1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
12 than -2003 did; needs libidn2 in addition to libidn.
14 JH/02 The path option on a pipe transport is now expanded before use.
16 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
17 Patch provided by "Björn", documentation fix added too.
19 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
20 missing a wire-to-host endian conversion.
22 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
23 close after a BDAT command line could be taken as a following command,
24 giving a synch failure. Fix by only checking for synch immediately
25 before acknowledging the chunk.
27 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
28 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
29 macro. Patches provided by Josh Soref.
31 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
32 Previously we did not; the RFC seems ambiguous and VRFY is not listed
33 by IANA as a service extension. However, John Klensin suggests that we
36 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
37 the dkim code may be unix-mode line endings rather than smtp wire-format
38 CRLF, so prepend a CR to any bare LF.
40 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
41 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
43 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
44 the first were themselves being wrongly included in the feed into dkim
45 processing; with most chunk sizes in use this resulted in an incorrect
46 body hash calculated value.
48 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
49 DKIM signature block, for verification. Although advised against by
50 standards it is specifically not ruled illegal.
52 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
54 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
55 missing a body hash (the bh= tag).
57 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
58 It seems that HAProxy sends the Proxy Protocol information in clear and
59 only then does a TLS startup, so do the same.
61 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
62 TCP connections (such as for Spamd) unless the daemon successfully set
63 Fast Open mode on its listening sockets. This fixes breakage seen on
64 too-old kernels or those not configured for Fast Open, at the cost of
65 requiring both directions being enabled for TFO, and TFO never being used
66 by non-daemon-related Exim processes.
68 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
69 endings, at least on the first header line. Try to canonify any that get
70 past that check, despite the cost.
72 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
73 now limited to an arbitrary five deep, while parsing addresses with the
74 strip_excess_angle_brackets option enabled.
76 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
77 instead leave the unprompted TLS handshake in socket buffer for the
78 TLS library to consume.
80 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
82 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
84 JH/16 Drop variables when they go out of scope. Memory management drops a whole
85 region in one operation, for speed, and this leaves assigned pointers
86 dangling. Add checks run only under the testsuite which checks all
87 variables at a store-reset and panics on a dangling pointer; add code
88 explicitly nulling out all the variables discovered. Fixes one known
89 bug: a transport crash, where a dangling pointer for $sending_ip_address
90 originally assigned in a verify callout, is re-used.
92 PP/06 Drop '.' from @INC in various Perl scripts.
94 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
96 PP/08 Reduce a number of compilation warnings under clang; building with
97 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
98 should be warning-free.
100 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
106 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
107 supports it and a size is available (ie. the sending peer gave us one).
109 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
110 years of being deprecated). The replacements are the ACLs
111 acl_smtp_mime and acl_not_smtp_mime.
113 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
114 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
115 or even an in-clear connection were permitted. Now, if the host lookup
116 was dnssec and dane was requested then the host is only used if the
117 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
118 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
119 if one fails this test.
120 This means that a poorly-configured remote DNS will make it incommunicado;
121 but it protects against a DNS-interception attack on it.
123 JH/04 Bug 1810: make continued-use of an open smtp transport connection
124 non-noisy when a race steals the message being considered.
126 JH/05 If main configuration option tls_certificate is unset, generate a
127 self-signed certificate for inbound TLS connections.
129 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
130 in rewrites and routers.
132 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
133 and logged a warning sing 4.83; now they are a configuration file error.
135 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
136 (lacking @domain). Apply the same qualification processing as RCPT.
138 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
140 JH/10 Support ${sha256:} applied to a string (as well as the previous
143 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
144 a cutthrough deliver is pending, as we always want to make a connection.
145 This also avoids re-routing the message when later placing the cutthrough
146 connection after a verify cache hit.
147 Do not update it with the verify result either.
149 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
150 when routing results in more than one destination address.
152 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
153 signing (which inhibits the cutthrough capability). Previously only
154 the presence of an option was tested; now an expansion evaluating as
155 empty is permissible (obviously it should depend only on data available
156 when the cutthrough connection is made).
158 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
159 the relevant preceding SMTP command did not note the pipelining mode.
161 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
162 Previously they were not counted.
164 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
165 as one having no matching records. Previously we deferred the message
166 that needed the lookup.
168 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
169 distinguished as "(=".
171 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
172 for missing MX records. Previously it only worked for missing A records.
174 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
176 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
177 after the data-go-ahead and data-ack. Patch from Jason Betts.
179 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
180 even for a "none" policy. Patch from Tony Meyer.
182 JH/22 Fix continued use of a connection for further deliveries. If a port was
183 specified by a router, it must also match for the delivery to be
186 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
187 When one of the recipients of a message was unsuitable for the connection
188 (has no matching addresses), we lost track of needing to mark it
189 deferred. As a result mail would be lost.
191 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
193 JH/25 Decoding ACL controls is now done using a binary search; the source code
194 takes up less space and should be simpler to maintain. Merge the ACL
195 condition decode tables also, with similar effect.
197 JH/26 Fix problem with one_time used on a redirect router which returned the
198 parent address unchanged. A retry would see the parent address marked as
199 delivered, so not attempt the (identical) child. As a result mail would
202 JH/27 Fix a possible security hole, wherein a process operating with the Exim
203 UID can gain a root shell. Credit to http://www.halfdog.net/ for
204 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
207 JH/28 Enable {spool,log} filesystem space and inode checks as default.
208 Main config options check_{log,spool}_{inodes,space} are now
209 100 inodes, 10MB unless set otherwise in the configuration.
211 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
212 Previously it only applied to the main-section connection policy
215 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
217 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
218 by me. Added RFC7919 DH primes as an alternative.
220 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
221 are not in the system include path.
223 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
224 GnuTLS, when a session startup failed (eg because the client disconnected)
225 Exim did stdio operations after fclose. This was exposed by a recent
226 change which nulled out the file handle after the fclose.
228 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
229 signed directly by the cert-signing cert, rather than an intermediate
230 OCSP-signing cert. This is the model used by LetsEncrypt.
232 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
234 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
235 an incoming connection.
237 HS/02 Bug 1802: Do not half-close the connection after sending a request
240 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
241 fallback to "prime256v1".
243 JH/34 SECURITY: Use proper copy of DATA command in error message.
244 Could leak key material. Remotely exploitable. CVE-2016-9963.
250 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
251 and 3.4.4 - once the server is enabled to respond to an OCSP request
252 it does even when not requested, resulting in a stapling non-aware
253 client dropping the TLS connection.
255 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
256 support variable-length bit vectors. No functional change.
258 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
259 The I= interface field on outgoing lines is now after the H= remote
260 host field, same as incoming lines. There is a separate
261 outgoing_interface log selector which allows you to disable the
264 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
265 If not running log_selector +smtp_connection the mainlog would be held
266 open indefinitely after a "too many connections" event, including to a
267 deleted file after a log rotate. Leave the per net connection logging
268 leaving it open for efficiency as that will be quickly detected by the
269 check on the next write.
271 HS/01 Bug 1671: Fix post transport crash.
272 Processing the wait-<transport> messages could crash the delivery
273 process if the message IDs didn't exist for some reason. When
274 using 'split_spool_directory=yes' the construction of the spool
275 file name failed already, exposing the same netto behaviour.
277 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
278 mime_regex ACL conditions.
280 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
281 to DSN fail messages (bounces): remote IP, remote greeting, remote response
282 to HELO, local diagnostic string.
284 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
285 log line to debug. Even when configured with a tls authenticator many
286 client connections are expected to not authenticate in this way, so
287 an authenticate fail is not an error.
289 HS/02 Add the Exim version string to the process info. This way exiwhat
290 gives some more detail about the running daemon.
292 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
293 matter for fast-change records such as DNSBLs.
295 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
296 retry record, even if constant. There may be multiple transports with
297 different interface settings and the retry behaviour needs to be kept
300 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
302 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
304 JH/10 Bug 840: fix log_defer_output option of pipe transport
306 JH/11 Bug 830: use same host for all RCPTS of a message, even under
307 hosts_randomize. This matters a lot when combined with mua_wrapper.
309 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
310 ${quote_pgsql:<string>} operator.
312 JH/13 Bug 1708: avoid misaligned access in cached lookup.
314 JH/14 Change header file name for freeradius-client. Relevant if compiling
315 with Radius support; from the Gentoo tree and checked under Fedora.
317 JH/15 Bug 1712: Introduce $prdr_requested flag variable
319 JH/16 Bug 1714: Permit an empty string as expansion result for transport
320 option transport_filter, meaning no filtering.
322 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
324 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
325 defaults to "*" (all hosts). The variable is now available when not built
326 with TLS, default unset, mainly to enable keeping the testsuite sane.
327 If a server certificate is not supplied (via tls_certificate) an error is
328 logged, and clients will find TLS connections fail on startup. Presumably
329 they will retry in-clear.
330 Packagers of Exim are strongly encouraged to create a server certificate
331 at installation time.
333 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
334 with the $config_file variable.
