s.
3101 PH/41 Added support for the use of login_cap features, on those BSD systems
3102 that have them, for controlling the resources used by pipe deliveries.
3104 PH/42 The content-scanning code uses fopen() to create files in which to put
3105 message data. Previously it was not paying any attention to the mode of
3106 the files. Exim runs with umask(0) because the rest of the code creates
3107 files with open(), and sets the required mode explicitly. Thus, these
3108 files were ending up world-writeable. This was not a big issue, because,
3109 being within the spool directory, they were not world-accessible. I have
3110 created a function called modefopen, which takes an additional mode
3111 argument. It sets umask(777), creates the file, chmods it to the required
3112 mode, then resets the umask. All the relevant calls to fopen() in the
3113 content scanning code have been changed to use this function.
3115 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3116 to 24 hours. This avoids potential overflow problems when processing G
3117 and H retry rules. I suspect nobody ever tinkers with this value.
3119 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3121 PH/45 When the plaintext authenticator is running as a client, the server's
3122 challenges are checked to ensure they are valid base64 strings. By
3123 default, the authentication attempt is cancelled if an invalid string is
3124 received. Setting client_ignore_invalid_base64 true ignores these errors.
3125 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3126 they are received. Thus, the responses can be made to depend on the
3127 challenges. If an invalid string is ignored, an empty string is placed in
3130 PH/46 Messages that are created by the autoreply transport now contains a
3131 References: header, in accordance with RFCs 2822 and 3834.
3133 PH/47 Added authenticated_sender_force to the smtp transport.
3135 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3137 PH/49 Installed latest patch from the Sieve maintainer.
3139 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3140 was also set, the appendfile transport was unnecessarily scanning a
3141 directory of message files (e.g. for maildir delivery) to find the count
3142 of files (along with the size), even though it did not need this
3143 information. It now does the scan only if it needs to find either the
3144 size of the count of files.
3146 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3148 PH/52 Two bugs concerned with error handling when the smtp transport is
3151 (i) Exim was not creating retry information for temporary errors given
3152 for individual recipients after the DATA command when the smtp transport
3153 was used in LMTP mode. This meant that they could be retried too
3154 frequently, and not timed out correctly.
3156 (ii) Exim was setting the flag that allows error details to be returned
3157 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3158 recipients that were returned after the DATA command.
3160 PH/53 This is related to PH/52, but is more general: for any failing address,
3161 when detailed error information was permitted to be returned to the
3162 sender, but the error was temporary, then after the final timeout, only
3163 "retry timeout exceeded" was returned. Now it returns the full error as
3164 well as "retry timeout exceeded".
3166 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3167 do this, and (what is worse) MTAs that accept it.
3169 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3170 will now be deprecated.
3172 PH/56 New os.c-cygwin from the Cygwin maintainer.
3174 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3175 formats (previously only available in exim formats via -bpr, -bpru,
3176 and -bpra. Now also available in native and exiqgrep formats)
3178 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3179 with very large, slow to parse queues
3181 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3183 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3185 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3186 responses to authentication challenges, though it was showing the
3187 challenges; (ii) I've removed the CR characters from the debug output for
3190 PH/58 Allow for the insertion of a newline as well as a space when a string
3191 is turned into more than one encoded-word during RFC 2047 encoding. The
3192 Sieve code now uses this.
3194 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3195 data_4xx, lost_connection, tls_required.
3197 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3198 message was being sent as an SMTP response.
3200 PH/61 Add -l and -k options to exicyclog.
3202 PH/62 When verifying, if an address was redirected to one new address, so that
3203 verification continued, and the new address failed or deferred after
3204 having set something in $address_data, the value of $address_data was not
3205 passed back to the ACL. This was different to the case when no
3206 redirection occurred. The value is now passed back in both cases.
3208 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3209 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3210 use login_cap.h, so on its own it isn't the distinguishing feature. The
3211 new name refers directly to the setclassresources() function.
3213 PH/65 Added configuration files for NetBSD3.
3215 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3217 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3218 is preferred over IPv4.
3220 PH/68 The bounce_return_message and bounce_return_body options were not being
3221 honoured for bounces generated during the reception of non-SMTP messages.
3222 In particular, this applied to messages rejected by the ACL. This bug has
3223 been fixed. However, if bounce_return_message is true and bounce_return_
3224 body is false, the headers that are returned for a non-SMTP message
3225 include only those that have been read before the error was detected.
3226 (In the case of an ACL rejection, they have all been read.)
3228 PH/69 The HTML version of the specification is now built in a directory called
3229 spec_html instead of spec.html, because the latter looks like a path with
3230 a MIME-type, and this confuses some software.
3232 PH/70 Catch two compiler warnings in sieve.c.
3234 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3235 function verify_get_ident() calls ip_connect() to connect a socket, but
3236 if the "connect()" function timed out, ip_connect() used to close the
3237 socket. However, verify_get_ident() also closes the socket later, and in
3238 between Exim writes to the log, which may get opened at this point. When
3239 the socket was closed in ip_connect(), the log could get the same file
3240 descriptor number as the socket. This naturally causes chaos. The fix is
3241 not to close the socket in ip_connect(); the socket should be closed by
3242 the function that creates it. There was only one place in the code where
3243 this was missing, in the iplookup router, which I don't think anybody now
3244 uses, but I've fixed it anyway.
3246 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3247 well as to direct DNS lookups. Otherwise the handling of names in host
3248 lists is inconsistent and therefore confusing.
3254 PH/01 Two changes to the default runtime configuration:
3256 (1) Move the checks for relay_from_hosts and authenticated clients from
3257 after to before the (commented out) DNS black list checks.
3259 (2) Add control=submission to the relay_from_hosts and authenticated
3260 clients checks, on the grounds that messages accepted by these
3261 statements are most likely to be submissions.
3263 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3265 (1) Generate an error if the third argument for the ${prvs expansion is
3268 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3271 (3) Reset the variables that are obtained from the first argument of
3272 ${prvscheck and used in the second argument before leaving the code,
3273 because their memory is reclaimed, so using them afterwards may do
3276 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3277 one (it's much easier than Tom thought :-).
3279 (5) Because of (4), we can now allow for the use of $prvscheck_result
3280 inside the third argument.
3282 PH/03 For some reason, the default setting of PATH when running a command from
3283 a pipe transport was just "/usr/bin". I have changed it to
3286 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3287 anything to be listed in the output from -bV.
3289 PH/05 When a filter generated an autoreply, the entire To: header line was
3290 quoted in the delivery log line, like this:
3292 => >A.N.Other <ano@some.domain> <original@ddress> ...
3294 This has been changed so that it extracts the operative address. There
3295 may be more than one such address. If so, they are comma-separated, like
3298 => >ano@some.domain,ona@other.domain <original@ddress> ...
3300 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3301 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3302 not being looked up in the rDNS to get a host name, Exim was showing the
3303 IP address twice in Received: lines, even though the IP addresses were
3304 identical. For example:
3306 Received: from [1.2.3.4] (helo=[1.2.3.4])
3308 However, if the real host name was known, it was omitting the HELO data
3309 if it matched the actual IP address. This has been tidied up so that it
3310 doesn't show the same IP address twice.
3312 PH/07 When both +timestamp and +memory debugging was on, the value given by
3313 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3314 called by the debug printing, thereby overwriting the timestamp buffer.
3315 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3317 PH/08 When the original message was included in an autoreply transport, it
3318 always said "this is a copy of the message, including all the headers",
3319 even if body_only or headers_only was set. It now gives an appropriate
3322 PH/09 Applied a patch from the Sieve maintainer which:
3324 o fixes some comments
3325 o adds the (disabled) notify extension core
3326 o adds some debug output for the result of if/elsif tests
3327 o points to the current vacation draft in the documentation
3328 and documents the missing references header update
3332 o fixes a bug in processing the envelope test (when testing
3333 multiple envelope elements, the last element determined the
3336 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3337 Electronic Mail") by including:
3339 Auto-submitted: auto-generated
3341 in the messages that it generates (bounce messages and others, such as
3342 warnings). In the case of bounce messages for non-SMTP messages, there was
3343 also a typo: it was using "Auto_submitted" (underscore instead of
3344 hyphen). Since every message generated by Exim is necessarily in response
3345 to another message, thes have all been changed to:
3347 Auto-Submitted: auto-replied
3349 in accordance with these statements in the RFC:
3351 The auto-replied keyword:
3353 - SHOULD be used on messages sent in direct response to another
3354 message by an automatic process,
3356 - MUST NOT be used on manually-generated messages,
3358 - MAY be used on Delivery Status Notifications (DSNs) and Message
3359 Disposition Notifications (MDNs),
3361 - MUST NOT be used on messages generated by automatic or periodic
3362 processes, except for messages which are automatic responses to
3365 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3366 to the default Received: header definition.
3368 PH/12 Added log selector acl_warn_skipped (default on).
3370 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3371 variables because (a) they are in the wrong storage pool and (b) even if
3372 they were copied, it wouldn't work properly because of the caching.
3374 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3375 checking when decoding. Apparently there are clients that generate
3376 overlong encoded strings. Why am I not surprised?
3378 PH/15 If the first argument of "${if match_address" was not empty, but did not
3379 contain an "@" character, Exim crashed. Now it writes a panic log message
3380 and treats the condition as false.
3382 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3384 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3385 extension "envelope-auth". The code is finished and in agreement with
3386 other implementations, but there is no documentation so far and in fact,
3387 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3388 not changing the active code.
3390 Print executed "if" and "elsif" statements when debugging is used. This
3391 helps a great deal to understand what a filter does.
3393 Document more things not specified clearly in RFC3028. I had all this
3394 sorted out, when out of a sudden new issues came to my mind. Oops."
3396 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3399 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3400 canonical form (as documented). However, after a host name lookup from
3401 the IP address, check_host() was doing a simple string comparison with
3402 addresses acquired from the DNS when checking that the found name did
3403 have the original IP as one of its addresses. Since any found IPv6
3404 addresses are likely to be in abbreviated form, the comparison could
3405 fail. Luckily, there already exists a function for doing the comparison
3406 by converting both addresses to binary, so now that is used instead of
3407 the text comparison.
3409 PH/20 There was another similar case to PH/19, when a complete host name was
3410 given in a host list; looking up its IP address could give an abbreviated
3411 form, whereas the current host's name might or might not be abbreviated.
3412 The same fix has been applied.
3418 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3419 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3422 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3423 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3425 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3427 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3428 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3429 Also, for "Auto-Submitted", treat anything other than "no" as
3430 non-personal, in accordance with RFC 3834. (Previously it treated
3431 anything starting "auto-" as non-personal.)
3433 TF/01 The control=submission/name=... option had a problem with syntax
3434 errors if the name included a slash character. The /name= option
3435 now slurps the rest of the string, so it can include any characters
3436 but it must come last in the list of options (after /sender_retain
3439 PH/05 Some modifications to the interface to the fake nameserver for the new
3447 TK/01 Added the "success_on_redirect" address verification option. See
3448 NewStuff for rationale and an example.
3450 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3452 PH/02 Patch to exigrep to allow it to work on syslog lines.
3454 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3455 fread() to skip over the body file's header line, because in Cygwin the
3456 header line is locked and is inaccessible.
3458 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3459 co-exist for some time) to make it clear that it is the Exim ID that is
3460 referenced, not the Message-ID: header line.
3462 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3463 string_format() function, because snprintf() does not exist on all
3466 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3467 ${stat: expansion item.
3469 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3470 protocol synchronization error", to keep the pedants happy.
3472 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3473 well as for IRIX systems, when gcc is being used. See the host.c source
3476 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3478 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3481 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3482 transport and to the smtp transport in LMTP mode.
3484 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3486 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3487 run a filter in a subprocess. This could lead to confusion in subsequent
3488 lookups in the parent process. There should also be a search_tidyup() at
3489 the end of the subprocess.
3491 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3492 only if the host matched helo_try_verify_hosts, which caused the
3493 verification to occur when the EHLO/HELO command was issued. The ACL just
3494 tested the remembered result. Now, if a previous verification attempt has
3495 not happened, "verify = helo" does it there and then.
3497 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3499 TK/03 Fix log output including CR from clamd.
3501 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3502 exist provoked a memory error which could cause a segfault.
3504 PH/15 Installed PCRE 6.2
3506 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3508 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3509 of the problem. Specifically, suggested +O2 rather than +O1 for the
3512 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3514 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3515 record, the port was not correctly logged when the outgoing_port log
3516 selector was set (it logged the transort's default port).
3518 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3519 fallback_hosts, and "hosts" in the smtp transport.
3521 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3522 host errors such as "Connection refused".
3524 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3525 authentication with radiusclient 0.4.9:
3527 - Error returned from rc_read_config was caught wrongly
3528 - Username/password not passed on to radius server due to wrong length.
3530 The presumption is that some radiusclient API changes for 4.51/PH/17
3531 were not taken care of correctly. The code is still untested by me (my
3532 Linux distribution still has 0.3.2 of radiusclient), but it was
3533 contributed by a Radius user.
3535 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3536 expanding the "port" option of the smtp transport.
3538 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3539 while reading a MIME header. Thanks to Tom Hughes for a patch.
3541 PH/24 Include config.h inside local_scan.h so that configuration settings are
3544 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3545 that in an ACL for RCPT (for example), you can examine exactly what was
3548 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3549 commands, but it was not correctly comparing the address with the actual
3550 client host address. Thus, it would show the EHLO address in Received:
3551 header lines when this was not necessary.
3553 PH/27 Added the % operator to ${eval:}.
3555 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3556 it should be ignoring failures (because with -C, for example, it has lost
3557 privilege). It wasn't ignoring creation failures other than "already
3560 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3563 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3564 because an input line was too long, either on its own, or by virtue of
3565 too many continuations, the temporary file was not being removed, and the
3566 return code was incorrect.
3568 PH/31 Missing "BOOL" in function definition in filtertest.c.
3570 PH/32 Applied Sieve patches from the maintainer.
3572 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3574 PH/33 Added "verify = not_blind".
3576 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3577 Local/Makefile (with some defaults set). These are used in built scripts
3578 such as exicyclog, but they have never been used in the exim_install
3579 script (though there are many overriding facilities there). I have
3580 arranged that the exim_install script now takes note of these two
3583 PH/35 Installed configuration files for Dragonfly.
3585 PH/36 When a locally submitted message by a trusted user did not contain a
3586 From: header, and the sender address was obtained from -f or from an SMTP
3587 MAIL command, and the trusted user did not use -F to supply a sender
3588 name, $originator_name was incorrectly used when constructing a From:
3589 header. Furthermore, $originator_name was used for submission mode
3590 messages from external hosts without From: headers in a similar way,
3591 which is clearly wrong.
3593 PH/37 Added control=suppress_local_fixups.
3595 PH/38 When log_selector = +received_sender was set, and the addition of the
3596 sender made the log line's construction buffer exactly full, or one byte
3597 less than full, an overflow happened when the terminating "\n" was
3600 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3601 when the result of a list match is failure because a DNS lookup failed.
3603 PH/40 RM_COMMAND is now used in the building process.
3605 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3606 the "build-* directories that it finds.
3608 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3609 address in a domain literal was a prefix of an interface address.
3611 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3612 when verifying a sender address, unless rewrite_headers is false.
3614 PH/44 Wrote a long comment about why errors_to addresses are verified as
3615 recipients, not senders.
3617 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3618 the ratelimit ACL was added.
3620 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3622 PH/46 Added extra information about PostgreSQL errors to the error string.