336 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
337 in transport context, after the attempt, and per-recipient. The latter type
338 is per host attempted. The event data is the error message, and the errno
339 information encodes the lookup type (A vs. MX) used for the (first) host,
340 and the trailing two digits of the smtp 4xx response.
342 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
343 to write to mainlog (or rejectlog, paniclog) in the window between file
344 creation and permissions/ownership being changed. Particularly affects
345 installations where exicyclog is run as root, rather than exim user;
346 result is that the running daemon panics and dies.
348 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
350 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
351 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
352 "pri" and "weight". Note that the previous implicit priority given by the
353 list order is no longer honoured.
355 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
358 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
359 by defining SUPPORT_SOCKS.
361 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
362 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
363 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
364 variables are renamed to proxy_{local,external}_{address,port}.
366 JH/27 Move Internationalisation support from Experimental to mainline, enabled
367 for a build by defining SUPPORT_I18N
369 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
370 of the query string, and make ${quote_redis:} do that quoting.
372 JH/29 Move Events support from Experimental to mainline, enabled by default
373 and removable for a build by defining DISABLE_EVENT.
375 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
377 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
378 cached by the daemon.
380 JH/32 Move Redis support from Experimental to mainline, enabled for a build
381 by defining LOOKUP_REDIS. The libhiredis library is required.
383 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
384 keys are given for lookup.
386 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
387 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
388 only supported when built with TLS support. The PolarSSL SHA routines
389 are still used when the TLS library is too old for convenient support.
391 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
392 openssl_options), for security. OpenSSL forces this from version 1.1.0
393 server-side so match that on older versions.
395 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
396 allocation for $value could be released as the expansion processing
397 concluded, but leaving the global pointer active for it.
399 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
400 and to use the domains and local_parts ACL conditions.
402 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
403 incorrectly not doubled on cutthrough transmission, hence seen as a
404 body-termination at the receiving system - resulting in truncated mails.
405 Commonly the sender saw a TCP-level error, and retransmitted the message
406 via the normal store-and-forward channel. This could result in duplicates
407 received - but deduplicating mailstores were liable to retain only the
408 initial truncated version.
410 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
412 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
414 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
415 we're in there, support oversigning also; bug 1309.
417 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
419 HS/04 Add support for keep_environment and add_environment options.
421 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
422 either intentional arithmetic overflow during PRNG, or testing config-
425 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
426 delivery resulted in actual delivery. Cancel cutthrough before DATA
429 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
430 rejects a recipient: pass the reject to the originator.
432 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
433 Many were false-positives and ignorable, but it's worth fixing the
436 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
437 for the new environment-manipulation done at startup. Move the routines
438 from being local to tls.c to being global via the os.c file.
440 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
441 an extract embedded as result-arg for a map, the first arg for extract
442 is unavailable so we cannot tell if this is a numbered or keyed
443 extraction. Accept either.
449 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
452 JH/02 The smtp transport option "multi_domain" is now expanded.
454 JH/03 The smtp transport now requests PRDR by default, if the server offers
457 JH/04 Certificate name checking on server certificates, when exim is a client,
458 is now done by default. The transport option tls_verify_cert_hostnames
459 can be used to disable this per-host. The build option
460 EXPERIMENTAL_CERTNAMES is withdrawn.
462 JH/05 The value of the tls_verify_certificates smtp transport and main options
463 default to the word "system" to access the system default CA bundle.
464 For GnuTLS, only version 3.0.20 or later.
466 JH/06 Verification of the server certificate for a TLS connection is now tried
467 (but not required) by default. The verification status is now logged by
468 default, for both outbound TLS and client-certificate supplying inbound
471 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
474 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
475 Status Notification (bounce) messages are now MIME format per RFC 3464.
476 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
477 under the control of the dsn_advertise_hosts option, and routers may
478 have a dsn_lasthop option.
480 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
481 default, modifiable by a malware= option. The list separator for
482 the options can now be changed in the usual way. Bug 68.
484 JH/10 The smtp_receive_timeout main option is now expanded before use.
486 JH/11 The incoming_interface log option now also enables logging of the
487 local interface on delivery outgoing connections.
489 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
490 if the interface and destination host and port all match.
492 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
495 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
496 Patch from Andrew Lewis.
498 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
499 now supports optional time-restrictions, weighting, and priority
500 modifiers per server. Patch originally by <rommer@active.by>.
502 JH/16 The spamd_address main option now supports a mixed list of local
503 and remote servers. Remote servers can be IPv6 addresses, and
504 specify a port-range.
506 JH/17 Bug 68: The spamd_address main option now supports an optional
507 timeout value per server.
509 JH/18 Bug 1581: Router and transport options headers_add/remove can
510 now have the list separator specified.
512 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
515 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
518 JH/21 Support for the A6 type of dns record is withdrawn.
520 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
521 rather than the verbs used.
523 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
524 from 255 to 1024 chars.
526 JH/24 Verification callouts now attempt to use TLS by default.
528 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
529 are generic router options now. The defaults didn't change.
531 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
532 Original patch from Alexander Shikoff, worked over by JH.
534 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
535 files if ZCAT_COMMAND is not executable.
537 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
539 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
541 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
542 Normally benign, it bites when the pair was led to by a CNAME;
543 modern usage is to not canonicalize the domain to a CNAME target
544 (and we were inconsistent anyway for A-only vs AAAA+A).
546 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
548 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
549 when evaluating $sender_host_dnssec.
551 JH/31 Check the HELO verification lookup for DNSSEC, adding new
552 $sender_helo_dnssec variable.
554 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
556 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
558 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
560 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
561 documented as working, but never had. Support all but $spam_report.
563 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
564 added for tls authenticator.
566 HS/03 Add perl_taintmode main config option
572 TL/01 When running the test suite, the README says that variables such as
573 no_msglog_check are global and can be placed anywhere in a specific
574 test's script, however it was observed that placement needed to be near
575 the beginning for it to behave that way. Changed the runtest perl
576 script to read through the entire script once to detect and set these
577 variables, reset to the beginning of the script, and then run through
578 the script parsing/test process like normal.
580 TL/02 The BSD's have an arc4random API. One of the functions to induce
581 adding randomness was arc4random_stir(), but it has been removed in
582 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
583 function when detected.
585 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
586 cause callback expansion.
588 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
589 syntax errors in an expansion can be treated as a string instead of
590 logging or causing an error, due to the internal use of bool_lax
591 instead of bool when processing it.
593 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
594 server certificates when making smtp deliveries.
596 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
598 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
600 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
602 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
603 Merged patch from Sebastian Wiedenroth.
605 JH/05 Fix results-pipe from transport process. Several recipients, combined
606 with certificate use, exposed issues where response data items split
607 over buffer boundaries were not parsed properly. This eventually
608 resulted in duplicates being sent. This issue only became common enough
609 to notice due to the introduction of connection certificate information,
610 the item size being so much larger. Found and fixed by Wolfgang Breyha.
612 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
613 size buffer was used, resulting in syntax errors when an expansion
616 JH/07 Add support for directories of certificates when compiled with a GnuTLS
617 version 3.3.6 or later.
619 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
620 is EXPERIMENTAL_EVENT, the main-configuration and transport options
621 both become "event_action", the variables become $event_name, $event_data
622 and $event_defer_errno. There is a new variable $verify_mode, usable in
623 routers, transports and related events. The tls:cert event is now also
624 raised for inbound connections, if the main configuration event_action
627 TL/06 In test suite, disable OCSP for old versions of openssl which contained
628 early OCSP support, but no stapling (appears to be less than 1.0.0).
630 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
631 server certificate names available under the smtp transport option
632 "tls_verify_cert_hostname" now do not permit multi-component wildcard
635 JH/10 Time-related extraction expansions from certificates now use the main
636 option "timezone" setting for output formatting, and are consistent
637 between OpenSSL and GnuTLS compilations. Bug 1541.
639 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
640 encoded parameter in the incoming message. Bug 1558.
642 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
643 include certificate info, eximon was claiming there were spoolfile
646 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
648 JH/14 Log delivery-related information more consistently, using the sequence
649 "H=<name> [<ip>]" wherever possible.
651 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
652 are problematic for Debian distribution, omit them from the release
655 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
657 JH/16 Fix string representation of time values on 64bit time_t architectures.
660 JH/17 Fix a null-indirection in certextract expansions when a nondefault
661 output list separator was used.
666 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
667 checkers that were complaining about end of non-void function with no
670 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
671 This was a regression introduced in 4.83 by another bugfix.
673 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
675 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
676 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
682 TF/01 Correctly close the server side of TLS when forking for delivery.
684 When a message was received over SMTP with TLS, Exim failed to clear up
685 the incoming connection properly after forking off the child process to
686 deliver the message. In some situations the subsequent outgoing
687 delivery connection happened to have the same fd number as the incoming
688 connection previously had. Exim would try to use TLS and fail, logging
689 a "Bad file descriptor" error.
691 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
692 utilities have not been installed.
694 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
695 temporary space as the ACL may create new global variables.
697 TL/01 LDAP support uses per connection or global context settings, depending
698 upon the detected version of the libraries at build time.