3624 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3625 avoiding the need to install special zones in a real server. This is
3626 backwards compatible; if it can't find the fake resolver, it drops back.
3627 Thus, both old and new test suites can be run.
3629 TF/02 Added util/ratelimit.pl
3631 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3632 clock is set back in time.
3634 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3635 Candler <B.Candler@pobox.com>.
3637 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3638 OK for addresses that are the result of redirections.
3640 PH/48 A number of further additions for the benefit of the new test suite,
3641 including a fake gethostbyname() that interfaces to the fake DNS resolver
3644 TF/06 The fix for widen_domains has also been applied to qualify_single and
3645 search_parents which are the other dnslookup options that can cause
3648 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3651 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3653 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3654 service" if it encountered an MX record with an empty target hostname.
3655 The message is now "an MX or SRV record indicated no SMTP service".
3657 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3658 if the DNS of the sending site is misconfigured. This is quite a
3659 common situation. This change restores the behaviour of treating a
3660 helo verification defer as a failure.
3662 PH/51 If self=fail was set on a router, the bounce message did not include the
3663 actual error message.
3669 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3671 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3672 command itself ended in error, the underlying message about the transport
3673 filter timeout was being overwritten with the pipe command error. Now the
3674 underlying error message should be appended to the second error message.
3676 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3678 PH/02 Reduce the amount of output that "make" produces by default. Full output
3679 can still be requested.
3681 PH/03 The warning log line about a condition test deferring for a "warn" verb
3682 was being output only once per connection, rather than after each
3683 occurrence (because it was using the same function as for successful
3684 "warn" verbs). This seems wrong, so I have changed it.
3686 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3687 it should not have, which might have caused a crash in the right
3688 circumstances, but probably never did.
3690 PH/04 Installed a modified version of Tony Finch's patch to make submission
3691 mode fix the return path as well as the Sender: header line, and to
3692 add a /name= option so that you can make the user's friendly name appear
3695 TF/03 Added the control = fakedefer ACL modifier.
3697 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3698 Mark Lowes for thorough testing.
3700 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3702 TK/03 Merged latest SRS patch from Miles Wilton.
3704 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3705 with the definition in sysexits.h (which is #included earlier).
3706 Fortunately, Exim does not actually use EX_OK. The code used to try to
3707 preserve the sysexits.h value, by assuming that macro definitions were
3708 scanned for macro replacements. I have been disabused of this notion,
3709 so now the code just undefines EX_OK before #including unistd.h.
3711 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3712 in the smtp transport. When a block could not be written in a single
3713 write() function, the timeout was being re-applied to each part-write.
3714 This seems wrong - if the receiver was accepting one byte at a time it
3715 would take for ever. The timeout is now adjusted when this happens. It
3716 doesn't have to be particularly precise.
3718 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3719 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3721 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3722 without a domain if the check to <postmaster@domain> fails.
3724 SC/01 Eximstats: added -xls and the ability to specify output files
3725 (patch written by Frank Heydlauf).
3727 SC/02 Eximstats: use FileHandles for outputting results.
3729 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3731 SC/04 Eximstats: fixed display of large numbers with -nvr option
3733 SC/05 Eximstats: fixed merging of reports with empty tables.
3735 SC/06 Eximstats: added the -include_original_destination flag
3737 SC/07 Eximstats: removed tabs and trailing whitespace.
3739 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3741 TK/06 MBOX spool code: Add real "From " MBOX separator line
3742 so the .eml file is really in mbox format (even though
3743 most programs do not really care). Patch from Alex Miller.
3745 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3746 The latter is generated from $received_to and is only set if the
3747 message has one envelope recipient. SA can use these headers,
3748 obviously out-of-the-box. Patch from Alex Miller.
3750 PH/08 The ${def test on a variable was returning false if the variable's
3751 value was "0", contrary to what the specification has always said!
3752 The result should be true unless the variable is empty.
3754 PH/09 The syntax error of a character other than { following "${if
3755 def:variable_name" (after optional whitespace) was not being diagnosed.
3756 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3757 accidental colon was present, for example, could give incorrect results.
3759 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3760 result is used (not including appendfile, where other changes are about
3763 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3764 This involved changing a lot of size variables from int to off_t. It
3765 should work with maildirs and everything.
3767 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3768 spamd dying while we are connected to it.
3770 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3773 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3774 function for BDB 4.3.
3776 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3778 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3779 the sources for PCRE have been re-arranged and more files are now
3782 PH/15 The code I had for printing potentially long long variables in PH/11
3783 above was not the best (it lost precision). The length of off_t variables
3784 is now inspected at build time, and an appropriate printing format (%ld
3785 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3786 to be "long long int" or "long int". This is needed for the internal
3787 formatting function string_vformat().
3789 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3790 the configuration file to be ":syslog", then the script "guesses" where
3791 the logs files are, rather than using the compiled in default. In our
3792 case the guess is not the same as the compiled default, so the script
3793 suddenly stopped working when I started to use syslog. The patch checks
3794 to see if log_file_path is "". If so, it attempts to read it from exim
3795 with no configuration file to get the compiled in version, before it
3796 falls back to the previous guessing code."
3798 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3799 implementing BATV in an Exim configuration. See NewStuff for the gory
3802 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3803 Makefile that are specific to HP-UX.
3805 PH/18 If the "use_postmaster" option was set for a recipient callout together
3806 with the "random" option, the postmaster address was used as the MAIL
3807 FROM address for the random test, but not for the subsequent recipient
3808 test. It is now used for both.
3810 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3811 patch removes a few documentation additions to RFC 3028, because the
3812 latest draft now contains them. It adds the new en;ascii-case comparator
3813 and a new error check for 8bit text in MIME parts. Comparator and
3814 require names are now matched exactly. I enabled the subaddress
3815 extension, but it is not well tested yet (read: it works for me)."
3817 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3818 rework some of the code of TK/09 above to avoid the hardwired use of
3819 "%lld" and "long long". Replaced the call to snprintf() with a call to
3822 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3823 records point to non-existent hosts", "retry timeout exceeded", and
3824 "retry time not reached for any host after a long failure period".
3826 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3827 experimental DomainKeys support:
3829 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3830 (2) On an error such as an illegally used "control", the wrong name for
3831 the control was given.
3833 These problems did NOT occur unless DomainKeys support was compiled.
3835 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3837 PH/24 Added ${if match_ip condition.
3839 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3840 databases so that it will be absolutely obvious if a crash occurs in the
3841 DB library. This is a regular occurrence (often caused by mis-matched
3844 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3845 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3846 picked up on a user's system that detects such things. There doesn't seem
3847 to be a gcc warning option for this - only an attribute that has to be
3848 put on the function's prototype. It seems that in Fedora Core 4 they have
3849 set this on a number of new functions. No doubt there will be more in due
3852 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3853 specify a transport. However, if an address that was verified by such a
3854 router was the subject of a callout, Exim crashed because it tried to
3855 read the rcpt_include_affixes from the non-existent transport. Now it
3856 just assumes that the setting of that option is false. This bug was
3857 introduced by 4.51/PH/31.
3859 PH/28 Changed -d+all to exclude +memory, because that information is very
3860 rarely of interest, but it makes the output a lot bigger. People tend to
3861 do -d+all out of habit.
3863 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3864 code in os-type was giving problems when libc.so lives in lib64, like on
3867 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3868 aren't the modern standard, and it seems that some systems' include files
3869 don't always have them. Exim was already checking for some of the newer
3870 ones like T_AAAA, and defining it itself. I've added checks for all the
3871 record types that Exim uses.
3873 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3874 not automatically generating a new one, as it is supposed to. This
3875 prevented TLS from working. If the file did exist, but contained invalid
3876 data, a new version was generated, as expected. It was only the case of a
3877 non-existent file that was broken.
3879 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3880 with a change in libdomainkeys > 0.64.
3882 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3883 from DNS. If the selector record carries the flag, it now has
3884 precedence over the domain-wide flag.
3886 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3888 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3889 the use of a port name that isn't defined in /etc/services) occurred, the
3890 message was deferred as in a normal delivery, and thus remained on the
3891 spool, instead of being failed because of the mua_wrapper setting. This
3892 is now fixed, and I tidied up some of the mua_wrapper messages at the
3895 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3896 the messages in a hash of arrays rather than using individual hashes.
3897 This is a bit cleaner and results in dramatic memory savings, albeit
3898 at a slight CPU cost.
3900 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3901 as requested by Marc Sherman.
3903 SC/10 Eximstats: added histograms for user specified patterns as requested
3906 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3908 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3909 fopen() in the content-scanning modules that did not already have it.
3915 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3916 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3918 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3920 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3922 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3923 to test Sieve filters that use "vacation".
3925 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3926 that changes the way the GnuTLS parameters are stored in the cache file.
3927 The new format can be generated externally. For backward compatibility,
3928 if the data in the cache doesn't make sense, Exim assumes it has read an
3929 old-format file, and it generates new data and writes a new file. This
3930 means that you can't go back to an older release without removing the
3933 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3934 work if there are any delivery delays because "one_time" forces the
3935 parent to be marked "delivered", so its unseen clone is never tried
3936 again. For this reason, Exim now forbids the simultaneous setting of
3939 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3940 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3941 entry if you want to know the details. The fix, however, neglected to
3942 consider the case when local delivery batching is involved. The test for
3943 "previously delivered" was not happening when checking to see if an
3944 address could be batched with a previous (undelivered) one; under
3945 certain circumstances this could lead to multiple deliveries to the same
3948 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3949 in its include files, and this causes problems building Exim.
3951 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3952 header_syntax) but Exim was just ignoring anything given after a slash.
3953 In particular, this caused confusion with an attempt to use "verify =
3954 reverse_host_lookup/defer_ok". An error is now given when options are
3955 supplied for verify items that do not have them. (Maybe reverse_host_
3956 lookup should have a defer_ok option, but that's a different point.)
3958 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3959 defined by RFC 821) to 2048, because there were problems with some AUTH
3960 commands, and RFC 1869 says the size should be increased for extended
3961 SMTP commands that take arguments.
3963 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3966 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3967 "unknown" error; now it says that the functionality isn't in the binary.
3969 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3970 an address' error message when a string expansion fails (syntax or
3971 whatever). Otherwise the password may appear in the log. Following change
3972 PH/42 below, there is no longer a chance of it appearing in a bounce
3975 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3977 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3978 of its addresses was ever tried. (Bugzilla bug #2.)
3980 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3981 the result incorrectly in the debug output. (It correctly added a newline
3982 to what was transported.)
3984 TF/01 Added $received_time.
3986 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3987 commented out examples of how to interface to a virus scanner and to
3988 SpamAssassin. Also added commented examples of av_scanner and
3989 spamd_address settings.
3991 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3992 and controls are allowed in which ACLs. There were a couple of minor
3993 errors. Some of the entries in the conditions table (which is a table of
3994 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3995 negation of where the condition IS allowed.
3997 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3999 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4000 header file does not have a version number, so I've had to invent a new
4001 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4002 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4003 radiusclient), but it was contributed by a Radius user.
4005 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4006 files or directories, for OpenSSL.
4008 PH/19 When an Exim process that is running as root has to create an Exim log
4009 file, it does so in a subprocess that runs as exim:exim so as to get the
4010 ownership right at creation (otherwise, other Exim processes might see
4011 the file with the wrong ownership). There was no test for failure of this
4012 fork() call, which would lead to the process getting stuck as it waited
4013 for a non-existent subprocess. Forks do occasionally fail when resources
4014 run out. I reviewed all the other calls to fork(); they all seem to check
4017 PH/20 When checking for unexpected SMTP input at connect time (before writing
4018 the banner), Exim was not dealing correctly with a non-positive return
4019 from the read() function. If the client had disconnected by this time,
4020 the result was a log entry for a synchronization error with an empty
4021 string after "input=" when read() returned zero. If read() returned -1
4022 (an event I could not check), uninitialized data bytes were printed.
4023 There were reports of junk text (parts of files, etc) appearing after
4026 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4028 PH/22 Added support for macro redefinition, and (re)definition in between
4029 driver and ACL definitions.
4031 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4032 forgetting to use the resulting value; it was using the unexpanded value.
4034 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4035 hadn't been configured. The fix is from Juergen Kreileder, who
4036 understands it better than I do:
4038 "Here's what I see happening with three configured cyrus_sasl
4039 authenticators configured (plain, login, cram-md5):
4041 On startup auth_cyrus_sasl_init() gets called for each of these.
4042 This means three calls to sasl_listmech() without a specified mech_list.
4043 => SASL tests which mechs of all available mechs actually work
4044 => three warnings about OTP not working
4045 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4047 With the patch, sasl_listmech() also gets called three times. But now
4048 SASL's mech_list option is set to the server_mech specified in the the
4049 authenticator. Or in other words, the answer from sasl_listmech()
4050 gets limited to just the mech you're testing for (which is different
4052 => the return list contains just 'plain' or 'login', 'cram-md5' or
4053 nothing depending on the value of ob->server_mech.
4055 I've just tested the patch: Authentication still works fine,
4056 unavailable mechs specified in the exim configuration are still
4057 caught, and the auth.log warnings about OTP are gone."
4059 PH/25 When debugging is enabled, the contents of the command line are added
4060 to the debugging output, even when log_selector=+arguments is not
4063 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4064 answer is "GNU", and only if the return is "GNU/something" is the answer
4067 PH/27 $acl_verify_message is now set immediately after the failure of a
4068 verification in an ACL, and so is available in subsequent modifiers. In
4069 particular, the message can be preserved by coding like this:
4071 warn !verify = sender
4072 set acl_m0 = $acl_verify_message
4074 Previously, $acl_verify_message was set only while expanding "message"
4075 and "log_message" when a very denied access.
4077 PH/28 Modified OS/os.c-Linux with
4079 -#ifndef OS_LOAD_AVERAGE
4080 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4082 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4083 nomenclature these days.)
4085 PH/29 Installed patch from the Sieve maintainer that adds the options
4086 sieve_useraddress and sieve_subaddress to the redirect router.
4088 PH/30 In these circumstances:
4089 . Two addresses routed to the same list of hosts;
4090 . First host does not offer TLS;
4091 . First host accepts first address;
4092 . First host gives temporary error to second address;
4093 . Second host offers TLS and a TLS session is established;
4094 . Second host accepts second address.
4095 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4096 and peerdn, if requested) that were in fact used only for the second
4099 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4100 attention to any local part prefix or suffix that was matched by the
4101 router that accepted the address. It now behaves in the same way as it
4102 does for delivery: the affixes are removed from the local part unless
4103 rcpt_include_affixes is set on the transport.
4105 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4106 timeout during the DATA phase of an incoming message.
4108 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4109 applied a patch sanctioned by the Sieve maintainer.
4111 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4112 the uid or gid is negative. A case of a negative gid caused this to be
4113 noticed. The fix allows for either to be negative.
4115 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4116 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4119 PH/36 The change PH/12 above was broken. Fixed it.
4121 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4122 the grounds that routing the same address twice would always produce the
4123 same answer. This might have been true once, but it is certainly no
4124 longer true now. Routing a child address may depend on the previous
4125 routing that produced that child. Some complicated redirection strategies
4126 went wrong when messages had multiple recipients, and made Exim's
4127 behaviour dependent on the order in which the addresses were given.
4129 I have moved the duplicate checking until after the routing is complete.
4130 Exim scans the addresses that are assigned to local and remote
4131 transports, and removes any duplicates. This means that more work will be
4132 done, as duplicates will always all be routed, but duplicates are
4133 presumably rare, so I don't expect this is of any significance.