700 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
701 to extract and use the src ip:port in logging and expansions as if it
702 were a direct connection from the outside internet. PPv2 support was
703 updated based on HAProxy spec change in May 2014.
705 JH/02 Add ${listextract {number}{list}{success}{fail}}.
707 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
708 Properly escape header and check for NULL return.
710 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
713 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
715 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
716 characters in header names, implemented as a verify condition.
717 Contributed by Michael Fischer v. Mollard.
719 TL/05 Rename SPF condition results err_perm and err_temp to standardized
720 results permerror and temperror. Previous values are deprecated but
721 still accepted. In a future release, err_perm and err_temp will be
722 completely removed, which will be a backward incompatibility if the
723 ACL tests for either of these two old results. Patch contributed by
724 user bes-internal on the mailing list.
726 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
728 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
729 selectors, in both main and reject logs.
731 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
734 JH/07 Add malware type "sock" for talking to simple daemon.
736 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
738 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
739 routers/transports under cutthrough routing.
741 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
742 numbers. Touch up "bool" conditional to keep the same definition.
744 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
746 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
748 JH/12 Expand items in router/transport headers_add or headers_remove lists
749 individually rather than the list as a whole. Bug 1452.
751 Required for reasonable handling of multiple headers_ options when
752 they may be empty; requires that headers_remove items with embedded
753 colons must have them doubled (or the list-separator changed).
755 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
756 view the policy declared in the DMARC record. Currently, $dmarc_status
757 is a combined value of both the record presence and the result of the
760 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
762 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
763 dnslookup router and the smtp transport (applying to the forward
766 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
767 of ldap servers used for a specific lookup. Patch provided by Heiko
770 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
771 New variable $lookup_dnssec_authenticated for observability.
773 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
774 Patch submitted by Lars Timman.
776 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
778 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
779 Requires trusted mode and valid format message id, aborts otherwise.
780 Patch contributed by Heiko Schlichting.
782 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
783 certextract with support for various fields. Bug 1358.
785 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
786 is requested by default, modifiable by smtp transport option
789 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
790 operate on certificate variables to give certificate fingerprints
791 Also new ${sha256:cert_variable}.
793 JH/23 The PRDR feature is moved from being Experimental into the mainline.
795 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
796 Christian Aistleitner.
798 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
800 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
801 file. Patch from Wolfgang Breyha.
803 JH/25 Expand the coverage of the delivery $host and $host_address to
804 client authenticators run in verify callout. Bug 1476.
806 JH/26 Port service names are now accepted for tls_on_connect_ports, to
807 align with daemon_smtp_ports. Bug 72.
809 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
810 support and error reporting did not work properly.
812 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
813 and is readable. Patch from Andrew Colin Kissa.
815 TL/14 Enhance documentation of ${run expansion and how it parses the
816 commandline after expansion, particularly in the case when an
817 unquoted variable expansion results in an empty value.
819 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
821 PP/02 Fix internal collision of T_APL on systems which support RFC3123
822 by renaming away from it. Addresses GH issue 15, reported by
825 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
827 TL/15 SECURITY: prevent double expansion in math comparison functions
828 (can expand unsanitized data). Not remotely exploitable.
835 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
837 PP/02 Make -n do something, by making it not do something.
838 When combined with -bP, the name of an option is not output.
840 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
843 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
844 $sender_host_name and config options to manage this, and basic check
847 PP/05 DSCP support for outbound connections and control modifier for inbound.
849 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
850 (Only plugin which currently uses this is kerberos4, which nobody should
851 be using, but we should make it available and other future plugins might
852 conceivably use it, even though it would break NAT; stuff *should* be
853 using channel bindings instead).
855 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
856 name; added for Sendmail compatibility; requires admin caller.
857 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
858 just ignore it); requires trusted caller.
859 Also parse but ignore: -Ac -Am -X<logfile>
862 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
864 TL/02 Add +smtp_confirmation as a default logging option.
866 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
867 Patch by Magnus Holmgren from 2007-02-20.
869 TL/04 Bugzilla 1281 - Spec typo.
870 Bugzilla 1283 - Spec typo.
871 Bugzilla 1290 - Spec grammar fixes.
873 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
875 TL/06 Add Experimental DMARC support using libopendmarc libraries.
877 TL/07 Fix an out of order global option causing a segfault. Reported to dev
878 mailing list by by Dmitry Isaikin.
880 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
882 JH/02 Support "G" suffix to numbers in ${if comparisons.
884 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
886 NM/01 Bugzilla 1197 - Spec typo
887 Bugzilla 1196 - Spec examples corrections
889 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
891 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
892 gnutls_enable_pkcs11, but renamed to more accurately indicate its
895 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
896 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
898 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
899 "acl {{name}{arg}...}", and optional args on acl condition
902 JH/05 Permit multiple router/transport headers_add/remove lines.
904 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
906 JH/07 Avoid using a waiting database for a single-message-only transport.
907 Performance patch from Paul Fisher. Bugzilla 1262.
909 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
912 JH/09 Add $headers_added variable, with content from use of ACL modifier
913 add_header (but not yet added to the message). Bugzilla 199.
915 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
916 Pulled from Bugzilla 817 by Wolfgang Breyha.
918 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
920 (nb: this is the same fix as in Exim 4.80.1)
922 JH/11 Add A= logging on delivery lines, and a client_set_id option on
925 JH/12 Add optional authenticated_sender logging to A= and a log_selector
928 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
930 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
931 advertise SMTP AUTH mechanism to us, instead of a generic
932 protocol violation error. Also, make Exim more robust to bad
933 data from the Dovecot auth socket.
935 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
937 When a queue runner is handling a message, Exim first routes the
938 recipient addresses, during which it prunes them based on the retry
939 hints database. After that it attempts to deliver the message to
940 any remaining recipients. It then updates the hints database using
943 So if a recipient address works intermittently, it can get repeatedly
944 deferred at routing time. The retry hints record remains fresh so the
945 address never reaches the final cutoff time.
947 This is a fairly common occurrence when a user is bumping up against
948 their storage quota. Exim had some logic in its local delivery code
949 to deal with this. However it did not apply to per-recipient defers
950 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
952 This change adds a proper retry rule check during routing so that the
953 final cutoff time is checked against the message's age. We only do
954 this check if there is an address retry record and there is not a
955 domain retry record; this implies that previous attempts to handle
956 the address had the retry_use_local_parts option turned on. We use
957 this as an approximation for the destination being like a local
958 delivery, as in LMTP.
960 I suspect this new check makes the old local delivery cutoff check
961 redundant, but I have not verified this so I left the code in place.
963 TF/02 Correct gecos expansion when From: is a prefix of the username.
965 Test 0254 submits a message to Exim with the header
969 When I ran the test suite under the user fanf2, Exim expanded
970 the header to contain my full name, whereas it should have added
971 a Resent-Sender: header. It erroneously treats any prefix of the
972 username as equal to the username.
974 This change corrects that bug.
976 GF/01 DCC debug and logging tidyup
977 Error conditions log to paniclog rather than rejectlog.
978 Debug lines prefixed by "DCC: " to remove any ambiguity.
980 TF/03 Avoid unnecessary rebuilds of lookup-related code.
982 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
983 Bug spotted by Jeremy Harris; was flawed since initial commit.
984 Would have resulted in OCSP responses post-SNI triggering an Exim
985 NULL dereference and crash.
987 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
989 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
990 Bug detection, analysis and fix by Samuel Thibault.
991 Bugzilla 1331, Debian bug #698092.
993 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
995 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
996 Server implementation by Todd Lyons, client by JH.
997 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
998 config variable "prdr_enable" controls whether the server
999 advertises the facility. If the client requests PRDR a new
1000 acl_data_smtp_prdr ACL is called once for each recipient, after
1001 the body content is received and before the acl_smtp_data ACL.
1002 The client is controlled by both of: a hosts_try_prdr option
1003 on the smtp transport, and the server advertisement.
1004 Default client logging of deliveries and rejections involving
1005 PRDR are flagged with the string "PRDR".
1007 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1008 fclose(). Diagnosis by Todd Lyons.
1010 PP/17 Update configure.default to handle IPv6 localhost better.
1011 Patch by Alain Williams (plus minor tweaks).
1014 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1015 This is now consistent with GnuTLS, and is now documented: the
1016 previous undocumented portable approach to treating the option as
1017 unset was to force an expansion failure. That still works, and
1018 an empty string is now equivalent.
1020 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1021 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1022 not performing validation itself.
1024 PP/20 Added force_command boolean option to pipe transport.
1025 Patch from Nick Koston, of cPanel Inc.
1027 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1030 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1032 PP/21 Fix eximon continuous updating with timestamped log-files.
1033 Broken in a format-string cleanup in 4.80, missed when I repaired the
1034 other false fix of the same issue.
1035 Report and fix from Heiko Schlichting.
1038 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1039 Report from Prashanth Katuri.
1041 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1042 It's SecureTransport, so affects any MacOS clients which use the
1043 system-integrated TLS libraries, including email clients.
1045 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1046 using a MIME ACL for non-SMTP local injection.
1047 Report and assistance in diagnosis by Warren Baker.
1049 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1051 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1053 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1054 last id that failed so it may be referenced in subsequent ACL's.