4135 For deliveries to pipes, files, and autoreplies, the duplicate checking
4136 still happens during the routing process, since they are not going to be
4139 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4140 It corrects a timeout issue with spamd. This is Ian's comment: "The
4141 background is that sometimes spamd either never reads data from a
4142 connection it has accepted, or it never writes response data. The exiscan
4143 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4144 blindly assumes that writes won't block so it may never time out."
4146 PH/39 Allow G after quota size as well as K and M.
4148 PH/40 The value set for $authenticated_id in an authenticator may not contain
4149 binary zeroes or newlines because the value is written to log lines and
4150 to spool files. There was no check on this. Now the value is run through
4151 the string_printing() function so that such characters are converted to
4152 printable escape sequences.
4154 PH/41 $message_linecount is a new variable that contains the total number of
4155 lines in the message. Compare $body_linecount, which is the count for the
4158 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4159 bounce and delay warning messages, except in certain special cases, which
4162 (a) An SMTP error message from a remote host;
4163 (b) A message specified in a :fail: redirection;
4164 (c) A message specified in a "fail" command in a system filter;
4165 (d) A message specified in a FAIL return from the queryprogram router;
4166 (e) A message specified by the cannot_route_message router option.
4168 In these cases only, Exim does include the error details in bounce and
4169 warning messages. There are also a few cases where bland messages such
4170 as "unrouteable address" or "local delivery error" are given.
4172 PH/43 $value is now also set for the "else" part of a ${run expansion.
4174 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4175 being worked on, but at least Exim now implements the latest version to
4178 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4179 process to complete was treated as a delivery failure, a timeout while
4180 writing the message to the pipe was logged, but erroneously treated as a
4181 successful delivery. Such timeouts include transport filter timeouts. For
4182 consistency with the overall process timeout, these timeouts are now
4183 treated as errors, giving rise to delivery failures by default. However,
4184 there is now a new Boolean option for the pipe transport called
4185 timeout_defer, which, if set TRUE, converts the failures into defers for
4186 both kinds of timeout. A transport filter timeout is now identified in
4189 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4190 systems where "make" and "gmake" are different, calling "gmake" at top
4191 level broke things. I've arranged for the value of $(MAKE) to be passed
4192 from the Makefile to this script so that it can call the same version of
4196 A note about Exim versions 4.44 and 4.50
4197 ----------------------------------------
4199 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4200 changes of various kinds. As a consequence, a big documentation update was
4201 needed. This delayed the release for rather longer than seemed good, especially
4202 in the light of a couple of (minor) security issues. Therefore, the changes
4203 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4204 release. So 4.44 and 4.50 are in effect two different branches that both start
4207 I have left the 4.50 change log unchanged; it contains all the changes since
4208 4.43. The change log for 4.44 is below; many of its items are identical to
4209 those for 4.50. This seems to be the most sensible way to preserve the
4210 historical information.
4216 1. Minor wording change to the doc/README.SIEVE file.
4218 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4219 computation of the current number of files was incorrect.
4221 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4222 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4225 4. Give more explanation in the error message when the command for a transport
4226 filter fails to execute.
4228 5. There are several places where Exim runs a non-Exim command in a
4229 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4230 was being done only for the command run by the queryprogram router. It is
4231 now done for all such subprocesses. The other cases are: ${run, transport
4232 filters, and the commands run by the lmtp and pipe transports.
4234 6. Added CONFIGURE_GROUP build-time option.
4236 7. Some older OS have a limit of 256 on the maximum number of file
4237 descriptors. Exim was using setrlimit() to set 1000 as a large value
4238 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4239 systems. I've change it so that if it can't get 1000, it tries for 256.
4241 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4242 was an oversight, and furthermore, ever since the addition of extra
4243 controls (e.g. 4.43/32), the checks on when to allow different forms of
4244 "control" were broken. There should now be diagnostics for all cases when a
4245 control that does not make sense is encountered.
4247 9. Added the /retain_sender option to "control=submission".
4249 10. $recipients is now available in the predata ACL (oversight).
4251 11. Tidy the search cache before the fork to do a delivery from a message
4252 received from the command line. Otherwise the child will trigger a lookup
4253 failure and thereby defer the delivery if it tries to use (for example) a
4254 cached ldap connection that the parent has called unbind on.
4256 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4257 of $address_data from the recipient verification was clobbered by the
4258 sender verification.
4260 13. The value of address_data from a sender verification is now available in
4261 $sender_address_data in subsequent conditions in the ACL statement.
4263 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4265 15. Added a new option "connect=<time>" to callout options, to set a different
4268 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4269 was its contents. (It was OK if the option was not defined at all.)
4271 17. A "Completed" log line is now written for messages that are removed from
4272 the spool by the -Mrm option.
4274 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4275 information about exactly what failed.
4277 19. Added -dd to debug only the daemon process.
4279 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4280 handles timeouts, both on the server side and network timeouts. Renamed the
4281 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4283 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4284 It is now set to "smtps".
4286 22. $host_address is now set to the target address during the checking of
4287 ignore_target_hosts.
4289 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4290 being passed; this would have caused $sender_host_name to have been used if
4291 matching the list had actually called for a host name (not very likely,
4292 since this list is usually IP addresses). A host name is now passed as
4295 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4296 code that specifies a non-restarting handler (typically sigaction() in
4297 modern systems) in an attempt to fix a rare and obscure crash bug.
4299 25. Narrowed the window for a race in the daemon that could cause it to ignore
4300 SIGCHLD signals. This is not a major problem, because they are used only to
4301 wake it up if nothing else does.
4303 26. A malformed maildirsize file could cause Exim to calculate negative values
4304 for the mailbox size or file count. Odd effects could occur as a result.
4305 The maildirsize information is now recalculated if the size or filecount
4308 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4309 support for a long time. Removed HAVE_SYS_VFS_H.
4311 28. Installed the latest version of exipick from John Jetmore.
4313 29. In an address list, if the pattern was not a regular expression, an empty
4314 subject address (from a bounce message) matched only if the pattern was an
4315 empty string. Non-empty patterns were not even tested. This was the wrong
4316 because it is perfectly reasonable to use an empty address as part of a
4317 database query. An empty address is now tested by patterns that are
4318 lookups. However, all the other forms of pattern expect the subject to
4319 contain a local part and a domain, and therefore, for them, an empty
4320 address still always fails if the pattern is not itself empty.
4322 30. Exim went into a mad DNS loop when attempting to do a callout where the
4323 host was specified on an smtp transport, and looking it up yielded more
4324 than one IP address.
4326 31. Re-factored the code for checking spool and log partition space into a
4327 function that finds that data and another that does the check. The former
4328 is then used to implement four new variables: $spool_space, $log_space,
4329 $spool_inodes, and $log_inodes.
4331 32. The RFC2047 encoding function was originally intended for short strings
4332 such as real names; it was not keeping to the 75-character limit for
4333 encoded words that the RFC imposes. It now respects the limit, and
4334 generates multiple encoded words if necessary. To be on the safe side, I
4335 have increased the buffer size for the ${rfc2047: expansion operator from
4338 33. It is now permitted to omit both strings after an "if" condition; if the
4339 condition is true, the result is "true". As before, when the second string
4340 is omitted, a false condition yields an empty string. This makes it less
4341 cumbersome to write custom ACL and router conditions.
4343 34. Failure to deliver a bounce message always caused it to be frozen, even if
4344 there was an errors_to setting on the router. The errors_to setting is now
4347 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4348 canonical form (fully expanded) before being placed in
4349 $sender_host_address.
4351 36. The table in the code that translates DNS record types into text (T_A to
4352 "A" for instance) was missing entries for NS and CNAME. It is just possible
4353 that this could have caused confusion if both these types were looked up
4354 for the same domain, because the text type is used as part of Exim's
4355 per-process caching. But the chance of anyone hitting this buglet seems
4358 37. The dnsdb lookup has been extended in a number of ways.
4360 (1) There is a new type, "zns", which walks up the domain tree until it
4361 finds some nameserver records. It should be used with care.
4363 (2) There is a new type, "mxh", which is like "mx" except that it returns
4364 just the host names, not the priorities.
4366 (3) It is now possible to give a list of domains (or IP addresses) to be
4367 looked up. The behaviour when one of the lookups defers can be
4368 controlled by a keyword.
4370 (4) It is now possible to specify the separator character for use when
4371 multiple records are returned.
4373 38. The dnslists ACL condition has been extended: it is now possible to supply
4374 a list of IP addresses and/or domains to be looked up in a particular DNS
4377 39. Added log_selector=+queue_time_overall.
4379 40. When running the queue in the test harness, wait just a tad after forking a
4380 delivery process, to get repeatability of debugging output.
4382 41. Include certificate and key file names in error message when GnuTLS fails
4383 to set them up, because the GnuTLS error message doesn't include the name
4384 of the failing file when there is a problem reading it.
4386 42. Allow both -bf and -bF in the same test run.
4388 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4390 44. The "Exiscan patch" is now merged into the mainline Exim source.
4392 45. Sometimes the final signoff response after QUIT could fail to get
4393 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4394 < 0 before doing a fflush(). This bug looks as though it goes back to the
4395 introduction of TLS in release 3.20, but "sometimes" must have been rare
4396 because the tests only now provoked it.
4398 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4399 (this can affect the format of dates).
4401 47. exim_tidydb, when checking for the continued existence of a message for
4402 which it has found a message-specific retry record, was not finding
4403 messages that were in split spool directories. Consequently, it was
4404 deleting retry records that should have stayed in existence.
4406 48. Steve fixed some bugs in eximstats.
4408 49. The SPA authentication driver was not abandoning authentication and moving
4409 on to the next authenticator when an expansion was forced to fail,
4410 contradicting the general specification for all authenticators. Instead it
4411 was generating a temporary error. It now behaves as specified.
4413 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4414 (the order specifies the preference for clients). The order is now AES256,
4415 AES128, 3DES, ARCFOUR128.
4417 51. Small patch to Sieve code - explicitly set From: when generating an
4420 52. Exim crashed if a remote delivery caused a very long error message to be
4421 recorded - for instance if somebody sent an entire SpamAssassin report back
4422 as a large number of 550 error lines. This bug was coincidentally fixed by
4423 increasing the size of one of Exim's internal buffers (big_buffer) that
4424 happened as part of the Exiscan merge. However, to be on the safe side, I
4425 have made the code more robust (and fixed the comments that describe what
4428 53. Now that there can be additional text after "Completed" in log lines (if
4429 the queue_time_overall log selector is set), a one-byte patch to exigrep
4430 was needed to allow it to recognize "Completed" as not the last thing in
4433 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4434 patch that reportedly fixes this has been added. I am not expert enough to
4435 create a test for it. This is what the patch creator wrote:
4437 "I found a little strange behaviour of ldap code when working with
4438 Windows 2003 AD Domain, where users was placed in more than one
4439 Organization Units. When I tried to give exim partial DN, the exit code
4440 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4441 But simultaneously result of request was absolutely normal ldap result,
4442 so I produce this patch..."
4444 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4445 REFERENCE, so I have modified the code to exclude the patch when that macro
4448 55. Some experimental protocols are using DNS PTR records for new purposes. The
4449 keys for these records are domain names, not reversed IP addresses. The
4450 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4451 leaves it alone. Component reversal etc. now happens only for IP addresses.
4454 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4456 57. Double the size of the debug message buffer (to 2048) so that more of very
4457 long debug lines gets shown.
4459 58. The exicyclog utility now does better if the number of log files to keep
4460 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4462 59. Two changes related to the smtp_active_hostname option:
4464 (1) $smtp_active_hostname is now available as a variable.
4465 (2) The default for smtp_banner uses $smtp_active_hostname instead
4466 of $primary_hostname.
4468 60. The host_aton() function is supposed to be passed a string that is known
4469 to be a valid IP address. However, in the case of IPv6 addresses, it was
4470 not checking this. This is a hostage to fortune. Exim now panics and dies
4471 if the condition is not met. A case was found where this could be provoked
4472 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4473 components; fortuitously, this particular loophole had already been fixed
4474 by change 4.50/55 above.
4476 If there are any other similar loopholes, the new check in host_aton()
4477 itself should stop them being exploited. The report I received stated that
4478 data on the command line could provoke the exploit when Exim was running as
4479 exim, but did not say which command line option was involved. All I could
4480 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4481 running as the user.
4484 61. There was a buffer overflow vulnerability in the SPA authentication code
4485 (which came originally from the Samba project). I have added a test to the
4486 spa_base64_to_bits() function which I hope fixes it.
4489 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4490 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4492 63. The daemon start-up calls getloadavg() while still root for those OS that
4493 need the first call to be done as root, but it missed one case: when
4494 deliver_queue_load_max is set with deliver_drop_privilege. This is
4495 necessary for the benefit of the queue runner, because there is no re-exec
4496 when deliver_drop_privilege is set.
4498 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4499 This has been fixed.
4501 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4502 was in use, was not putting the data itself into the right store pool;
4503 consequently, it could be overwritten for a subsequent message in the same
4504 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4507 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4509 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4510 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4511 the function were treating the return as a boolean value, which happened to
4512 work because 0=false and not-0=true, but is not correct code.
4514 68. The host_aton() function was not handling scoped IPv6 addresses (those
4515 with, for example, "%eth0" on the end) correctly.
4517 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4518 negated items (that is, ~something) in unsigned ints. Some compilers
4519 apparently mutter when there is no cast.
4521 70. If an address verification called from an ACL failed, and did not produce a
4522 user-specific message (i.e. there was only a "system" message), nothing was
4523 put in $acl_verify_message. In this situation, it now puts the system
4526 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4527 session; change 4.31/43 added the unwanted input to the log line - except
4528 that it did not do this in the start of session case. It now does.
4530 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4531 This is wrong and can cause the other end to generate a synchronization
4532 error if it is another Exim or anything else that does the synchronization
4533 check. A QUIT command is no longer sent after a timeout.
4535 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4536 during host lookups.
4538 74. The defer_ok option of callout verification was not working if it was used
4539 when verifying addresses in header lines, that is, for this case:
4541 verify = header_sender/callout=defer_ok
4543 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4544 those file descriptors could be used for SMTP connections. If anything
4545 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4546 could be sent to the SMTP client, causing chaos. The daemon now opens
4547 stdin, stdout, and stderr to /dev/null when it puts itself into the
4550 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4551 log by default. The user can override this with suitable Perl magic.
4553 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4554 the log message when discard triggers, was not working for the DATA ACL or
4555 for the non-SMTP ACL.
4557 78. Error message wording change in sieve.c.
4559 79. If smtp_accept_max_per_host was set, the number of connections could be
4560 restricted to fewer than expected, because the daemon was trying to set up
4561 a new connection before checking whether the processes handling previous
4562 connections had finished. The check for completed processes is now done
4563 earlier. On busy systems, this bug wouldn't be noticed because something
4564 else would have woken the daemon, and it would have reaped the completed
4567 80. If a message was submitted locally by a user whose login name contained one
4568 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4569 It caused a spool format error. I have fixed the spool reading code. A
4570 related problem was that the "from" clause in the Received: line became
4571 illegal because of the space(s). It is now covered by ${quote_local_part.
4573 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4576 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4578 83. Patch from Sieve maintainer to fix unterminated string problem in
4579 "vacation" handling.
4581 84. Some minor changes to the Linux configuration files to help with other
4582 OS variants using glibc.
4584 85. One more patch for Sieve to update vacation handling to latest spec.
4587 ----------------------------------------------------
4588 See the note above about the 4.44 and 4.50 releases.
4589 ----------------------------------------------------
4595 1. Change 4.43/35 introduced a bug that caused file counts to be
4596 incorrectly computed when quota_filecount was set in an appendfile
4599 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4600 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4603 3. Give more explanation in the error message when the command for a transport
4604 filter fails to execute.
4606 4. There are several places where Exim runs a non-Exim command in a
4607 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4608 was being done only for the command run by the queryprogram router. It is
4609 now done for all such subprocesses. The other cases are: ${run, transport
4610 filters, and the commands run by the lmtp and pipe transports.
4612 5. Some older OS have a limit of 256 on the maximum number of file
4613 descriptors. Exim was using setrlimit() to set 1000 as a large value
4614 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4615 systems. I've change it so that if it can't get 1000, it tries for 256.
4617 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4618 was an oversight, and furthermore, ever since the addition of extra
4619 controls (e.g. 4.43/32), the checks on when to allow different forms of
4620 "control" were broken. There should now be diagnostics for all cases when a
4621 control that does not make sense is encountered.
4623 7. $recipients is now available in the predata ACL (oversight).
4625 8. Tidy the search cache before the fork to do a delivery from a message
4626 received from the command line. Otherwise the child will trigger a lookup
4627 failure and thereby defer the delivery if it tries to use (for example) a
4628 cached ldap connection that the parent has called unbind on.
4630 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4631 of $address_data from the recipient verification was clobbered by the
4632 sender verification.
4634 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4635 was its contents. (It was OK if the option was not defined at all.)
4637 11. A "Completed" log line is now written for messages that are removed from
4638 the spool by the -Mrm option.
4640 12. $host_address is now set to the target address during the checking of
4641 ignore_target_hosts.
4643 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4644 being passed; this would have caused $sender_host_name to have been used if
4645 matching the list had actually called for a host name (not very likely,
4646 since this list is usually IP addresses). A host name is now passed as
4649 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4650 code that specifies a non-restarting handler (typically sigaction() in
4651 modern systems) in an attempt to fix a rare and obscure crash bug.
4653 15. Narrowed the window for a race in the daemon that could cause it to ignore
4654 SIGCHLD signals. This is not a major problem, because they are used only to
4655 wake it up if nothing else does.
4657 16. A malformed maildirsize file could cause Exim to calculate negative values
4658 for the mailbox size or file count. Odd effects could occur as a result.
4659 The maildirsize information is now recalculated if the size or filecount
4662 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4663 support for a long time. Removed HAVE_SYS_VFS_H.
4665 18. Updated exipick to current release from John Jetmore.
4667 19. Allow an empty sender to be matched against a lookup in an address list.
4668 Previously the only cases considered were a regular expression, or an
4671 20. Exim went into a mad DNS lookup loop when doing a callout where the
4672 host was specified on the transport, if the DNS lookup yielded more than
4675 21. The RFC2047 encoding function was originally intended for short strings
4676 such as real names; it was not keeping to the 75-character limit for
4677 encoded words that the RFC imposes. It now respects the limit, and
4678 generates multiple encoded words if necessary. To be on the safe side, I
4679 have increased the buffer size for the ${rfc2047: expansion operator from
4682 22. Failure to deliver a bounce message always caused it to be frozen, even if
4683 there was an errors_to setting on the router. The errors_to setting is now
4686 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4687 canonical form (fully expanded) before being placed in
4688 $sender_host_address.
4690 24. Updated eximstats to version 1.33
4692 25. Include certificate and key file names in error message when GnuTLS fails
4693 to set them up, because the GnuTLS error message doesn't include the name
4694 of the failing file when there is a problem reading it.
4696 26. Expand error message when OpenSSL has problems setting up cert/key files.
4699 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4700 (this can affect the format of dates).
4702 28. exim_tidydb, when checking for the continued existence of a message for
4703 which it has found a message-specific retry record, was not finding
4704 messages that were in split spool directories. Consequently, it was
4705 deleting retry records that should have stayed in existence.
4707 29. eximstats updated to version 1.35
4708 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4709 1.35 - bugfix such that pie charts by volume are generated correctly
4711 30. The SPA authentication driver was not abandoning authentication and moving
4712 on to the next authenticator when an expansion was forced to fail,
4713 contradicting the general specification for all authenticators. Instead it
4714 was generating a temporary error. It now behaves as specified.
4716 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4717 (the order specifies the preference for clients). The order is now AES256,
4718 AES128, 3DES, ARCFOUR128.
4720 31. Small patch to Sieve code - explicitly set From: when generating an
4723 32. Exim crashed if a remote delivery caused a very long error message to be
4724 recorded - for instance if somebody sent an entire SpamAssassin report back
4725 as a large number of 550 error lines. This bug was coincidentally fixed by
4726 increasing the size of one of Exim's internal buffers (big_buffer) that
4727 happened as part of the Exiscan merge. However, to be on the safe side, I
4728 have made the code more robust (and fixed the comments that describe what
4731 33. Some experimental protocols are using DNS PTR records for new purposes. The
4732 keys for these records are domain names, not reversed IP addresses. The
4733 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4734 leaves it alone. Component reversal etc. now happens only for IP addresses.
4737 34. The host_aton() function is supposed to be passed a string that is known
4738 to be a valid IP address. However, in the case of IPv6 addresses, it was
4739 not checking this. This is a hostage to fortune. Exim now panics and dies
4740 if the condition is not met. A case was found where this could be provoked
4741 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4742 components; fortuitously, this particular loophole had already been fixed
4743 by change 4.50/55 or 4.44/33 above.
4745 If there are any other similar loopholes, the new check in host_aton()
4746 itself should stop them being exploited. The report I received stated that
4747 data on the command line could provoke the exploit when Exim was running as
4748 exim, but did not say which command line option was involved. All I could
4749 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4750 running as the user.
4753 35. There was a buffer overflow vulnerability in the SPA authentication code
4754 (which came originally from the Samba project). I have added a test to the
4755 spa_base64_to_bits() function which I hope fixes it.
4758 36. The daemon start-up calls getloadavg() while still root for those OS that
4759 need the first call to be done as root, but it missed one case: when
4760 deliver_queue_load_max is set with deliver_drop_privilege. This is
4761 necessary for the benefit of the queue runner, because there is no re-exec
4762 when deliver_drop_privilege is set.
4764 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4765 was in use, was not putting the data itself into the right store pool;
4766 consequently, it could be overwritten for a subsequent message in the same
4767 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4770 38. Sometimes the final signoff response after QUIT could fail to get
4771 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4772 < 0 before doing a fflush(). This bug looks as though it goes back to the
4773 introduction of TLS in release 3.20, but "sometimes" must have been rare
4774 because the tests only now provoked it.
4780 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4781 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4782 Now it returns an integer. A number of calls were still expecting a T/F
4783 return. Fortuitously, in all cases, the tests worked in OK situations,
4784 which is the norm. However, things would have gone wrong on any write
4785 failures on the smtp file descriptor. This function is used when sending
4786 messages over SMTP and also when doing verify callouts.
4788 2. When Exim is called to do synchronous delivery of a locally submitted
4789 message (the -odf or -odi options), it no longer closes stderr before doing
4792 3. Implemented the mua_wrapper option.
4794 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4796 5. Implemented the functions header_remove(), header_testname(),
4797 header_add_at_position(), and receive_remove_recipient(), and exported them
4800 6. If an ACL "warn" statement specified the addition of headers, Exim already
4801 inserted X-ACL-Warn: at the start if there was no header name. However, it
4802 was not making this test for the second and subsequent header lines if
4803 there were newlines in the string. This meant that an invalid header could
4804 be inserted if Exim was badly configured.
4806 7. Allow an ACL "warn" statement to add header lines at the start or after all
4807 the Received: headers, as well as at the end.
4809 8. Added the rcpt_4xx retry error code.
4811 9. Added postmaster_mailfrom=xxx to callout verification option.
4813 10. Added mailfrom=xxxx to the callout verification option, for verify=
4816 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4817 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4819 12. Inserted some casts to stop certain compilers warning when using pointer
4820 differences as field lengths or precisions in printf-type calls (mostly
4821 affecting debugging statements).
4823 13. Added optional readline() support for -be (dynamically loaded).
4825 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4826 same clock tick as a message's arrival, so that its received time was the
4827 same as the "first fail" time on the retry record, and that message
4828 remained on the queue past the ultimate address timeout, every queue runner
4829 would try a delivery (because it was past the ultimate address timeout) but
4830 after another failure, the ultimate address timeout, which should have then
4831 bounced the address, did not kick in. This was a "< instead of <=" error;
4832 in most cases the first failure would have been in the next clock tick
4833 after the received time, and all would be well.
4835 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4836 being recognized when the domain list was tested by the match_domain
4837 condition in an expansion string.
4839 16. Added the ${str2b64: operator.
4841 17. Exim was always calling setrlimit() to set a large limit for the number of
4842 processes, without checking whether the existing limit was already
4843 adequate. (It did check for the limit on file descriptors.) Furthermore,
4844 errors from getrlimit() and setrlimit() were being ignored. Now they are
4845 logged to the main and panic logs, but Exim does carry on, to try to do its
4846 job under whatever limits there are.
4848 18. Imported PCRE 5.0.
4850 19. Trivial typo in log message " temporarily refused connection" (the leading
4853 20. If the log selector return_path_on_delivery was set and an address was
4854 redirected to /dev/null, the delivery process crashed because it assumed
4855 that a return path would always be set for a "successful" delivery. In this
4856 case, the whole delivery is bypassed as an optimization, and therefore no
4859 21. Internal re-arrangement: the function for sending a challenge and reading
4860 a response while authentication was assuming a zero-terminated challenge
4861 string. It's now changed to take a pointer and a length, to allow for
4862 binary data in such strings.
4864 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4866 23. Exim was not respecting finduser_retries when seeking the login of the
4867 uid under which it was called; it was always trying 10 times. (The default
4868 setting of finduser_retries is zero.) Also, it was sleeping after the final
4869 failure, which is pointless.
4871 24. Implemented tls_on_connect_ports.
4873 25. Implemented acl_smtp_predata.
4875 26. If the domain in control=submission is set empty, Exim assumes that the
4876 authenticated id is a complete email address when it generates From: or
4877 Sender: header lines.
4879 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4880 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4881 chown and chgrp in /bin and hostname in /usr/bin.
4883 28. Exim was keeping the "process log" file open after each use, just as it
4884 does for the main log. This opens the possibility of it remaining open for
4885 long periods when the USR1 signal hits a daemon. Occasional processlog
4886 errors were reported, that could have been caused by this. Anyway, it seems
4887 much more sensible not to leave this file open at all, so that is what now
4890 29. The long-running daemon process does not normally write to the log once it
4891 has entered its main loop, and it closes the log before doing so. This is
4892 so that log files can straightforwardly be renamed and moved. However,
4893 there are a couple of unusual error situations where the daemon does write
4894 log entries, and I had neglected to close the log afterwards.
4896 30. The text of an SMTP error response that was received during a remote
4897 delivery was being truncated at 512 bytes. This is too short for some of
4898 the long messages that one sometimes sees. I've increased the limit to
4901 31. It is now possible to make retry rules that apply only when a message has a
4902 specific sender, in particular, an empty sender.
4904 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4905 it possible to be selective about when SMTP synchronization is enforced.
4907 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4909 32. Implemented hosts_connection_nolog.
4911 33. Added an ACL for QUIT.
4913 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4916 35. Added mailbox_size and mailbox_filecount to appendfile.
4918 36. Added control = no_multiline_responses to ACLs.
4920 37. There was a bug in the logic of the code that waits for the clock to tick
4921 in the case where the clock went backwards by a substantial amount such
4922 that the microsecond fraction of "now" was more than the microsecond
4923 fraction of "then" (but the whole seconds number was less).
4925 38. Added support for the libradius Radius client library this is found on
4926 FreeBSD (previously only the radiusclient library was supported).
4932 1. When certain lookups returned multiple values in the form name=value, the
4933 quoting of the values was not always being done properly. Specifically:
4934 (a) If the value started with a double quote, but contained no whitespace,
4936 (b) If the value contained whitespace other than a space character (i.e.
4937 tabs or newlines or carriage returns) it was not quoted.
4938 This fix has been applied to the mysql and pgsql lookups by writing a
4939 separate quoting function and calling it from the lookup code. The fix
4940 should probably also be applied to nisplus, ibase and oracle lookups, but
4941 since I cannot test any of those, I have not disturbed their existing code.
4943 2. A hit in the callout cache for a specific address caused a log line with no
4944 reason for rejecting RCPT. Now it says "Previous (cached) callout
4945 verification failure".
4947 3. There was an off-by-one bug in the queryprogram router. An over-long
4948 return line was truncated at 256 instead of 255 characters, thereby
4949 overflowing its buffer with the terminating zero. As well as fixing this, I
4950 have increased the buffer size to 1024 (and made a note to document this).
4952 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4953 when Exim is waiting for an SMTP response from a remote server, Exim
4954 restarts its select() call on the socket, thereby resetting its timeout.
4955 This is not a problem when such interrupts are rare. Somebody set up a cron
4956 job to run exiwhat every 2 minutes, which is less than the normal select()
4957 timeout (5 or 10 minutes). This meant that the select() timeout never
4958 kicked in because it was always reset. I have fixed this by comparing the
4959 time when an interrupt arrives with the time at the start of the first call
4960 to select(). If more time than the timeout has elapsed, the interrupt is
4961 treated as a timeout.
4963 5. Some internal re-factoring in preparation for the addition of Sieve
4964 extensions (by MH). In particular, the "personal" test is moved to a
4965 separate function, and given an option for scanning Cc: and Bcc: (which is
4966 not set for Exim filters).
4968 6. When Exim created an email address using the login of the caller as the
4969 local part (e.g. when creating a From: or Sender: header line), it was not
4970 quoting the local part when it contained special characters such as @.
4972 7. Installed new OpenBSD configuration files.
4974 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4975 try to make them clearer.
4977 9. Callout options, other than the timeout value, were being ignored when
4978 verifying sender addresses in header lines. For example, when using
4980 verify = header_sender/callout=no_cache
4982 the cache was (incorrectly) being used.
4984 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4985 only the Cygwin environment.
4987 11. When return_path_on_delivery was set as a log selector, if different remote
4988 addresses in the same message used different return paths and parallel
4989 remote delivery occurred, the wrong values would sometimes be logged.
4990 (Whenever a remote delivery process finished, the return path value from
4991 the most recently started remote delivery process was logged.)
4993 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4994 lines when AUTH and/or TLS are in use. This is the "received protocol"
4995 field. Exim used to use "asmtp" for authenticated SMTP, without any
4996 indication (in the protocol name) for TLS use. Now it follows the RFC and
4997 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4998 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4999 names appear in log lines as well as in Received: header lines.
5001 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5002 extensions, and comparison tests, and to fix some bugs.
5004 14. Changes to the "personal" filter test:
5006 (1) The test was buggy in that it was just doing the equivalent of
5007 "contains" tests on header lines. For example, if a user's address was
5008 anne@some.where, the "personal" test would incorrectly be true for
5010 To: susanne@some.where
5012 This test is now done by extracting each address from the header in turn,
5013 and checking the entire address. Other tests that are part of "personal"
5014 are now done using regular expressions (for example, to check local parts
5015 of addresses in From: header lines).
5017 (2) The list of non-personal local parts in From: addresses has been
5018 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5019 taken from the Sieve specification recommendations.
5021 (3) If the message contains any header line starting with "List-" it is
5022 treated as non-personal.