1056 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1059 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1060 ldap library initialization, allowing self-signed CA's to be
1061 used. Also properly sets require_cert option later in code by
1062 using NULL (global ldap config) instead of ldap handle (per
1063 session). Bug diagnosis and testing by alxgomz.
1065 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1066 the src/util/ subdirectory.
1068 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1069 renamed to Transport Post Delivery Action by Jeremy Harris, as
1072 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1073 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1074 redis_servers = needs to be configured which will be used by the redis
1075 lookup. Patch from Warren Baker, of The Packet Hub.
1077 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1079 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1080 hostname or reverse DNS when processing a host list. Used suggestions
1081 from multiple comments on this bug.
1083 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1085 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1086 Missed a few lines, added it to make the runtest require no keyboard
1089 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1090 contains upper case chars. Make router use caseful_local_part.
1092 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1093 support when GnuTLS has been built with p11-kit.
1099 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1101 This, or similar/improved, will also be change PP/11 of 4.82.
1107 PP/01 Handle short writes when writing local log-files.
1108 In practice, only affects FreeBSD (8 onwards).
1109 Bugzilla 1053, with thanks to Dmitry Isaikin.
1111 NM/01 Bugzilla 949 - Documentation tweak
1113 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1116 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1118 PP/02 Implemented gsasl authenticator.
1120 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1122 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1123 `pkg-config foo` for cflags/libs.
1125 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1126 with rest of GSASL and with heimdal_gssapi.
1128 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1129 `pkg-config foo` for cflags/libs for the TLS implementation.
1131 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1132 properties get this fed in as external SSF. A number of robustness
1133 and debugging improvements to the cyrus_sasl authenticator.
1135 PP/08 cyrus_sasl server now expands the server_realm option.
1137 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1138 Patch by Jeremy Harris.
1140 PP/10 Added dbmjz lookup type.
1142 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1144 PP/12 MAIL args handles TAB as well as SP, for better interop with
1145 non-compliant senders.
1146 Analysis and variant patch by Todd Lyons.
1148 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1149 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1150 Patch from Dirk Mueller <dmueller@suse.com>
1152 PP/13 tls_peerdn now print-escaped for spool files.
1153 Observed some $tls_peerdn in wild which contained \n, which resulted
1154 in spool file corruption.
1156 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1157 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1158 or write after TLS renegotiation, which otherwise led to messages
1161 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1162 as a tracking header (ie: a signed header comes before the signature).
1163 Patch from Wolfgang Breyha.
1165 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1166 comma-sep list; embedded commas doubled.
1168 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1170 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1172 Report and patch from Dmitry Banschikov.
1174 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1175 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1176 failures. We appear to now support TLS1.1+ with Exim.
1178 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1179 lets Exim select keys and certificates based upon TLS SNI from client.
1180 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1181 before an outbound SMTP session. New log_selector, +tls_sni.
1183 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1184 NULL dereference. Report and patch from Alun Jones.
1186 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1187 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1188 Not seeing resolver debug output on NetBSD, but suspect this is a
1189 resolver implementation change.
1191 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1192 Left warnings. Added "eximon gdb" invocation mode.
1194 PP/21 Defaulting "accept_8bitmime" to true, not false.
1196 PP/22 Added -bw for inetd wait mode support.
1198 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1199 locate the relevant includes and libraries. Made this the default.
1201 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1202 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1204 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1205 This may cause build issues on older platforms.
1207 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1208 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1209 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1210 Added SNI support via GnuTLS too.
1211 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1213 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1215 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1216 Applied second patch from Janne, implementing suggestion to default
1217 multiple-strings-in-record handling to match SPF spec.
1219 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1221 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1222 read-only, out of scope).
1223 Patch from Wolfgang Breyha, report from Stuart Northfield.
1225 PP/29 Fix three issues highlighted by clang analyser static analysis.
1226 Only crash-plausible issue would require the Cambridge-specific
1227 iplookup router and a misconfiguration.
1228 Report from Marcin Mirosław.
1230 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1232 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1233 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1234 As part of this, removing so much warning spew let me fix some minor
1235 real issues in debug logging.
1237 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1238 assignment on my part. Fixed.
1240 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1241 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1242 Janne Snabb (who went above and beyond: thank you).
1244 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1245 string otherwise requires a connection and a bunch more work and it's
1246 relatively easy to get wrong. Should also expose TLS library linkage
1249 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1250 64-bit ${eval} (JH/03).
1252 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1253 GNU libc to support some of the 64-bit stuff, should not lead to
1254 conflicts. Defined before os.h is pulled in, so if a given platform
1255 needs to override this, it can.
1257 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1258 protection layer was required, which is not implemented.
1259 Bugzilla 1254, patch from Wolfgang Breyha.
1261 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1262 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1263 tls_dhparam take prime identifiers. Also unbreak combination of
1264 OpenSSL+DH_params+TLSSNI.
1266 PP/39 Disable SSLv2 by default in OpenSSL support.
1272 PP/01 Solaris build fix for Oracle's LDAP libraries.
1273 Bugzilla 1109, patch from Stephen Usher.
1275 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1277 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1280 TF/02 Fix a couple more cases where we did not log the error message
1281 when unlink() failed. See also change 4.74-TF/03.
1283 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1284 lock up or crash if it happened to be inside a call to libc when it
1285 got a SIGUSR1 from exiwhat.
1287 The SIGUSR1 handler appends the current process status to the process
1288 log which is later printed by exiwhat. It used to use the general
1289 purpose logging code to do this, but several functions it calls are
1290 not safe for signals.
1292 The new output code in the SIGUSR1 handler is specific to the process
1293 log, and simple enough that it's easy to inspect for signal safety.
1294 Removing some special cases also simplifies the general logging code.
1295 Removing the spurious timestamps from the process log simplifies
1298 TF/04 Improved ratelimit ACL condition.
1300 The /noupdate option has been deprecated in favour of /readonly which
1301 has clearer semantics. The /leaky, /strict, and /readonly update modes
1302 are mutually exclusive. The update mode is no longer included in the
1303 database key; it just determines when the database is updated. (This
1304 means that when you upgrade Exim will forget old rate measurements.)
1306 Exim now checks that the per_* options are used with an update mode that
1307 makes sense for the current ACL. For example, when Exim is processing a
1308 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1309 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1310 must specify per_mail/readonly. If you omit the update mode it defaults to
1311 /leaky where that makes sense (as before) or /readonly where required.
1313 The /noupdate option is now undocumented but still supported for
1314 backwards compatibility. It is equivalent to /readonly except that in
1315 ACLs where /readonly is required you may specify /leaky/noupdate or
1316 /strict/noupdate which are treated the same as /readonly.
1318 A useful new feature is the /count= option. This is a generalization
1319 of the per_byte option, so that you can measure the throughput of other
1320 aggregate values. For example, the per_byte option is now equivalent
1321 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1323 The per_rcpt option has been generalized using the /count= mechanism
1324 (though it's more complicated than the per_byte equivalence). When it is
1325 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1326 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1327 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1328 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1329 Note that using per_rcpt with a non-readonly update mode in more than
1330 one ACL will cause the recipients to be double-counted. (The per_mail
1331 and per_byte options don't have this problem.)
1333 The handling of very low rates has changed slightly. If the computed rate
1334 is less than the event's count (usually one) then this event is the first
1335 after a long gap. In this case the rate is set to the same as this event's
1336 count, so that the first message of a spam run is counted properly.
1338 The major new feature is a mechanism for counting the rate of unique
1339 events. The new per_addr option counts the number of different
1340 recipients that someone has sent messages to in the last time period. It
1341 behaves like per_rcpt if all the recipient addresses are different, but
1342 duplicate recipient addresses do not increase the measured rate. Like
1343 the /count= option this is a general mechanism, so the per_addr option
1344 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1345 example, measure the rate that a client uses different sender addresses
1346 with the options per_mail/unique=$sender_address. There are further
1347 details in the main documentation.
1349 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1351 TF/06 Removed a few PCRE remnants.
1353 TF/07 Automatically extract Exim's version number from tags in the git
1354 repository when doing development or release builds.
1356 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1357 Bugzilla 879. Patch from Paul Fisher.
1359 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1360 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1363 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1365 PP/05 Variable $av_failed, true if the AV scanner deferred.
1366 Bugzilla 1078. Patch from John Horne.
1368 PP/06 Stop make process more reliably on build failure.
1369 Bugzilla 1087. Patch from Heiko Schlittermann.
1371 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1372 Bugzilla 1089. Patch from Heiko Schlittermann.
1374 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1375 Bugzilla 1131. Patch from Holger Weiß.
1377 PP/09 Handle IPv6 addresses with SPF.
1378 Bugzilla 860. Patch from Wolfgang Breyha.
1380 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1382 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1385 PP/11 match_* no longer expand right-hand-side by default.
1386 New compile-time build option, EXPAND_LISTMATCH_RHS.
1387 New expansion conditions, "inlist", "inlisti".
1389 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1391 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1393 PP/14 fix log_write() format string regression from TF/03.
1394 Bugzilla 1152. Patch from Dmitry Isaikin.
1400 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1402 PP/02 Harmonised TLS library version reporting; only show if debugging.
1403 Layout now matches that introduced for other libraries in 4.74 PP/03.
1405 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1407 PP/04 New "dns_use_edns0" global option.
1409 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1412 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1413 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1415 TK/01 Updated PolarSSL code to 0.14.2.
1416 Bugzilla 1097. Patch from Andreas Metzler.
1418 PP/07 Catch divide-by-zero in ${eval:...}.
1419 Fixes bugzilla 1102.
1421 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1424 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1425 format-string attack -- SECURITY: remote arbitrary code execution.
1427 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1428 time unintentionally subject to list matching rules, letting the header
1429 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1430 arbitrary string expansion). This allowed for information disclosure.
1432 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1433 INT_MIN/-1 -- value coerced to INT_MAX.
1439 NM/01 Workaround for PCRE version dependency in version reporting
1442 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1443 This fixes portability to compilers other than gcc, notably
1444 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1446 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1447 makefiles for portability to HP-UX and POSIX correctness.
1449 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1450 Also via indented variable definition in the Makefile.
1451 (Debugging by Oliver Heesakkers).
1453 PP/02 Restore caching of spamd results with expanded spamd_address.
1454 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1456 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1457 Improves build reliability. Fix from: Frank Elsner
1459 NM/02 Fix wide character breakage in the rfc2047 coding
1460 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1462 NM/03 Allow underscore in dnslist lookups
1463 Fixes bug 1026. Patch from Graeme Fowler
1465 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1466 Code patches from Adam Ciarcinski of NetBSD.
1468 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1471 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1472 is logged, to avoid truncation. Patch from John Horne.
1474 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1475 Patch from Jakob Hirsch.
1477 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1478 SQL string expansion failure details.
1479 Patch from Andrey Oktyabrski.
1481 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1482 Patch from Simon Arlott.
1484 PP/09 New lookups functionality failed to compile on old gcc which rejects
1485 extern declarations in function scope.
1486 Patch from Oliver Fleischmann
1488 PP/10 Use sig_atomic_t for flags set from signal handlers.
1489 Check getgroups() return and improve debugging.
1490 Fixed developed for diagnosis in bug 927 (which turned out to be
1493 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1494 Patch from Mark Zealey.
1496 PP/12 Bugzilla 1056: Improved spamd server selection.
1497 Patch from Mark Zealey.
1499 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1500 Based on patch from Heiko Schlittermann.
1502 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1503 Patch from Uwe Doering, sign-off by Michael Haardt.
1505 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1506 variable declaration deep within a block. Bug and patch from
1509 PP/15 lookups-Makefile IRIX compatibility coercion.
1511 PP/16 Make DISABLE_DKIM build knob functional.
1513 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1514 Patch by Simon Arlott
1516 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1517 variable argument macros. Our copy now differs from upstream.
1523 TF/01 Failure to get a lock on a hints database can have serious
1524 consequences so log it to the panic log.
1526 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1527 controlled using the smtp_confirmation log selector.
1529 TF/03 Include the error message when we fail to unlink a spool file.
1531 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1532 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1533 for maintaining out-of-tree patches for some time.
1535 PP/01 Bugzilla 139: Documentation and portability issues.
1536 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1537 Handle per-OS dynamic-module compilation flags.
1539 PP/02 Let /dev/null have normal permissions.
1540 The 4.73 fixes were a little too stringent and complained about the
1541 permissions on /dev/null. Exempt it from some checks.
1542 Reported by Andreas M. Kirchwitz.
1544 PP/03 Report version information for many libraries, including
1545 Exim version information for dynamically loaded libraries. Created
1546 version.h, now support a version extension string for distributors
1547 who patch heavily. Dynamic module ABI change.
1549 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1550 privilege escalation vulnerability whereby the Exim run-time user
1551 can cause root to append content of the attacker's choosing to
1554 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1557 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1558 If dropping privileges for untrusted macros, we disabled normal logging
1559 on the basis that it would fail; for the Exim run-time user, this is not
1560 the case, and it resulted in successful deliveries going unlogged.
1561 Fixed. Reported by Andreas Metzler.
1567 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1568 only prepend for the Resent-* case. Fixes regression introduced in
1569 Exim 4.70 by NM/22 for Bugzilla 607.
1571 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1572 increasing numbers of administrators be bitten by this.
1574 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1576 PP/03 Bugzilla 994: added openssl_options main configuration option.
1578 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1580 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1582 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1584 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1585 without a peer certificate, leading to a segfault because of an
1586 assumption that peers always have certificates. Be a little more
1587 paranoid. Problem reported by Martin Tscholak.
1589 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1590 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1591 NB: ClamAV planning to remove STREAM in "middle of 2010".
1592 CL also introduces -bmalware, various -d+acl logging additions and
1593 more caution in buffer sizes.
1595 PP/09 Implemented reverse_ip expansion operator.
1597 PP/10 Bugzilla 937: provide a "debug" ACL control.
1599 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1601 PP/12 Bugzilla 973: Implement --version.
1603 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1605 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1607 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1609 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1610 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1611 ignore trailing whitespace.
1613 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1615 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1618 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1619 Notification from Dr Andrew Aitchison.
1621 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1622 ExtendedDetectionInfo response format.
1623 Notification from John Horne.
1625 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1628 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1629 XSL and documented dependency on system catalogs, with examples of how
1632 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1635 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1636 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1637 configuration file which is writeable by the Exim user or group.
1639 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1640 of configuration files to cover files specified with the -C option if
1641 they are going to be used with root privileges, not just the default
1644 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1645 option (effectively making it always true).
1647 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1648 files to be used while preserving root privileges.
1650 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1651 that rogue child processes cannot use them.
1653 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1654 run-time user, instead of root.
1656 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1657 Exim run-time user without dropping privileges.
1659 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1660 result string, instead of calling string_vformat() twice with the same
1663 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1664 for other users. Others should always drop root privileges if they use
1665 -C on the command line, even for a whitelisted configure file.
1667 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1669 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1675 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1676 $data_path, and $header_path variables; fixed documentation bugs and
1679 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1680 exipick to access non-standard spools, including the "frozen" queue
1683 NM/01 Bugzilla 965: Support mysql stored procedures.
1684 Patch from Alain Williams
1686 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1688 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1689 Patch from Andreas Metzler
1691 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1692 Patch from Kirill Miazine
1694 NM/05 Bugzilla 671: Added umask to procmail example.
1696 JJ/03 installed exipick 20100323.0, fixing doc bug
1698 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1699 directory. Notification and patch from Dan Rosenberg.
1701 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1703 TK/02 Improve log output when DKIM signing operation fails.
1705 MH/01 Treat the transport option dkim_domain as a colon separated
1706 list, not as a single string, and sign the message with each element,
1707 omitting multiple occurences of the same signer.
1709 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1710 Bugzilla 985, 986. Patch by Simon Arlott
1712 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1713 Patch by Simon Arlott
1715 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1716 MBX locking. Notification from Dan Rosenberg.
1722 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1724 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1726 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1728 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1730 NM/04 Bugzilla 915: Flag broken perl installation during build.
1736 TK/01 Added patch by Johannes Berg that expands the main option
1737 "spamd_address" if it starts with a dollar sign.
1739 TK/02 Write list of recipients to X-Envelope-Sender header when building
1740 the mbox-format spool file for content scanning (suggested by Jakob
1743 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1744 (http://www.dcc-servers.net/) support via dccifd. Activated by
1745 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1747 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1748 by Mark Daniel Reidel <mr@df.eu>.
1750 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1751 When building exim an external PCRE library is now needed -
1752 PCRE is a system library on the majority of modern systems.
1753 See entry on PCRE_LIBS in EDITME file.
1755 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1756 conversation. Added nologin parameter to request.
1757 Patch contributed by Kirill Miazine.
1759 TF/01 Do not log submission mode rewrites if they do not change the address.
1761 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1763 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1764 log files in place. Contributed by Roberto Lima.
1766 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1768 TF/03 Bugzilla 615: When checking the local_parts router precondition
1769 after a local_part_suffix or local_part_prefix option, Exim now
1770 does not use the address's named list lookup cache, since this
1771 contains cached lookups for the whole local part.
1773 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1774 Robert Millan. Documentation is in experimental-spec.txt.
1776 TF/04 Bugzilla 668: Fix parallel build (make -j).
1778 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1780 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1781 Patch provided by Jan Srzednicki.
1783 TF/05 Leading white space used to be stripped from $spam_report which
1784 wrecked the formatting. Now it is preserved.
1786 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1787 that they are available at delivery time.
1789 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1791 TF/08 TLS error reporting now respects the incoming_interface and
1792 incoming_port log selectors.
1794 TF/09 Produce a more useful error message if an SMTP transport's hosts
1795 setting expands to an empty string.
1797 NM/06 Bugzilla 744: EXPN did not work under TLS.
1798 Patch provided by Phil Pennock.
1800 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1801 Patch provided by Richard Godbee.
1803 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1804 acl_smtp_notquit, added index entry.