5024 (4) The test for "circular" in the Subject: header line has been removed
5025 because it now seems ill-conceived.
5027 15. Minor typos in src/EDITME comments corrected.
5029 16. Installed latest exipick from John Jetmore.
5031 17. If headers_add on a router specified a text string that was too long for
5032 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5033 of string_sprintf() is now avoided.
5035 18. $message_body_size was not set (it was always zero) when running the DATA
5036 ACL and the local_scan() function.
5038 19. For the "mail" command in an Exim filter, no default was being set for
5039 the once_repeat time, causing a random time value to be used if "once" was
5040 specified. (If the value happened to be <= 0, no repeat happened.) The
5041 default is now 0s, meaning "never repeat". The "vacation" command was OK
5042 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5043 (I found it when inspecting the code).
5045 20. There is now an overall timeout for performing a callout verification. It
5046 defaults to 4 times the callout timeout, which applies to individual SMTP
5047 commands during the callout. The overall timeout applies when there is more
5048 than one host that can be tried. The timeout is checked before trying the
5049 next host. This prevents very long delays if there are a large number of
5050 hosts and all are timing out (e.g. when the network connections are timing
5051 out). The value of the overall timeout can be changed by specifying an
5052 additional sub-option for "callout", called "maxwait". For example:
5054 verify = sender/callout=5s,maxwait=20s
5056 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5057 to the end before writing, but this should make it even safer).
5059 22. Exim was forgetting that it had advertised PIPELINING for the second and
5060 subsequent messages on an SMTP connection. It was also not resetting its
5061 memory on STARTTLS and an internal HELO.
5063 23. When Exim logs an SMTP synchronization error within a session, it now
5064 records whether PIPELINING has been advertised or not.
5066 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5067 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5068 rather than long int.
5070 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5072 26. Added the never_mail option to autoreply.
5078 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5079 crash if the getsockname() call failed; this can happen if a connection is
5080 closed very soon after it is established. The problem was simply in the
5081 order in which certain operations were done, causing Exim to try to write
5082 to the SMTP stream before it had set up the file descriptor. The bug has
5083 been fixed by making things happen in the correct order.
5089 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5090 before the connection was closed, thus losing the rejection response.
5092 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5093 some early Solaris releases, but causes trouble in current releases where
5094 socklen_t is defined.
5096 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5099 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5102 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5103 that releasing the top of it at the end releases what was used for sub-
5104 expansions (unless the block got too big). However, discard this block if
5105 the first thing is a variable or header, so that we can use its block when
5106 it is dynamic (useful for very large $message_headers, for example).
5108 6. Lookups now cache *every* query, not just the most recent. A new, separate
5109 store pool is used for this. It can be recovered when all lookup caches are
5110 flushed. Lookups now release memory at the end of their result strings.
5111 This has involved some general refactoring of the lookup sources.
5113 7. Some code has been added to the store_xxx() functions to reduce the amount
5114 of flapping under certain conditions.
5116 8. log_incoming_interface used to affect only the <= reception log lines. Now
5117 it causes the local interface and port to be added to several more SMTP log
5118 lines, for example "SMTP connection from", and rejection lines.
5120 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5122 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5124 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5125 could be overwritten at the end of the current message (or the start of a
5126 new message if it was set in a HELO ACL). The value is now preserved for
5127 the duration of the SMTP connection.
5129 12. If a transport had a headers_rewrite setting, and a matching header line
5130 contained an unqualified address, that address was qualified, even if it
5131 did not match any rewriting rules. The underlying bug was that the values
5132 of the flags that permit the existence of unqualified sender and recipient
5133 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5134 non-local messages, and by -bnq for local messages) were not being
5135 preserved with the message after it was received.
5137 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5138 "next input=" as part of the text comprising the host identity instead of
5139 the correct text. The code was using the same buffer for two different
5140 strings. However, depending on which order the printing function evaluated
5141 its arguments, the bug did not always show up. Under Linux, for example, my
5142 test suite worked just fine.
5144 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5145 doesn't work with some older versions of Perl. It has been changed to "my",
5146 which in any case is probably the better facility to use.
5148 15. A really picky compiler found some instances of statements for creating
5149 error messages that either had too many or two few arguments for the format
5152 16. The size of the buffer for calls to the DNS resolver has been increased
5153 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5154 for addresses that have a lot of PTR records. This alleviates a problem; it
5155 does not fully solve it.
5157 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5158 buffer now truncates the list and logs the incident, which is the same
5159 action as happens when Exim is looking up a host name and its aliases.
5160 Previously in this situation something unpredictable would happen;
5161 sometimes it was "internal error: store_reset failed".
5163 18. If a server dropped the connection unexpectedly when an Exim client was
5164 using GnuTLS and trying to read a response, the client delivery process
5165 crashed while trying to generate an error log message.
5167 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5168 string, for example:
5170 warn message = H1: something\nH2: something
5172 the text was added as a single header line from Exim's point of view
5173 though it ended up OK in the delivered message. However, searching for the
5174 second and subsequent header lines using $h_h2: did not work. This has been
5175 fixed. Similarly, if a system filter added multiple headers in this way,
5176 the routers could not see them.
5178 20. Expanded the error message when iplsearch is called with an invalid key to
5179 suggest using net-iplsearch in a host list.
5181 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5182 ACLs are no longer actually imposed (and a message to that effect is
5185 22. If a "gecos" field in a passwd entry contained escaped characters, in
5186 particular, if it contained a \" sequence, Exim got it wrong when building
5187 a From: or a Sender: header from that name. A second bug also caused
5188 incorrect handling when an unquoted " was present following a character
5189 that needed quoting.
5191 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5192 was not being matched caselessly.
5194 24. Arranged for all hyphens in the exim.8 source to be escaped with
5197 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5198 properly. Recipient callout cache records were still being keyed to include
5199 the sender, even when use_sender was set false. This led to far more
5200 callouts that were necessary. The sender is no longer included in the key
5201 when use_sender is false.
5203 26. Added "control = submission" modifier to ACLs.
5205 27. Added the ${base62d: operator to decode base 62 numbers.
5207 28. dnsdb lookups can now access SRV records.
5209 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5210 the configuration file.
5212 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5213 (-v) mode. This makes the output for a verbose queue run more intelligible.
5215 31. Added a use_postmaster feature to recipient callouts.
5217 32. Added the $body_zerocount variable, containing the number of binary zero
5218 bytes in the message body.
5220 33. The time of last modification of the "new" subdirectory is now used as the
5221 "mailbox time last read" when there is a quota error for a maildir
5224 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5226 35. Added +ignore_unknown as a special item in host lists.
5228 36. Code for decoding IPv6 addresses in host lists is now included, even if
5229 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5230 address was recognized as an IP address, but was then not correctly decoded
5231 into binary, causing unexpected and incorrect effects when compared with
5238 1. Very minor rewording of debugging text in manualroute to say "list of
5239 hosts" instead of "hostlist".
5241 2. If verify=header_syntax was set, and a header line with an unqualified
5242 address (no domain) and a large number of spaces between the end of the
5243 name and the colon was received, the reception process suffered a buffer
5244 overflow, and (when I tested it) crashed. This was caused by some obsolete
5245 code that should have been removed. The fix is to remove it!
5247 3. When running in the test harness, delay a bit after writing a bounce
5248 message to get a bit more predictability in the log output.
5250 4. Added a call to search_tidyup() just before forking a reception process. In
5251 theory, someone could use a lookup in the expansion of smtp_accept_max_
5252 per_host which, without the tidyup, could leave open a database connection.
5254 5. Added the variables $recipient_data and $sender_data which get set from a
5255 lookup success in an ACL "recipients" or "senders" condition, or a router
5256 "senders" option, similar to $domain_data and $local_part_data.
5258 6. Moved the writing of debug_print from before to after the "senders" test
5261 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5262 problems for message scanning, either using a data ACL, or using
5263 local_scan() because the Received: header was not generated till after they
5264 were called (in order to set the time as the time of reception completion).
5265 I have revised the way this works. The header is now generated after the
5266 body is received, but before the ACL or local_scan() are called. After they
5267 are run, the timestamp in the header is updated.
5273 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5274 before starting a queue runner without re-exec. This happened only when
5275 deliver_drop_privilege was set or when the Exim user was set to root. The
5276 effect of the bug was that timeouts during subsequent deliveries caused
5277 crashes instead of being properly handled. The handler is now left at its
5278 default (and expected) setting.
5280 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5281 message, again when deliver_drop_privilege is set or Exim is run as root.
5282 The bug described in (1) was not present in this case, but the tidying up
5283 of the other signals was missing. I have made the two cases consistent.
5285 3. The ignore_target_hosts setting on a manualroute router was being ignored
5286 for hosts that were looked up using the /MX notation.
5288 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5291 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5292 operated on the sender address. After changing the $sender_address to <>
5293 for the sender address verify, Exim was re-instated it as the original
5294 (before rewriting) address, but remembering that it had rewritten it, so it
5295 wasn't rewriting it again. This bug also had the effect of breaking the
5296 sender address verification caching when the sender address was rewritten.
5298 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5299 This has been changed so that if the ip literal address matches
5300 ignore_target_hosts, the router declines.
5302 7. Added expansion conditions match_domain, match_address, and match_local_
5303 part (NOT match_host).
5305 8. The placeholder for the Received: header didn't have a length field set.
5307 9. Added code to Exim itself and to exim_lock to test for a specific race
5308 condition that could lead to file corruption when using MBX delivery. The
5309 issue is with the lockfile that is created in /tmp. If this file is removed
5310 after a process has opened it but before that process has acquired a lock,
5311 there is the potential for a second process to recreate the file and also
5312 acquire a lock. This could lead to two Exim processes writing to the file
5313 at the same time. The added code performs the same test as UW imapd; it
5314 checks after acquiring the lock that its file descriptor still refers to
5315 the same named file.
5317 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5318 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5321 11. Added the smtp_active_hostname option. If used, this will typically be made
5322 to depend on the incoming interface address. Because $interface_address is
5323 not set up until the daemon has forked a reception process, error responses
5324 that can happen earlier (such as "too many connections") no longer contain
5327 12. If an expansion in a condition on a "warn" statement fails because a lookup
5328 defers, the "warn" statement is abandoned, and the next ACL statement is
5329 processed. Previously this caused the whole ACL to be aborted.
5331 13. Added the iplsearch lookup type.
5333 14. Added ident_timeout as a log selector.
5335 15. Added tls_certificate_verified as a log selector.
5337 16. Added a global option tls_require_ciphers (compare the smtp transport
5338 option of the same name). This controls incoming TLS connections.
5340 17. I finally figured out how to make tls_require_ciphers do a similar thing
5341 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5342 before starting the TLS session.
5344 18. Tabs are now shown as \t in -bP output.
5346 19. If the log selector return_path_on_delivery was set, Exim crashed when
5347 bouncing a message because it had too many Received: header lines.
5349 20. If two routers both had headers_remove settings, and the first one included
5350 a superfluous trailing colon, the final name in the first list and the
5351 first name in the second list were incorrectly joined into one item (with a
5352 colon in the middle).
5358 1. Added -C and -D options to the exinext utility, mainly to make it easier
5359 to include in the automated testing, but these could be helpful when
5360 multiple configurations are in use.
5362 2. The exinext utility was not formatting the output nicely when there was
5363 an alternate port involved in the retry record key, nor when there was a
5364 message id as well (for retries that were specific to a specific message
5365 and a specific host). It was also confused by IPv6 addresses, because of
5366 the additional colons they contain. I have fixed the IPv4 problem, and
5367 patched it up to do a reasonable job for IPv6.
5369 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5370 delivery, the log line now contains "pipelined" if PIPELINING was used.
5372 4. An SMTP transport process used to panic and die if the bind() call to set
5373 an explicit outgoing interface failed. This has been changed; it is now
5374 treated in the same way as a connect() failure.
5376 5. A reference to $sender_host_name in the part of a conditional expansion
5377 that was being skipped was still causing a DNS lookup. This no longer
5380 6. The def: expansion condition was not recognizing references to header lines
5381 that used bh_ and bheader_.
5383 7. Added the _cache feature to named lists.
5385 8. The code for checking quota_filecount in the appendfile transport was
5386 allowing one more file than it should have been.
5388 9. For compatibility with Sendmail, the command line option
5396 and sets the incoming protocol and host name (for trusted callers). The
5397 host name and its colon can be omitted when only the protocol is to be set.
5398 Note the Exim already has two private options, -pd and -ps, that refer to
5399 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5400 "s", but I don't think that's a major issue.
5402 10. A number of refactoring changes to the code, none of which should affect
5405 (a) The number of logging options was getting close to filling up the
5406 32-bit word that was used as a bit map. I have split them into two classes:
5407 those that are passed in the argument to log_write(), and those that are
5408 only ever tested independently outside of that function. These are now in
5409 separate 32-bit words, so there is plenty of room for expansion again.
5410 There is no change in the user interface or the logging behaviour.
5412 (b) When building, for example, log lines, the code previously used a
5413 macro that called string_cat() twice, in order to add two strings. This is
5414 not really sufficiently general. Furthermore, there was one instance where
5415 it was actually wrong because one of the argument was used twice, and in
5416 one call a function was used. (As it happened, calling the function twice
5417 did not affect the overall behaviour.) The macro has been replaced by a
5418 function that can join an arbitrary number of extra strings onto a growing
5421 (c) The code for expansion conditions now uses a table and a binary chop
5422 instead of a serial search (which was left over from when there were very
5423 few conditions). Also, it now recognizes conditions like "pam" even when
5424 the relevant support is not compiled in: a suitably worded error message is
5425 given if an attempt is made to use such a condition.
5427 11. Added ${time_interval:xxxxx}.
5429 12. A bug was causing one of the ddress fields not to be passed back correctly
5430 from remote delivery subprocesses. The field in question was not being
5431 subsequently used, so this caused to problems in practice.
5433 13. Added new log selectors queue_time and deliver_time.
5435 14. Might have fixed a bug in maildirsizefile handling that threw up
5436 "unexpected character" debug warnings, and recalculated the data
5437 unnecessarily. In any case, I expanded the warning message to give more
5440 15. Added the message "Restricted characters in address" to the statements in
5441 the default ACL that block characters like @ and % in local parts.
5443 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5444 Three changes have been made:
5446 (a) There was a serious bug; a negative response to MAIL caused the whole
5447 recipient domain to be cached as invalid, thereby blocking all messages
5448 to all local parts at the same domain, from all senders. This bug has
5449 been fixed. The domain is no longer cached after a negative response to
5450 MAIL if the sender used is not empty.
5452 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5455 (c) A new callout option, "use_sender" has been added for people who want
5456 the modified behaviour.
5462 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5465 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5466 indeed breaks things for older releases.
5468 3. Added additional logging to the case where there is a problem reading data
5469 from a filter that is running in a subprocess using a pipe, in order to
5470 try to track down a specific problem.
5472 4. Testing facility fudge: when running in the test harness and attempting
5473 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5474 getting "No route to host". Convert this to a timeout.
5476 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5479 6. Some OS don't have socklen_t but use size_t instead. This affects the
5480 fifth argument of getsockopt() amongst other things. This is now
5481 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5482 can be set for individual OS. I have set it for SunOS5, OSF1, and
5483 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5484 some earlier ones do not.
5486 7. Change 4.30/15 was not doing the test caselessly.
5488 8. The standard form for an IPv6 address literal was being rejected by address
5489 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5490 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5491 this, as well as the form without the "IPv6" on the front (but only when
5492 address literals are enabled, of course).
5494 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5496 10. Exim crashed if a message with an empty sender address specified by -f
5497 encountered a router with an errors_to setting. This could be provoked only
5498 by a command such as
5502 where an empty string was supplied; "<>" did not hit this bug.
5504 11. Installed PCRE release 4.5.
5506 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5507 remained set. It is now erased.
5509 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5510 times from message ids (which are base 36 rather than the normal 62).
5512 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5513 were being counted as actual protocol errors, and logged if the log
5514 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5515 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5516 and DATA following a set of rejected RCPTs do not count as protocol errors.
5517 In other words, Exim assumes they were pipelined, though this may not
5518 actually be the case. Of course, in all cases the client gets an
5519 appropriate error code.
5521 15. If a lookup fails in an ACL condition, a message about the failure may
5522 be available; it is used if testing the ACL cannot continue, because most
5523 such messages specify what the cause of the deferral is. However, some
5524 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5525 that caused an old message to be retained and used if a later statement
5526 caused a defer, replacing the real cause of the deferral.
5528 16. If an IP address had so many PTR records that the DNS lookup buffer
5529 was not large enough to hold them, Exim could crash while trying to process
5530 the truncated data. It now detects and logs this case.
5532 17. Further to 4.21/58, another change has been made: if (and only if) the
5533 first line of a message (the first header line) ends with CRLF, a bare LF
5534 in a subsequent header line has a space inserted after it, so as not to
5535 terminate the header.
5537 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5538 unnecessarily, used atoi() instead of strtol(), and didn't check the
5539 termination when getting file sizes from file names by regex.
5541 19. Completely re-implemented the support for maildirsize files, in the light
5542 of a number of problems with the previous contributed implementation
5543 (4.30/29). In particular:
5545 . If the quota is zero, the maildirsize file is maintained, but no quota is
5548 . If the maildir directory does not exist, it is created before any attempt
5549 to write a maildirsize file.
5551 . The quota value in the file is just a cache; if the quota is changed in
5552 the transport, the new value overrides.
5554 . A regular expression is available for excluding directories from the
5557 20. The autoreply transport checks the characters in options that define the
5558 message's headers; it allows continued headers, but it was checking with
5559 isspace() after an embedded newline instead of explicitly looking for a
5562 21. If all the "regular" hosts to which an address was routed had passed their
5563 expiry times, and had not reached their retry times, the address was
5564 bounced, even if fallback hosts were defined. Now Exim should go on to try
5567 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5568 equivalent code in the SMTP transport. Some hosts send humungous responses
5569 to HELO/EHLO, more than 1024 it seems.
5571 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5572 gives compiler warnings in some environments. I've now done it "properly",
5575 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5576 (because of problems with the built-in one) was declared to return uschar *
5577 instead of char *, causing compiler failure.
5579 25. Fixed a file descriptor leak when processing alias/forward files.
5581 26. Fixed a minor format string issue in dbfn.c.
5583 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5585 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5586 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5589 29. When a sender address is verified, it is cached, to save repeating the test
5590 when there is more than one recipient in a message. However, when the
5591 verification involves a callout, it is possible for different callout
5592 options to be set for different recipients. It is too complicated to keep
5593 track of this in the cache, so now Exim always runs a verification when a
5594 callout is required, relying on the callout cache for the optimization.
5595 The overhead is duplication of the address routing, but this should not be
5598 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5599 to be verified with callout=postmaster, and the main callout worked but the
5600 postmaster check failed, the verification correctly failed. However, if a
5601 subsequent RCPT command asked for sender verification *without* the
5602 postmaster check, incorrect caching caused this verification also to fail,
5605 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5606 it was not caching the DNS options (qualify_single, search_parents) that
5607 were used when the lookup failed. A subsequent lookup with different
5608 options therefore always gave the same answer, though there were cases
5609 where it should not have. (Example: a "domains = !$mx_any" option on a
5610 dnslookup router: the "domains" option is always processed without any
5611 widening, but the router might have qualify_single set.) Now Exim uses the
5612 cached value only when the same options are set.
5614 32. Added John Jetmore's "exipick" utility to the distribution.
5616 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5617 than a timeout (e.g. a certificate is required, and is not provided), an
5618 Exim server now closes the connection immediately. Previously it waited for
5619 the client to close - but if the client is SSL, it seems that they each
5620 wait for each other, leading to a delay before one of them times out.
5622 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5623 maintained 0.8.x compatibility because I don't think many are using it, and
5624 it is clearly obsolete.
5626 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5629 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5630 client certificate was expired. A simple patch fixes this, though I don't
5631 understand the full logic of why the verify callback is called multiple
5634 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5635 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5636 which causes problems with some clients (such as the Certicom SSL Plus
5637 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5638 disables the coutermeasure allowing Eudora to connect."
5640 38. Exim was not checking that a write() to a log file succeeded. This could
5641 lead to Bad Things if a log got too big, in particular if it hit a file
5642 size limit. Exim now panics and dies if it cannot write to a log file, just
5643 as it does if it cannot open a log file.
5645 39. Modified OS/Makefile-Linux so that it now contains
5647 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5649 The two -D definitions ensure that Exim is compiled with large file
5650 support, which makes it possible to handle log files that are bigger than
5653 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5654 instance) a domain was checked against a named list that involved a lookup,
5655 causing $domain_data to be set, then another domain was checked against the
5656 same list, then the first domain was re-checked, the value of $domain_data
5657 after the final check could be wrong. In particular, if the second check
5658 failed, it could be set empty. This bug probably also applied to
5661 41. The strip_trailing_dot option was not being applied to the address given
5662 with the -f command-line option.
5664 42. The code for reading a message's header from the spool was incrementing
5665 $received_count, but never initializing it. This meant that the value was
5666 incorrect (doubled) while delivering a message in the same process in which
5667 it was received. In the most common configuration of Exim, this never
5668 happens - a fresh exec is done - but it can happen when
5669 deliver_drop_privilege is set.
5671 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5672 it now includes up to 150 characters of the unexpected data in the log
5675 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5676 and building data strings. The size of both of these buffers was 10 000
5677 bytes - far larger than anybody would *ever* want, thought I. Needless to
5678 say, somebody hit the limit. I have increased the maximum line length to
5679 20 000 and the maximum data length of concatenated lines to 100 000. I have
5680 also fixed two bugs, because there was no checking on these buffers. Tsk,
5681 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5682 buffer is too small.
5684 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5685 lsearch lookups. Now it does.
5687 46. When parsing a route_list item in a manualroute router, a fixed-length
5688 buffer was used for the list of hosts. I made this 1024 bytes long,
5689 thinking that nobody would ever have a list of hosts that long. Wrong.
5690 Somebody had a whole pile of complicated expansion conditions, and the
5691 string was silently truncated, leading to an expansion error. It turns out
5692 that it is easier to change to an unlimited length (owing to other changes
5693 that have happened since this code was originally written) than to build
5694 structure for giving a limitation error. The length of the item that
5695 expands into the list of hosts is now unlimited.
5697 47. The lsearch lookup could not handle data where the length of text line was
5698 more than 4095 characters. Such lines were truncated, leading to shortened
5699 data being returned. It should now handle lines of any length.
5701 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5702 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5705 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5706 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5707 no longer happens in comment lines. A list of replacements is now placed
5708 at the head of all of the source files, except those whose only change is
5709 to replace PERL_COMMAND in the very first #! line.
5711 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5712 messages on the queue, with a bottom-up merge sort, using code contributed
5713 by Michael Haardt. This should make operations like -bp somewhat faster on
5714 large queues. It won't affect queue runners, except when queue_run_in_order
5717 51. Installed eximstats 1.31 in the distribution.
5719 52. Added support for SRV lookups to the dnslookup router.
5721 53. If an ACL referred to $message_body or $message_body_end, the value was not
5722 reset for any messages that followed in the same SMTP session.
5724 54. The store-handling optimization for building very long strings was not
5725 differentiating between the different store pools. I don't think this
5726 actually made any difference in practice, but I've tidied it.
5728 55. While running the routers to verify a sender address, $sender_address
5729 was still set to the sender address. This is wrong, because when routing to
5730 send a bounce to the sender, it would be empty. Therefore, I have changed
5731 it so that, while verifying a sender address, $sender_address is set to <>.
5732 (There is no change to what happens when verifying a recipient address.)
5734 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5735 target A or AAAA records (if not already returned) without resetting the
5736 qualify_single or search_parents options of the DNS resolver. These are
5737 inappropriate in this case because the targets of MX and SRV records must
5738 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5739 target that, when qualified, matched something in the local domain. These
5740 two options are now turned off when doing these lookups.
5742 57. It seems that at least some releases of Reiserfs (which does not have the
5743 concept of a fixed number of inodes) returns zero and not -1 for the
5744 number of available inodes. This interacted badly with check_spool_inodes,
5745 which assumed that -1 was the "no such thing" setting. What I have done is
5746 to check that the total number of inodes is greater than zero before doing
5747 the test of how many are available.
5749 58. When a "warn" ACL statement has a log_message modifier, the message is
5750 remembered, and not repeated. This is to avoid a lot of repetition when a
5751 message has many recipients that cause the same warning to be written.
5752 However, Exim was preserving the list of already written lines for an
5753 entire SMTP session, which doesn't seem right. The memory is now reset if a
5754 new message is started.
5756 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5757 debugging output unless log_rewrite was also set.
5759 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5760 of (int)(handle) when we know that handle contains (void *)(-1).
5762 61. The Exim daemon panic-logs an error return when it closes the incoming
5763 connection. However "connection reset by peer" seems to be common, and
5764 isn't really an error worthy of noting specially, so that particular error
5767 62. When Exim is trying to find all the local interfaces, it used to panic and
5768 die if the ioctl to get the interface flags failed. However, it seems that
5769 on at least one OS (Solaris 9) it is possible to have an interface that is
5770 included in the list of interfaces, but for which you get a failure error
5771 for this call. This happens when the interface is not "plumbed" into a
5772 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5773 failure of the "get flags" call assumes that the interface is down.
5775 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5776 makes life easier for people who are doing arithmetic on fields extracted
5777 from dates, where you often get leading zeros that should not be
5778 interpreted as octal.
5780 64. Added qualify_domain to the redirect router, to override the global
5783 65. If a pathologically long header line contained very many addresses (the
5784 report of this problem mentioned 10 000) and each of them was rewritten,
5785 Exim could use up a very large amount of memory. (It kept on making new
5786 copies of the header line as it rewrote, and never released the old ones.)
5787 At the expense of a bit more processing, the header rewriting function has
5788 been changed so that it no longer eats memory in this way.
5790 66. The generation of the Received: header has been moved from the time that a
5791 message starts to be received, to the time that it finishes. The timestamp
5792 in the Received: header should now be very close to that of the <= log
5793 line. There are two side-effects of this change:
5795 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5796 logged header lines no longer include the local Received: line, because
5797 it has not yet been created. The same applies to a copy of the message
5798 that is returned to a non-SMTP sender when a message is rejected.
5800 (b) When a filter file is tested using -bf, no additional Received: header
5801 is added to the test message. After some thought, I decided that this
5804 This change does not affect the value of $received_for. It is still set
5805 after address rewriting, but before local_scan() is called.
5807 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5809 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5810 gave an unhelpful panic error message, and a defer error. I have managed to
5811 change this behaviour so that it now rejects any supplied certificate,
5812 which seems right, as the list of acceptable certificates is empty.
5814 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5815 gave an unhelpful defer error. I have not managed to make this reject any
5816 supplied certificates, but the error message it gives is "no certificate
5817 supplied", which is not helpful.
5819 70. exigrep's output now also includes lines that are not associated with any
5820 message, but which match the given pattern. Implemented by a patch from
5821 Martin Sluka, which also tidied up the Perl a bit.
5823 71. Recipient callout verification, like sender verification, was using <> in
5824 the MAIL FROM command. This isn't really the right thing, since the actual
5825 sender may affect whether the remote host accepts the recipient or not. I
5826 have changed it to use the actual sender in the callout; this means that
5827 the cache record is now keyed on a recipient/sender pair, not just the
5828 recipient address. There doesn't seem to be a real danger of callout loops,
5829 since a callout by the remote host to check the sender would use <>.
5830 [SEE ABOVE: changed after hitting problems.]
5832 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5833 temporary errors. However, in the case of such a code being given after
5834 the end of a data transmission (i.e. after ".") Exim was failing to write
5835 a retry record for the message. (Yes, there was some broken host that was
5836 actually sending 8xx at this point.)
5838 73. An unknown lookup type in a host list could cause Exim to panic-die when
5839 the list was checked. (An example that provoked this was putting <; in the
5840 middle of a list instead of at the start.) If this happened during a DATA
5841 ACL check, a -D file could be left lying around. This kind of configuration
5842 error no longer causes Exim to die; instead it causes a defer error. The
5843 incident is still logged to the main and panic logs.
5845 74. Buglet left over from Exim 3 conversion. The message "too many messages
5846 in one connection" was written to the rejectlog but not the mainlog, except
5847 when address rewriting (yes!) was being logged.
5849 75. Added write_rejectlog option.
5851 76. When a system filter was run not as root (that is, when system_filter_user
5852 was set), the values of the $n variables were not being returned to the
5853 main process; thus, they were not subsequently available in the $sn
5856 77. Added +return_path_on_delivery log selector.
5858 78. A connection timeout was being treated differently from recipients deferred
5859 when testing hosts_max_try with a message that was older than the host's
5860 retry timeout. (The host should not be counted, thus allowing all hosts to
5861 be tried at least once before bouncing.) This may have been the cause of an
5862 occasionally reported bug whereby a message would remain on the queue
5863 longer than the retry timeout, but would be bounced if a delivery was
5864 forced. I say "may" because I never totally pinned down the problem;
5865 setting up timeout/retry tests is difficult. See also the next item.
5867 79. The ultimate address timeout was not being applied to errors that involved
5868 a combination of host plus message (for example, a timeout on a MAIL
5869 command). When an address resolved to a number of possible hosts, and they
5870 were not all tried for each delivery (e.g. because of hosts_max_try), a
5871 message could remain on the queue longer than the retry timeout.
5873 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5876 81. Fixed an obscure SMTP outgoing bug which required at least the following
5877 conditions: (a) there was another message waiting for the same server;
5878 (b) the server returned 5xx to all RCPT commands in the first message so
5879 that the message was not completed; (c) the server dropped the connection
5880 or gave a negative response to the RSET that Exim sends to abort the
5881 transaction. The observed case was a dropped connection after DATA that had
5882 been sent in pipelining mode. That is, the server had advertised PIPELINING
5883 but was not implementing it correctly. The effect of the bug was incorrect
5884 behaviour, such as trying another host, and this could lead to a crash.
5890 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5891 and daemon.c were passed as pointers to ints; they should have been
5892 pointers to socklen_t variables (which are typically unsigned ints).
5894 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5897 3. Fixed a really odd bug that affected only the testing scheme; patching a
5898 certain fixed string in the binary changed the value of another string that
5899 happened to be identical to the end of the original first string.
5901 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5902 name", it returns that address as the IP address. On some operating
5903 systems (e.g. Solaris), it also passes back the IP address string as the
5904 "host name". However, on others (e.g. Linux), it passes back an empty
5905 string. Exim wasn't checking for this, and was changing the host name to an
5906 empty string, assuming it had been canonicalized.
5908 5. Although rare, it is permitted to have more than one PTR record for a given
5909 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5910 all the names associated with an address, because they do in Solaris.