1806 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1807 Patch provided by Eugene Bujak.
1809 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1810 accept(). Patch provided by Maxim Dounin.
1812 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1813 Patch provided by Phil Pennock.
1815 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1817 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1818 Patch provided by Brad "anomie" Jorsch.
1820 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1821 Patch provided by Dean Brooks.
1823 TK/05 Add native DKIM support (does not depend on external libraries).
1825 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1826 Patch provided by Graeme Fowler.
1828 NM/16 Bugzilla 851: Documentation example syntax fix.
1830 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1832 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1835 NM/19 Bugzilla 745: TLS version reporting.
1836 Patch provided by Phil Pennock.
1838 NM/20 Bugzilla 167: bool: condition support.
1839 Patch provided by Phil Pennock.
1841 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1842 clients. Patch provided by Phil Pennock.
1844 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1845 Patch provided by Brad "anomie" Jorsch.
1847 NM/23 Bugzilla 687: Fix misparses in eximstats.
1848 Patch provided by Heiko Schlittermann.
1850 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1851 Patch provided by Heiko Schlittermann.
1853 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1854 plus update to original patch.
1856 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1858 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1859 Patch provided by David Brownlee.
1861 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1863 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1865 NM/30 Bugzilla 888: TLS documentation bugfixes.
1867 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1869 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1870 Unlike the original bugzilla I have changed all shell scripts in src tree.
1872 NM/33 Bugzilla 898: Transport filter timeout fix.
1873 Patch by Todd Rinaldo.
1875 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1876 Patch by Serge Demonchaux.
1878 NM/35 Bugzilla 39: Base64 decode bug fixes.
1879 Patch by Jakob Hirsch.
1881 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1883 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1885 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1887 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1893 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1894 ALT-N's libdkim that I have put here:
1895 http://duncanthrax.net/exim-experimental/
1897 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1898 were called 'true' and it seems that C99 defines that as a reserved
1899 keyword to be used with 'bool' variable types. That means you could
1900 not include C99-style headers which use bools without triggering
1901 build errors in sieve.c.
1903 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1904 as mailq or other aliases. Changed the --help handling significantly
1905 to do whats expected. exim_usage() emits usage/help information.
1907 SC/01 Added the -bylocaldomain option to eximstats.
1909 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1911 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1913 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1919 PH/01 Another patch from the Sieve maintainer.
1921 PH/02 When an IPv6 address is converted to a string for single-key lookup
1922 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1923 dots are used instead of colons so that keys in lsearch files need not
1924 contain colons. This was done some time before quoting was made available
1925 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1926 (notated using the quote facility) so as to distinguish them from IPv4
1927 keys. This meant that lookups for IP addresses in host lists did not work
1928 for iplsearch lookups.
1930 This has been fixed by arranging for IPv6 addresses to be expressed with
1931 colons if the lookup type is iplsearch. This is not incompatible, because
1932 previously such lookups could never work.
1934 The situation is now rather anomalous, since one *can* have colons in
1935 ordinary lsearch keys. However, making the change in all cases is
1936 incompatible and would probably break a number of configurations.
1938 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1941 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1942 conversion specification without a maximum field width, thereby enabling
1943 a rogue spamd server to cause a buffer overflow. While nobody in their
1944 right mind would setup Exim to query an untrusted spamd server, an
1945 attacker that gains access to a server running spamd could potentially
1946 exploit this vulnerability to run arbitrary code as the Exim user.
1948 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1949 $primary_hostname instead of what libspf2 thinks the hosts name is.
1951 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1952 a directory entry by the name of the lookup key. Previously, if a
1953 symlink pointed to a non-existing file or a file in a directory that
1954 Exim lacked permissions to read, a lookup for a key matching that
1955 symlink would fail. Now it is enough that a matching directory entry
1956 exists, symlink or not. (Bugzilla 503.)
1958 PH/03 The body_linecount and body_zerocount variables are now exported in the
1961 PH/04 Added the $dnslist_matched variable.
1963 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1964 This means they are set thereafter only if the connection becomes
1967 PH/06 Added the client_condition to authenticators so that some can be skipped
1968 by clients under certain conditions.
1970 PH/07 The error message for a badly-placed control=no_multiline_responses left
1971 "_responses" off the end of the name.
1973 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1975 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1976 (without spaces) instead of just copying the configuration text.
1978 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1980 PH/11 Added $max_received_linelength.
1982 PH/12 Added +ignore_defer and +include_defer to host lists.
1984 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1985 way in which PCRE > 7.0 is built.
1987 PH/14 Implemented queue_only_load_latch.
1989 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1990 MAIL command. The effect was to mangle the value on 64-bit systems.
1992 PH/16 Another patch from the Sieve maintainer.
1994 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1996 PH/18 If a system quota error occurred while trying to create the file for
1997 a maildir delivery, the message "Mailbox is full" was not appended to the
1998 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1999 only to a quota excession during the actual writing of the file.
2001 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2002 characters?) which causes problems in log lines. The DN values are now
2003 passed through string_printing() before being added to log lines.
2005 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2006 and InterBase are left for another time.)
2008 PH/21 Added message_body_newlines option.
2010 PH/22 Guard against possible overflow in moan_check_errorcopy().
2012 PH/23 POSIX allows open() to be a macro; guard against that.
2014 PH/24 If the recipient of an error message contained an @ in the local part
2015 (suitably quoted, of course), incorrect values were put in $domain and
2016 $local_part during the evaluation of errors_copy.
2022 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2023 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2026 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2027 issue a MAIL command.
2029 PH/02 In an ACL statement such as
2031 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2033 if a client was not listed at all, or was listed with a value other than
2034 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2035 the condition was not true (as it should be), so access was not denied.
2036 The bug was that the ! inversion was incorrectly passed on to the second
2037 item. This has been fixed.
2039 PH/03 Added additional dnslists conditions == and =& which are different from
2040 = and & when the dns lookup returns more than one IP address.
2042 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2043 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2045 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2046 FSYNC, which compiles an option called disable_fsync that allows for
2047 bypassing fsync(). The documentation is heavily laced with warnings.
2049 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2051 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2052 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2053 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2054 including adding "make clean"; (3) Added -fPIC when compiling the test
2055 dynamically loaded module, to get rid of a warning.
2057 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2058 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2059 = 0s. The bug is otherwise harmless.
2061 PH/07 There was a bug in the dovecot authenticator such that the value of
2062 $auth1 could be overwritten, and so not correctly preserved, after a
2063 successful authentication. This usually meant that the value preserved by
2064 the server_setid option was incorrect.
2066 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2068 PH/09 Installed PCRE release 7.0.
2070 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2071 run for batched SMTP input. It is now run at the start of every message
2072 in the batch. While fixing this I discovered that the process information
2073 (output by running exiwhat) was not always getting set for -bs and -bS
2074 input. This is fixed, and it now also says "batched" for BSMTP.
2076 PH/11 Added control=no_pipelining.
2078 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2079 patch, slightly modified), and move the expansion of helo_data till after
2080 the connection is made in the smtp transport (so it can use these
2083 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2085 PH/14 Added log_selector = +pid.
2087 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2089 PH/16 Add ${if forany and ${if forall.
2091 PH/17 Added dsn_from option to vary the From: line in DSNs.
2093 PH/18 Flush SMTP output before performing a callout, unless control =
2094 no_callout_flush is set.
2096 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2097 was true (the default) a successful delivery failed to delete the retry
2098 item, thus causing premature timeout of the address. The bug is now
2101 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2103 PH/21 Long custom messages for fakedefer and fakereject are now split up
2104 into multiline responses in the same way that messages for "deny" and
2105 other ACL rejections are.
2107 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2108 with slight modification.
2110 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2111 draft, changing the syntax and factoring some duplicate code".
2113 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2114 for deliveries of the second and subsequent messages over the same SMTP
2117 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2118 ${reduce, with only minor "tidies".
2120 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2122 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2123 expansion side effects.
2125 PH/27 When a message times out after an over-quota error from an Exim-imposed
2126 quota, the bounce message says "mailbox is full". This message was not
2127 being given when it was a system quota that was exceeded. It now should
2130 MH/03 Made $recipients available in local_scan(). local_scan() already has
2131 better access to the recipient list through recipients_list[], but
2132 $recipients can be useful in postmaster-provided expansion strings.
2134 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2135 in the case of a MAIL command with additional options following the
2136 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2137 were accidentally chopped off.
2139 PH/29 SMTP synchronization checks are implemented when a command is read -
2140 there is a check that no more input is waiting when there shouldn't be
2141 any. However, for some commands, a delay in an ACL can mean that it is
2142 some time before the response is written. In this time, more input might
2143 arrive, invalidly. So now there are extra checks after an ACL has run for
2144 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2145 pipelining has not been advertised.
2147 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2149 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2150 returned in the SMTP response when smtp_return_error_details was false.
2151 This has been fixed.
2153 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2154 instead of the C I/O that was originally supplied, because problems were
2155 reported on Solaris.