5911 However, it seems that they do not in Linux for data that comes from the
5912 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5913 I found this out when I moved to a new Linux workstation and tried to run
5914 the Exim test suite.
5916 To get round this problem I have changed the code so that it now does its
5917 own call to the DNS to look up PTR records when searching for a host name.
5918 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5919 addresses that are only in /etc/hosts are still found.
5921 This behaviour is, however, controlled by an option called host_lookup_
5922 order, which defaults to "bydns:byaddr". If people want to use the other
5923 order, or indeed, just use one or the other means of lookup, they can
5924 specify it in this variable.
5926 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5927 some operating systems, this comes back from gethostbyaddr() as an empty
5928 string, and this is what Exim used to test for. However, it seems that in
5929 other systems, "." is yielded. Exim now tests for this case too.
5931 7. The values of check_spool_space and check_log_space are now held internally
5932 as a number of kilobytes instead of an absolute number of bytes. If a
5933 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5934 kilobyte. This means that much larger values can be stored.
5936 8. Exim monitor: an attempt to get the action menu when not actually pointing
5937 at a message produces an empty menu entitled "No message selected". This
5938 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5939 no entries in it ("Shell widget menu has zero width and/or height"). So I
5940 have added a single, blank menu entry in this case.
5942 9. Added ${quote_local_part.
5944 10. MIME decoding is now applied to the contents of Subject: header lines when
5947 11. Now that a reference to $sender_host_address automatically causes a reverse
5948 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5949 host lookup before query-style lookups in lists that might use this
5950 variable. This has therefore been abolished, and the "net-" prefix is no
5951 longer necessary for query-style lookups.
5953 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5954 have been a typo for LFLAGS, so it has been changed.
5956 13. The install script calls Exim with "-C /dev/null" in order to find the
5957 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5958 to be output. However, since Exim outputs its version number before the
5959 error, it didn't break the script. It just looked ugly. I fixed this by
5960 always allowing "-C /dev/null" if the caller is root.
5962 14. Ignore overlarge ACL variable number when reading spool file - insurance
5963 against a later release with more variables having written the file.
5965 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5966 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5967 this, as well as the form without the "IPv6" on the front.
5969 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5970 OS/Makefile-Darwin file.
5972 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5973 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5975 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5976 variable. However, some OS use ints for inodes. I've added cast to long int
5977 to get rid of the compiler warning.
5979 19. I had forgotten to lock out "/../" in configuration file names when
5980 ALT_CONFIG_PREFIX was set.
5982 20. Routers used for verification do not need to specify transports. However,
5983 if such a router generated a host list, and callout was configured, Exim
5984 crashed, because it could not find a port number from the (non-existent)
5985 transport. It now assumes port 25 in this circumstance.
5987 21. Added the -t option to exigrep.
5989 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5990 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5991 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5992 them is set, LOOKUP_LSEARCH is forced.
5994 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5996 24. Added sender and host information to the "rejected by local_scan()" log
5997 line; previously there was no indication of these.
5999 25. Added .include_if_exists.
6001 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6002 Linux. It turns out that not all file systems support this. Apparently some
6003 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6004 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6005 not supported on the file descriptor, is now ignored when Exim is trying to
6006 sync a directory. This applies only to Linux.
6008 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6010 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6011 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6014 29. Added support for maildirsize files from supplied patch (modified a bit).
6016 30. The use of :fail: followed by an empty string could lead Exim to respond to
6017 sender verification failures with (e.g.):
6019 550 Verification failed for <xxx>
6020 550 Sender verify failed
6022 where the first response line was missing the '-' that indicates it is not
6023 the final line of the response.
6025 31. The loop for finding the name of the user that called Exim had a hardwired
6026 limit of 10; it now uses the value of finduser_retries, which is used for
6027 all other user lookups.
6029 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6032 33. Exim was neglecting to zero errno before one call of strtol() when
6033 expanding a string and expecting an integer value. On some systems this
6034 resulted in spurious "integer overflow" errors. Also, it was casting the
6035 result into an int without checking.
6037 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6038 did not work. The code looks as if it has *never* worked, though it appears
6039 to have been documented since at least release 1.62. I have made it work.
6041 35. The "timeout_DNS" error in retry rules, also documented since at least
6042 1.62, also never worked. As it isn't clear exactly what this means, and
6043 clearly it isn't a major issue, I have abolished the feature by treating it
6044 as "timeout", and writing a warning to the main and panic logs.
6046 36. The display of retry rules for -brt wasn't always showing the error code
6049 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6050 timeout_connect_A, timeout_connect_MX.
6052 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6053 to the empty sender.
6055 39. The daemon was not analysing the content of -oX till after it had closed
6056 stderr and disconnected from the controlling terminal. This meant that any
6057 syntax errors were only noted on the panic log, and the return code from
6058 the command was 0. By re-arranging the code a little, I've made the
6059 decoding happen first, so such errors now appear on stderr, and the return
6060 code is 1. However, the actual setting up of the sockets still happens in
6061 the disconnected process, so errors there are still only recorded on the
6064 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6065 connections (as happens on some IP stacks) was logged at start up time as
6066 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6067 it from "IPv6 and IPv4", which means that two separate sockets are being
6070 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6071 says whether AF_INET or AF_INET6 was passed as an argument.
6073 42. Exiwhat output was messed up when time zones were included in log
6076 43. Exiwhat now gives more information about the daemon's listening ports,
6077 and whether -tls-on-connect was used.
6079 44. The "port" option of the smtp transport is now expanded.
6081 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6082 silently ignored. Now an error message is written to the main and panic
6085 46. There's a new ACL modifier called "logwrite" which writes to a log file
6086 as soon as it is encountered.
6088 47. Added $local_user_uid and $local_user_gid at routing time.
6090 48. Exim crashed when trying to verify a sender address that was being
6093 49. Exim was recognizing only a space character after ".include". It now also
6094 recognizes a tab character.
6096 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6097 extracting the relevant information from the specification. The man page no
6098 longer contains scrambled data for the -d option, and I've added a section
6099 at the front about calling Exim under different names.
6101 51. Added "extra_headers" argument to the "mail" command in filter files.
6103 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6106 53. Installed eximstats 1.29.
6108 54. Added transport_filter_timeout as a generic transport option.
6110 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6111 Cc: header lines. This was required by RFC 822, but it not required by RFC
6114 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6115 incoming messages that did not have them. Now it does so only if the
6116 message originates locally, that is, if there is no associated remote host
6117 address. When Resent- header lines are present, this applies to the Resent-
6118 lines rather than the non-Resent- lines.
6120 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6121 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6123 58. Messages for configuration errors now include the name of the main
6124 configuration file - useful now that there may be more than one file in a
6125 list (.included file names were always shown).
6127 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6128 for those rare installations that do not start the daemon as root or run it
6129 setuid root. I've cut out the call to initgroups() if the daemon is not
6132 60. The Exim user and group can now be bound into the binary as text strings
6133 that are looked up at the start of Exim's processing.
6135 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6137 62. Added $mailstore_basename variable.
6139 63. Installed patch to sieve.c from Michael Haardt.
6141 64. When Exim failed to open the panic log after failing to open the main log,
6142 the original message it was trying to log was written to stderr and debug
6143 output, but if they were not available (the usual case in production), it
6144 was lost. Now it is written to syslog before the two lines that record the
6145 failures to open the logs.
6147 65. Users' Exim filters run in subprocesses under the user's uid. It is
6148 possible for a "deliver" command or an alias in a "personal" command to
6149 provoke an address rewrite. If logging of address rewriting is configured,
6150 this fails because the process is not running as root or exim. There may be
6151 a better way of dealing with this, but for the moment (because 4.30 needs
6152 to be released), I have disabled address rewrite logging when running a
6153 filter in a non-root, non-exim process.
6159 1. The buildconfig auxiliary program wasn't quoting the value set for
6160 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6161 not defined. This bug was masked in 4.22 by the effect that was fixed in
6164 2. Some messages that were rejected after a message id was allocated were
6165 shown as "incomplete" by exigrep. It no longer does this for messages that
6166 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6168 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6169 have allow_domain_literals set, the ID did not get logged in the <= line.
6170 Domain literals are now always recognized in Message-ID: header lines.
6172 4. The first argument for a ${extract expansion item is the key name or field
6173 number. Leading and trailing spaces in this item were not being ignored,
6174 causing some misleading effects.
6176 5. When deliver_drop_privilege was set, single queue runner processes started
6177 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6178 same command in the process it spins off) were not dropping privilege.
6180 6. When the daemon running as "exim" started a queue runner, it always
6181 re-executed Exim in the spun-off process. This is a waste of effort when
6182 deliver_drop_privilege is set. The new process now just calls the
6183 queue-runner function directly.
6189 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6192 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6193 ignored. Though the use of -oP was forcing the writing of a pid file, it
6194 was always written to the default place.
6196 3. If the message "no IP address found for host xxxx" is generated during
6197 incoming verification, it is now followed by identification of the incoming
6198 connection (so you can more easily find what provoked it).
6200 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6202 5. Added some features to "harden" Exim a bit more against certain attacks:
6204 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6205 be put in Local/Makefile. This is like the never_users runtime option,
6206 but it cannot be overridden. The default setting is "root".
6208 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6209 prefix string with which any file named in a -C command line option
6212 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6213 is retained for -C and -D only if the caller of Exim is root. Without
6214 it, the exim user may also use -C and -D and retain privilege.
6216 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6217 command line option is disabled.
6219 6. Macro names set by the -D option must start with an upper case letter, just
6220 like macro names defined in the configuration file.
6222 7. Added "dereference=" facility to LDAP.
6224 8. Two instances of the typo "uknown" in the source files are fixed.
6226 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6227 the Configure-Makefile script screwed up while processing it.
6229 10. Incorporated PCRE 4.4.
6231 11. The SMTP synchronization check was not operating right at the start of an
6232 SMTP session. For example, it could not catch a HELO sent before the client
6233 waited for the greeting. There is now a check for outstanding input at the
6234 point when the greeting is written. Because of the duplex, asynchronous
6235 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6236 way, but not yet received, when the check is performed.
6238 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6239 on TCP/IP sockets, because this apparently causes some broken clients to
6242 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6243 unchanged) from the Cygwin maintainer.
6245 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6246 was supported instead of testing for mbx. Effectively a typo.
6248 15. The spa authenticator server code was not checking that the input it
6249 received was valid base64.
6251 16. The debug output line for the "set" modifier in ACLs was not showing the
6252 name of the variable that was being set.
6254 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6256 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6257 lookup on its own. Something else was needed to trigger the lookup. For
6258 example, a match in host_lookup or the need for a host name in a host list.
6259 Now, if $sender_host_name is referenced and the host name has not yet been
6260 looked up, a lookup is performed. If the lookup fails, the variable remains
6261 empty, and $host_lookup_failed is set to "1".
6263 19. Added "eqi" as a case-independent comparison operator.
6265 20. The saslauthd authentication condition could segfault if neither service
6266 nor realm was specified.
6268 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6269 error message that was logged was misleading, and incoming SMTP
6270 connections were dropped. The message is now more accurate, and temporary
6271 errors are given to SMTP connections.
6273 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6274 (except RSET) with a 421 error, until QUIT is received. However, it was
6275 failing to send a response to QUIT.
6277 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6278 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6279 ACL is now run after the helo_try_verify_hosts code.
6281 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6282 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6283 case-independent, so other case variants are also recognized). Apparently
6284 some systems use these upper case variants.
6286 25. If more than two messages were waiting for the same host, and a transport
6287 filter was specified for the transport, Exim sent two messages over the
6288 same TCP/IP connection, and then failed with "socket operation on non-
6289 socket" when it tried to send the third.
6291 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6293 27. The extern definition of crypt16() in expand.c was not being excluded when
6294 the OS had its own crypt16() function.
6296 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6297 as a preferred synonym for return_size_limit, both as an option and as an
6300 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6302 30. Changed the default configuration ACL to relax the local part checking rule
6303 for addresses that are not in any local domains. For these addresses,
6304 slashes and pipe symbols are allowed within local parts, but the sequence
6305 /../ is explicitly forbidden.
6307 31. SPA server authentication was not clearing the challenge buffer before
6310 32. log_message in a "warn" ACL statement was writing to the reject log as
6311 well as to the main log, which contradicts the documentation and doesn't
6312 seem right (because no rejection is happening). So I have stopped it.
6314 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6315 However, I am unable to do any testing of this.
6317 34. Fixed an infelicity in the appendfile transport. When checking directories
6318 for a mailbox, to see if any needed to be created, it was accidentally
6319 using path names with one or more superfluous leading slashes; tracing
6320 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6322 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6323 contents are added to the log line that is written for every discarded
6324 recipient. (Previously a log_message setting was ignored.)
6326 36. The ${quote: operator now quotes the string if it is empty.
6328 37. The install script runs exim in order to find its version number. If for
6329 some reason other than non-existence or emptiness, which it checks, it
6330 could not run './exim', it was installing it with an empty version number,
6331 i.e. as "exim-". This error state is now caught, and the installation is
6334 38. An argument was missing from the function that creates an error message
6335 when Exim fails to connect to the socket for saslauthd authentication.
6336 This could cause Exim to crash, or give a corrupted message.
6338 39. Added isip, isip4, and isip6 to ${if conditions.
6340 40. The ACL variables $acl_xx are now saved with the message, and can be
6341 accessed later in routers, transports, and filters.
6343 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6344 strings in the file are not string-expanded.
6346 42. If a MAIL command specified a SIZE value that was too large to fit into an
6347 int variable, the check against message_size_limit failed. Such values are
6348 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6349 day this will have to be increased, but I don't think I want to be around
6350 when emails are that large.
6357 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6358 iconv() is not standard in FreeBSD.
6360 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6361 IPv6 enabled. The observed symptom was a segmentation fault on return from
6362 the function os_common_find_running_interfaces() in src/os.c.
6364 3. In the check_special_case() function in daemon.c I had used "errno" as an
6365 argument name, which causes warnings on some systems. This was basically a
6366 typo, since it was named "eno" in the comments!
6368 4. The code that waits for the clock to tick (at a resolution of some fraction
6369 of a second) so as to ensure message-id uniqueness was always waiting for
6370 at least one whole tick, when it could have waited for less. [This is
6371 almost certainly not relevant at current processor speeds, where it is
6372 unlikely to ever wait at all. But we try to future-proof.]
6374 5. The function that sleeps for a time interval that includes fractions of a
6375 second contained a race. It did not block SIGALRM between setting the
6376 timer, and suspending (a couple of lines later). If the interval was short
6377 and the sigsuspend() was delayed until after it had expired, the suspension
6378 never ended. On busy systems this could lead to processes getting stuck for
6381 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6382 process, before it forks any delivery processes. The open lookup caching
6383 mechanism meant that the open file or database connection was passed into
6384 the delivery process. The problem was that delivery processes always tidy
6385 up cached lookup data. This could cause a problem for the next delivery
6386 process started by the queue runner, because the external queue runner
6387 process does not know about the closure. So the next delivery process
6388 still has data in the lookup cache. In the case of a file lookup, there was
6389 no problem because closing a file descriptor in a subprocess doesn't affect
6390 the parent. However, if the lookup was caching a connection to a database,
6391 the connection was closed, and the second delivery process was likely to
6392 see errors such as "PGSQL: query failed: server closed the connection
6393 unexpectedly". The problem has been fixed by closing all cached lookups
6394 in a queue runner before running a delivery process.
6396 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6397 seem to have the "const" qualifier which it has on other OS. I've
6400 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6401 *of the same type* (client or server) with the same public name that an
6402 error should be diagnosed.
6404 9. When Exim looked up a host name for an IP address, but failed to find the
6405 original IP address when looking up the host name (a safety check), it
6406 output the message "<ip address> does not match any IP for NULL", which was
6407 confusing, to say the least. The bug was that the host name should have
6408 appeared instead of "NULL".
6410 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6411 user that is built into the binary, and the -C or -D options is used, root
6412 privilege is dropped before the configuration file is read. In addition,
6413 logging is switched to stderr instead of the normal log files. If the
6414 configuration then re-defines the Exim user, the unprivileged environment
6415 is probably not what is expected, so Exim logs a panic warning message (but
6418 However, if deliver_drop_privilege is set, the unprivileged state may well
6419 be exactly what is intended, so the warning has been cut out in that case,
6420 and Exim is allowed to try to write to its normal log files.
6426 1. smtp_return_error_details was not giving details for temporary sender
6427 or receiver verification errors.
6429 2. Diagnose a configuration error if two authenticators have the same public
6432 3. Exim used not to create the message log file for a message until the first
6433 delivery attempt. This could be confusing when incoming messages were held
6434 for policy or load reasons. The message log file is now created at the time
6435 the message is received, and an initial "Received" line is written to it.
6437 4. The automatically generated man page for command line options had a minor
6438 bug that caused no ill effects; however, a more serious problem was that
6439 the procedure for building the man page automatically didn't always
6440 operate. Consequently, release 4.20 contains an out-of-date version. This
6441 shouldn't happen again.
6443 5. When building Exim with embedded Perl support, the script that builds the
6444 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6445 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6447 6. The freeze_tell option was not being used for messages that were frozen on
6448 arrival, either by an ACL or by local_scan().
6450 7. Added the smtp_incomplete_transaction log selector.
6452 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6453 was accepting AUTH without a new EHLO.
6455 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6456 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6459 10. Logging of TCP/IP connections (when configured) now happens in the main
6460 daemon process instead of the child process, so that the TCP/IP connection
6461 count is more accurate (but it can never be perfect).
6463 11. The use of "drop" in a nested ACL was not being handled correctly in the
6464 outer ACL. Now, if condition failure induced by the nested "drop" causes
6465 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6466 or "require"), the connection is dropped.
6468 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6469 that yield "discard" can now be used with an "accept" or a "discard" verb,
6470 but an error is generated for any others (because I can't see a useful way
6471 to define what should happen).
6473 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6474 are now processed in the same way as lines in the Exim configuration file.
6475 In particular, continuation lines are supported.
6477 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6479 15. Added -ti meaning -t -i.
6481 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6482 domains, and warn by logging if others are found.
6484 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6485 returned from GIFCONF when Exim is trying to find the list of interfaces on
6486 a host. The code in os.c has been modified to copy each ifreq to an aligned
6487 structure in all cases.
6489 Also, in some cases, the returned ifreq's were being copied to a 'struct
6490 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6491 means the last couple of bytes of an IPv6 address could be chopped if the
6492 ifreq contained only a normal sockaddr (14 bytes storage).
6494 18. Named domain lists were not supported in the hosts_treat_as_local option.
6495 An entry such as +xxxx was not recognized, and was treated as a literal
6498 19. Ensure that header lines added by a DATA ACL are included in the reject log
6499 if the ACL subsequently rejects the message.
6501 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6502 MD5 (which is deprecated).
6504 21. When testing a filter file using -bf, Exim was writing a message when it
6505 took the sender from a "From " line in the message, but it was not doing so
6506 when it took $return_path from a Return-Path: header line. It now does.
6508 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6509 with a valid header line field name (a series of printing characters
6510 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6512 23. Changed "disc" in the source to "disk" to conform to the documentation and
6513 the book and for uniformity.
6515 24. Ignore Sendmail's -Ooption=value command line item.
6517 25. When execve() failed while trying to run a command in a pipe transport,
6518 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6519 could be confused with a return value of 69 from the command itself. This
6520 has been changed to 127, the value the shell returns if it is asked to run
6521 a non-existent command. The wording for the related log line suggests a
6522 non-existent command as the problem.
6524 26. If received_header_text expands to an empty string, do not add a Received:
6525 header line to the message. (Well, it adds a token one on the spool, but
6526 marks it "old" so that it doesn't get used or transmitted.)
6528 27. Installed eximstats 1.28 (addition of -nt option).
6530 28. There was no check for failure on the call to getsockname() in the daemon
6531 code. This can fail if there is a shortage of resources on the system, with
6532 ENOMEM, for example. A temporary error is now given on failure.
6534 29. Contrary to the C standard, it seems that in some environments, the
6535 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6536 program. Exim now does this explicitly; it affects the formatting of
6537 timestamps using strftime().
6539 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6540 complaints. I've now initialized all the variables, to avoid this.
6542 32. Header lines added by a system filter were not being "seen" during
6543 transport-time rewrites.
6545 33. The info_callback() function passed to OpenSSL is set up with type void
6546 (*)(SSL *, int, int), as described somewhere. However, when calling the
6547 function (actually a macro) that sets it up, the type void(*)() is
6548 expected. I've put in a cast to prevent warnings from picky compilers.
6550 34. If a DNS black list lookup found a CNAME record, but there were no A
6551 records associated with the domain it pointed at, Exim crashed.
6553 35. If a DNS black list lookup returned more than one A record, Exim ignored
6554 all but the first. It now scans all returned addresses if a particular IP
6555 value is being sought. In this situation, the contents of the
6556 $dnslist_value variable are a list of all the addresses, separated by a
6559 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6560 to accept a host name and all its aliases if the forward lookup for any of
6561 them yielded the IP address of the incoming connection. Now it accepts only
6562 those names whose forward lookup yields the correct IP address. Any other
6563 names are discarded. This closes a loophole whereby a rogue DNS
6564 administrator could create reverse DNS records to break through a
6565 wildcarded host restriction in an ACL.
6567 37. If a user filter or a system filter that ran in a subprocess used any of
6568 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6569 the wrong values were passed to the pipe command ($thisaddress had the
6570 value of $0, $0 had the value of $1, etc). This bug was introduced by
6571 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6573 38. Improved the line breaking for long SMTP error messages from ACLs.
6574 Previously, if there was no break point between 40 and 75 characters, Exim
6575 left the rest of the message alone. Two changes have been made: (a) I've
6576 reduced the minimum length to 35 characters; (b) if it can't find a break
6577 point between 35 and 75 characters, it looks ahead and uses the first one
6578 that it finds. This may give the occasional overlong line, but at least the
6579 remaining text gets split now.
6581 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6582 file descriptors might be low, and that setting 1000 would always raise it.
6583 It turns out that in some environments, the limit is already over 1000 and
6584 that lowering it causes trouble. So now Exim takes care not to decrease it.
6586 40. When delivering a message, the value of $return_path is set to $sender_
6587 address at the start of routing (routers may change the value). By an
6588 oversight, this default was not being set up when an address was tested by
6589 -bt or -bv, which affected the outcome if any router or filter referred to
6592 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6593 log only when "message" or "log_message" are set. However, if one of the
6594 conditions was an address verification, or a call to a nested ACL, the
6595 messages generated by the underlying test were being passed through. This
6596 no longer happens. The underlying message is available in $acl_verify_
6597 message for both "message" and "log_message" expansions, so it can be
6598 passed through if needed.
6600 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6601 new expansion $bh_ to give the encoded byte string without charset
6602 translation. Translation happens only if iconv() is available; HAVE_ICONV
6603 indicates this at build time. HEADERS_CHARSET gives the charset to
6604 translate to; headers_charset can change it in the configuration, and
6605 "headers charset" can change it in an individual filter file.
6607 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6608 that creates RFC 2047 encoded "words" labels them as that charset instead
6609 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6610 expansion operator; (ii) when Exim creates a From: line for a local
6611 message; (iii) when a header line is rewritten to include a "phrase" part.
6613 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6614 buggy, causing it to skip the first lines of messages whose message ID
6615 ended in 'D'. This would not have bitten before Exim release 4.14, because
6616 message IDs were unlikely to end in 'D' before then. The effect was to have
6617 incorrect size information for certain domains.
6619 45. #include "config.h" was missing at the start of the crypt16.c module. This
6620 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6623 46. If there was a timeout during a "random" callout check, Exim treated it as
6624 a failure of the random address, and carried on sending RSET and the real
6625 address. If the delay was just some slowness somewhere, the response to the
6626 original RCPT would be taken as a response to RSET and so on, causing
6627 mayhem of various kinds.
6629 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6630 when I implemented it. It didn't allow for the fact that some option values
6631 may legitimately be negative (e.g. size_addition), and it didn't even do
6632 the right test for positive values.
6634 48. Domain names in DNS records are case-independent. Exim always looks them up
6635 in lower case. Some resolvers return domain names in exactly the case they
6636 appear in the zone file, that is, they may contain uppercase letters. Not
6637 all resolvers do this - some return always lower case. Exim was treating a
6638 change of case by a resolver as a change of domain, similar to a widening
6639 of a domain abbreviation. This triggered its re-routing code and so it was
6640 trying to route what was effectively the same domain again. This normally
6641 caused routing to fail (because the router wouldn't handle the domain
6642 twice). Now Exim checks for this case specially, and just changes the
6643 casing of the domain that it ultimately uses when it transmits the message
6646 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6649 50. If a filter generated a file delivery with a non-absolute name (possible if
6650 no home directory exists for the router), the forbid_file option was not
6653 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6654 the existing equality matching.
6656 52. Exim was using ints instead of ino_t variables in some places where it was
6657 dealing with inode numbers.
6659 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6660 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6661 called TMPDIR, and if it finds it is different, it changes its value.
6663 54. The smtp_printf() function is now made available to local_scan() so
6664 additional output lines can be written before returning. There is also an
6665 smtp_fflush() function to enable the detection of a dropped connection.
6666 The variables smtp_input and smtp_batched_input are exported to
6669 55. Changed the default runtime configuration: the message "Unknown user"
6670 has been removed from the ACL, and instead placed on the localuser router,
6671 using the cannot_route_message feature. This means that any verification
6672 failures that generate their own messages won't get overridden. Similarly,
6673 the "Unrouteable address" message that was in the ACL for unverifiable
6674 relay addresses has also been removed.
6676 56. Added hosts_avoid_esmtp to the smtp transport.
6678 57. The exicyclog script was not checking for the esoteric option
6679 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6680 will work only if exicyclog is run under the appropriate euid.
6682 58. Following a discussion on the list, the rules by which Exim recognises line
6683 endings on incoming messages have been changed. The -dropcr and drop_cr
6684 options are now no-ops, retained only for backwards compatibility. The
6685 following line terminators are recognized: LF CRLF CR. However, special
6686 processing applies to CR:
6688 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6689 nor a local message in the state where . is a terminator.
6691 (ii) If a bare CR is encountered in a header line, an extra space is added
6692 after the line terminator so as not to end the header. The reasoning
6693 behind this is that bare CRs in header lines are most likely either
6694 to be mistakes, or people trying to play silly games.
6696 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6697 was being incorrectly given as 18 bytes larger than it should have been.
6698 This is a VOB (very old bug).
6700 60. This may never have affected anything current, but just in case it has:
6701 When the local host is found other than at the start of a list of hosts,
6702 the local host, those with the same MX, and any that follow, are discarded.
6703 When the list in question was part of a longer list of hosts, the following
6704 hosts (not currently being processed) were also being discarded. This no
6705 longer happens. I'm not sure if this situation could ever has previously
6708 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6711 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6712 header. This is something that is recommended in a new Internet Draft, and
6713 is something that is documented as being done by Sendmail. There are two
6714 possible values. For messages generated by the autoreply transport, Exim
6717 Auto-Submitted: auto-replied
6719 whereas for all other generated messages (e.g. bounces) it adds
6721 Auto-Submitted: auto-generated
6723 63. The "personal" condition in filters now includes a test for the
6724 Auto-Submitted: header. If it contains the string "auto-" the message it
6725 not considered personal.
6727 64. Added rcpt_include_affixes as a generic transport option.
6729 65. Added queue_only_override (default true).
6731 66. Added the syslog_duplication option.
6733 67. If what should have been the first header line of a message consisted of
6734 a space followed by a colon, Exim was mis-interpreting it as a header line.
6735 It isn't of course - it is syntactically invalid and should therefore be
6736 treated as the start of the message body. The misbehaviour could have
6737 caused a number of strange effects, including loss of data in subsequent
6738 header lines, and spool format errors.
6740 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6741 client host had authenticated. This control can now be exercised by an ACL
6742 for more flexibility.
6744 69. By default, callouts do not happen when testing with -bh. There is now a
6745 variant, -bhc, which does actually run the callout code, including
6746 consulting and updating the callout cache.
6748 70. Added support for saslauthd authentication, courtesy of Alexander
6751 71. If statvfs() failed on the spool or log directories while checking their
6752 size for availability, Exim confusingly gave the error "space shortage".
6753 Furthermore, in debugging mode it crashed with a floating point exception.
6754 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6755 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6756 serious problem, Exim now writes to the main and panic logs when this
6757 happens, with details of the failure. It then refuses to accept the
6758 incoming message, giving the message "spool directory problem" or "log
6759 directory problem" with a 421 code for SMTP messages.
6761 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6762 0, 1, and 2 exist, because some OS complain for execs without them (see
6763 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6764 descriptors. However, the code omitted to check that the open succeeded,
6765 causing mysterious errors if for some reason the permissions on /dev/null
6766 got screwed. Now Exim writes a message to the main and panic logs, and
6767 bombs out if it can't open /dev/null.
6769 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6770 interact so that it is all more flexible. It is supposed to remain
6771 backwards compatible. Also added extra_local_interfaces.
6773 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6774 to bomb out with an assertion failure - to the client this appears as a
6775 connection drop. This problem occurs in the part of the code that was taken
6776 from the Samba project. Fortunately, the assertion is in a very simple
6777 function, so I have fixed this by reproducing the function inline in the
6778 one place where it is called, and arranging for authentication to fail
6779 instead of killing the process with assert().
6781 75. The SPA client code was not working when the server requested OEM rather
6782 than Unicode encoding.
6784 76. Added code to make require_files with a specific uid setting more usable in
6785 the case where statting the file as root fails - usually a non-root-mounted
6786 NFS file system. When this happens and the failure is EACCES, Exim now
6787 forks a subprocess and does the per-uid checking as the relevant uid.
6789 77. Added process_log_path.
6791 78. If log_file_path was not explicitly set, a setting of check_log_space or
6792 check_log_inodes was ignored.
6794 79. If a space check for the spool or log partitions fails, the incident is now
6795 logged. Of course, in the latter case the data may get lost...
6797 80. Added the %p formatting code to string_format() so that it can be used to
6798 print addresses in debug_print(). Adjusted all the address printing in the
6799 debugging in store.c to use %p rather than %d.
6801 81. There was a concern that a line of code in smtp_in.c could overflow a
6802 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6803 initially expressed, the concern was not well-founded, because trailing
6804 spaces are removed early. However, if the trailing spaces were followed by
6805 a NULL, they did not get removed, so the overflow was possible. Two fixes
6808 (a) I re-wrote the offending code in a cleaner fashion.
6809 (b) If an incoming SMTP command contains a NULL character, it is rejected
6812 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6813 runs initgroups(), so that if the Exim user is in any additional groups,
6814 they will be used during message reception.
6820 The change log for 4.20 and earlier releases has been archived.