2157 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2158 Exim which did not show up earlier: it was assuming that a call to
2159 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2160 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2161 was a macro that became an assignment, so it seemed to work. This has
2162 changed to a proper function call with a void return, hence the compile
2163 error. Exim's code has been fixed.
2165 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2168 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2170 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2172 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2173 because it turns out to be too big or there is a timeout) there is no
2174 "Completed" line in the log. When some messages of this type were
2175 selected by exigrep, they were listed as "not completed". Others were
2176 picked up by some special patterns. I have improved the selection
2177 criteria to be more general.
2179 PH/38 The host_find_failed option in the manualroute router can now be set
2180 to "ignore", to completely ignore a host whose IP address cannot be
2181 found. If all hosts are ignored, the behaviour is controlled by the new
2182 host_all_ignored option.
2184 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2185 homing or because of multiple MX records with /mx) generated more than
2186 one IP address, and the following item turned out to be the local host,
2187 all the secondary addresses of the first item were incorrectly removed
2188 from the list, along with the local host and any following hosts (which
2189 is what is supposed to happen).
2191 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2192 whoever called Exim into the -H file. In the case of the daemon it was
2193 behaving confusingly. When first started, it used values for whoever
2194 started the daemon, but after a SIGHUP it used the Exim user (because it
2195 calls itself on a restart). I have changed the code so that it now always
2198 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2199 message are rejected with the same error (e.g. no authentication or bad
2200 sender address), and a DATA command is nevertheless sent (as can happen
2201 with PIPELINING or a stupid MUA), the error message that was given to the
2202 RCPT commands is included in the rejection of the DATA command. This is
2203 intended to be helpful for MUAs that show only the final error to their
2206 PH/42 Another patch from the Sieve maintainer.
2208 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2209 Eximstats - Fixed some broken HTML links and added missing column headers
2211 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2212 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2214 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2216 PH/43 Yet another patch from the Sieve maintainer.
2218 PH/44 I found a way to check for a TCP/IP connection going away before sending
2219 the response to the final '.' that terminates a message, but only in the
2220 case where the client has not sent further data following the '.'
2221 (unfortunately, this is allowed). However, in many cases there won't be
2222 any further data because there won't be any more messages to send. A call
2223 to select() can be used: if it shows that the input is "ready", there is
2224 either input waiting, or the socket has been closed. An attempt to read
2225 the next input character can distinguish the two cases. Previously, Exim
2226 would have sent an OK response which the client would never have see.
2227 This could lead to message repetition. This fix should cure that, at
2228 least in a lot of common cases.
2230 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2231 advertised in response to EHLO.
2237 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2238 fixed by 4.65/MH/01 (is this a record?) are fixed:
2240 (i) An empty string was always treated as zero by the numeric comparison
2241 operators. This behaviour has been restored.
2243 (ii) It is documented that the numeric comparison operators always treat
2244 their arguments as decimal numbers. This was broken in that numbers
2245 starting with 0 were being interpreted as octal.
2247 While fixing these problems I realized that there was another issue that
2248 hadn't been noticed. Values of message_size_limit (both the global option
2249 and the transport option) were treated as octal if they started with 0.
2250 The documentation was vague. These values are now always treated as
2251 decimal, and I will make that clear in the documentation.
2257 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2258 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2261 MH/01 Don't check that the operands of numeric comparison operators are
2262 integers when their expansion is in "skipping" mode (fixes bug
2263 introduced by 4.64-PH/07).
2265 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2266 child addresses, Exim now panics and dies. Previously, because the count
2267 is held in a short int, deliveries were likely to be lost. As such a
2268 large number of recipients for a single message is ridiculous
2269 (performance will be very, very poor), I have chosen to impose a limit
2270 rather than extend the field.
2276 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2277 leftover -K file (the existence of which was triggered by #402).
2278 While we were at it, introduced process PID as part of the -K
2279 filename. This should rule out race conditions when creating
2282 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2283 processing considerably. Previous code took too long for large mails,
2284 triggering a timeout which in turn triggers #401.
2286 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2287 in the DK code in transports.c. sendfile() is not really portable,
2288 hence the _LINUX specificness.
2290 TF/01 In the add_headers option to the mail command in an Exim filter,
2291 there was a bug that Exim would claim a syntax error in any
2292 header after the first one which had an odd number of characters
2295 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2296 callout verification, Exim cached a "reject" for the entire domain. This
2297 is correct for most verifications, but it is not correct for a recipient
2298 verification with use_sender or use_postmaster set, because in that case
2299 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2300 case of MAIL FROM:<> rejection from other early rejections (e.g.
2301 rejection of HELO). When verifying a recipient using a non-null MAIL
2302 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2303 Whatever the result of the callout, the value of the domain cache is
2304 left unchanged (for any other kind of callout, getting as far as trying
2305 RCPT means that the domain itself is ok).
2307 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2310 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2311 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2312 session, the daemon ignores ECONNECT errors and logs others; it now
2313 ignores EPIPE as well.
2315 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2316 (quoted-printable decoding).
2318 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2319 later the small subsequent patch to fix an introduced bug.
2321 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2323 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2325 PH/08 An error is now given if message_size_limit is specified negative.
2327 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2328 to be given (somewhat) arbitrary names.
2330 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2333 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2334 miscellaneous code fixes
2336 PH/10 Added the log_reject_target ACL modifier to specify where to log
2339 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2340 hostname. This is wrong, because it relates to the incoming message (and
2341 probably the interface on which it is arriving) and not to the outgoing
2342 callout (which could be using a different interface). This has been
2343 changed to use the value of the helo_data option from the smtp transport
2344 instead - this is what is used when a message is actually being sent. If
2345 there is no remote transport (possible with a router that sets up host
2346 addresses), $smtp_active_hostname is used.
2348 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2349 tweaks were necessary in order to get it to work (see also 21 below):
2350 (a) The code assumed that strncpy() returns a negative number on buffer
2351 overflow, which isn't the case. Replaced with Exim's string_format()
2353 (b) There were several signed/unsigned issues. I just did the minimum
2354 hacking in of casts. There is scope for a larger refactoring.
2355 (c) The code used strcasecmp() which is not a standard C function.
2356 Replaced with Exim's strcmpic() function.
2357 (d) The code set only $1; it now sets $auth1 as well.
2358 (e) A simple test gave the error "authentication client didn't specify
2359 service in request". It would seem that Dovecot has changed its
2360 interface. Fortunately there's a specification; I followed it and
2361 changed what the client sends and it appears to be working now.
2363 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2366 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2367 address A is aliased to B and C, where B exists and C does not. Without
2368 -v the output is "A verified" because verification stops after a
2369 successful redirection if more than one address is generated. However,
2370 with -v the child addresses are also verified. Exim was outputting "A
2371 failed to verify" and then showing the successful verification for C,
2372 with its parentage. It now outputs "B failed to verify", showing B's
2373 parentage before showing the successful verification of C.
2375 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2376 look up a TXT record in a specific list after matching in a combined
2379 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2380 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2381 they consult the DNS. I had assumed they would set it the way they
2382 wanted; and indeed my experiments on Linux seem to show that in some
2383 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2384 To be on the safe side, however, I have now made the interface to
2385 host_find_byname() similar to host_find_bydns(), with an argument
2386 containing the DNS resolver options. The host_find_byname() function now
2387 sets these options at its start, just as host_find_bydns() does. The smtp
2388 transport options dns_qualify_single and dns_search_parents are passed to
2389 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2390 of host_find_byname() use the default settings of RES_DEFNAMES
2391 (qualify_single) but not RES_DNSRCH (search_parents).
2393 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2394 spool_read_header() do less string testing, by means of a preliminary
2395 switch on the second character of optional "-foo" lines. (This is
2396 overdue, caused by the large number of possibilities that now exist.
2397 Originally there were few.) While I was there, I also converted the
2398 str(n)cmp tests so they don't re-test the leading "-" and the first
2399 character, in the hope this might squeeze out yet more improvement.
2401 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2402 flag allowing group syntax was set by the header_syntax check but not
2403 turned off, possible causing trouble later; (2) The flag was not being
2404 set at all for the header_verify test, causing "group"-style headers to
2405 be rejected. I have now set it in this case, and also caused header_
2406 verify to ignore an empty address taken from a group. While doing this, I
2407 came across some other cases where the code for allowing group syntax
2408 while scanning a header line wasn't quite right (mostly, not resetting
2409 the flag correctly in the right place). These bugs could have caused
2410 trouble for malformed header lines. I hope it is now all correct.
2412 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2413 with the "reply" argument non-NULL. The code, however (which originally
2414 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2415 but it didn't always do it. This confused somebody who was copying the
2416 code for some other use. I have removed all the tests.
2418 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2419 feature that was used to support insecure browsers during the U.S. crypto
2420 embargo. It requires special client support, and Exim is probably the
2421 only MTA that supported it -- and would never use it because real RSA is
2422 always available. This code has been removed, because it had the bad
2423 effect of slowing Exim down by computing (never used) parameters for the
2424 RSA_EXPORT functionality.
2426 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2427 authenticator to fail if there's a tab character in the incoming data
2428 (there should never be unless someone is messing about, as it's supposed
2429 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2430 if the connection is using TLS or if the remote IP is the same as the
2431 local IP, and the "valid-client-cert option" if a client certificate has
2434 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2435 authenticators. This can be used for authorization after authentication
2436 succeeds. (In the case of plaintext, it servers for both authentication
2439 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2440 if any retry times were supplied.
2442 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2443 connection, where there is no client IP address to check. In this
2444 situation, the verify now always succeeds.
2446 PH/25 Applied John Jetmore's -Mset patch.
2448 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2450 PH/27 In a string expansion for a processed (not raw) header when multiple
2451 headers of the same name were present, leading whitespace was being
2452 removed from all of them, but trailing whitespace was being removed only
2453 from the last one. Now trailing whitespace is removed from each header
2454 before concatenation. Completely empty headers in a concatenation (as
2455 before) are ignored.
2457 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2458 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2460 PH/29 [Removed. This was a change that I later backed out, and forgot to
2461 correct the ChangeLog entry (that I had efficiently created) before
2462 committing the later change.]
2464 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2465 address errors (4xx response to RCPT) over the same connection as other
2466 messages routed to the same hosts. Such deliveries are always "forced",
2467 so retry times are not inspected. This resulted in far too many retries
2468 for the affected addresses. The effect occurred only when there were more
2469 hosts than the hosts_max_try setting in the smtp transport when it had
2470 the 4xx errors. Those hosts that it had tried were not added to the list
2471 of hosts for which the message was waiting, so if all were tried, there
2472 was no problem. Two fixes have been applied:
2474 (i) If there are any address or message errors in an SMTP delivery, none
2475 of the hosts (tried or untried) are now added to the list of hosts
2476 for which the message is waiting, so the message should not be a
2477 candidate for sending over the same connection that was used for a
2478 successful delivery of some other message. This seems entirely
2479 reasonable: after all the message is NOT "waiting for some host".
2480 This is so "obvious" that I'm not sure why it wasn't done
2481 previously. Hope I haven't missed anything, but it can't do any
2482 harm, as the worst effect is to miss an optimization.
2484 (ii) If, despite (i), such a delivery is accidentally attempted, the
2485 routing retry time is respected, so at least it doesn't keep
2486 hammering the server.
2488 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2489 in ${readsocket because some servers need this prod.
2491 PH/32 Added some extra debug output when updating a wait-xxx database.
2493 PH/33 The hint "could be header name not terminated by colon", which has been
2494 given for certain expansion errors for a long time, was not being given
2495 for the ${if def:h_colon_omitted{... case.
2497 PH/34 The spec says: "With one important exception, whenever a domain list is
2498 being scanned, $domain contains the subject domain." There was at least
2499 one case where this was not true.
2501 PH/35 The error "getsockname() failed: connection reset by peer" was being
2502 written to the panic log as well as the main log, but it isn't really
2503 panic-worthy as it just means the connection died rather early on. I have
2504 removed the panic log writing for the ECONNRESET error when getsockname()
2507 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2508 runs only) independently of the message's sender address. This meant
2509 that, if the 4xx error was in fact related to the sender, a different
2510 message to the same recipient with a different sender could confuse
2511 things. In particular, this can happen when sending to a greylisting
2512 server, but other circumstances could also provoke similar problems.
2513 I have changed the default so that the retry time for these errors is now
2514 based a combination of the sender and recipient addresses. This change
2515 can be overridden by setting address_retry_include_sender=false in the
2518 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2519 remote server are returned as part of bounce messages. This was not
2520 happening for LMTP over a pipe (the lmtp transport), but now it is the
2521 same for both kinds of LMTP.
2523 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2524 in header lines that were in fact CNAMEs. This is no longer the case.
2526 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2527 and queue runs started by the daemon processed all messages. This has
2528 been fixed so that -R and -S can now usefully be given with -q<time>.
2530 PH/40 Import PCRE release 6.7 (fixes some bugs).
2532 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2534 PH/42 Give an error if -q is specified more than once.
2536 PH/43 Renamed the variables $interface_address and $interface_port as
2537 $received_ip_address and $received_port, to make it clear that these
2538 values apply to message reception, and not to the outgoing interface when
2539 a message is delivered. (The old names remain recognized, of course.)
2541 PH/44 There was no timeout on the connect() call when using a Unix domain
2542 socket in the ${readsocket expansion. There now is.
2544 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2545 be meaningful with "accept".
2547 SC/01 Eximstats V1.43
2548 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2550 SC/02 Eximstats V1.44
2551 Use a glob alias rather than an array ref in the generated
2552 parser. This improves both readability and performance.
2554 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2555 Collect SpamAssassin and rejection statistics.
2556 Don't display local sender or destination tables unless
2557 there is data to show.
2558 Added average volumes into the top table text output.
2560 SC/04 Eximstats V1.46
2561 Collect data on the number of addresses (recipients)
2562 as well as the number of messages.
2564 SC/05 Eximstats V1.47
2565 Added 'Message too big' to the list of mail rejection
2566 reasons (thanks to Marco Gaiarin).
2568 SC/06 Eximstats V1.48
2569 Mainlog lines which have GMT offsets and are too short to
2570 have a flag are now skipped.
2572 SC/07 Eximstats V1.49 (Alain Williams)
2573 Added the -emptyok flag.
2575 SC/08 Eximstats V1.50
2576 Fixes for obtaining the IP address from reject messages.
2578 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2579 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2580 whitespace changes from 4.64-PH/27
2582 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2585 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2586 are found, allow negative numbers in numeric criteria)
2588 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2590 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2591 to match changes made in 4.64-PH/43
2593 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2595 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2596 "domain" is actually an IP address, and force a failure. This locks out
2597 those revolvers/nameservers that support "A-for-A" lookups, in
2598 contravention of the specifications.
2600 PH/48 When a host name was looked up from an IP address, and the subsequent
2601 forward lookup of the name timed out, the host name was left in
2602 $sender_host_name, contrary to the specification.
2604 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2605 restricted to single-key lookups, Exim was not diagnosing an error if
2606 * or *@ was used with a query-style lookup.
2608 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2610 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2611 long ago, but noone interested enough thought of it. Let's just say that
2612 the "1.1" means that there are some new functions that weren't there at
2613 some point in the past.
2615 PH/51 Error processing for expansion failure of helo_data from an smtp
2616 transport during callout processing was broken.
2618 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2619 tested/used via the -bh/-bhc/-bs options.
2621 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2622 bug, fixed in subsequent PCRE releases).
2624 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2625 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2627 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2633 SC/01 Use a glob alias rather than an array ref in eximstats generated
2634 parser. This improves both readability and performance.
2636 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2637 Don't display local sender or destination tables in eximstats unless
2638 there is data to show.
2639 Added average volumes into the eximstats top table text output.
2641 SC/03 Collect data on the number of addresses (recipients) as well
2642 as the number of messages in eximstats.
2644 TF/01 Correct an error in the documentation for the redirect router. Exim
2645 does (usually) call initgroups() when daemonizing.
2647 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2648 with consistent privilege compared to when running as a daemon.
2650 TF/03 Note in the spec that $authenticated_id is not set for local
2651 submissions from trusted users.
2653 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2654 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2656 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2657 by adding some example configuration directives to the default
2658 configuration file. A little bit of work is required to uncomment the
2659 directives and define how usernames and passwords are checked, but
2660 there is now a framework to start from.
2662 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2663 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2664 without this. I don't know how relevant this is to other LDAP libraries.
2666 PH/02 Add the verb name to the "unknown ACL verb" error.
2668 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2670 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2672 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2673 directory not expanded when it should be if an expanded home directory
2674 was set for the address (which is overridden by the transport).
2676 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2679 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2680 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2681 because it is too late at that time, and has no effect.
2683 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2684 security issue with \' (bugzilla #107). I could not use the
2685 PQescapeStringConn() function, because it needs a PGconn value as one of
2688 PH/08 When testing addresses using -bt, indicate those final addresses that
2689 are duplicates that would not cause an additional delivery. At least one
2690 person was confused, thinking that -bt output corresponded to deliveries.
2691 (Suppressing duplicates isn't a good idea as you lose the information
2692 about possibly different redirections that led to the duplicates.)
2694 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2695 systems where poll() doesn't work, in particular OS X.
2697 PH/10 Added more information to debugging output for retry time not reached.
2699 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2700 operations in malware.c.
2702 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2705 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2706 syslog_duplication set false, log lines that would normally be written
2707 both the the main log and to the reject log were not written to syslog at
2710 PH/14 In the default configuration, change the use of "message" in ACL warn
2711 statements to "add_header".
2713 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2714 not followed by a command (e.g. "seen endif").
2716 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2717 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2720 PH/17 Added extra conditions to the default value of delay_warning_condition
2724 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2725 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2726 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2729 The Auto-Submitted: and various List- headers are standardised, whereas I
2730 don't think Precedence: ever was.
2732 PH/18 Refactored debugging code in route_finduser() to show more information,
2733 in particular, the error code if getpwnam() issues one.
2735 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2736 This is apparently needed in addition to the PH/07 change above to avoid
2737 any possible encoding problems.
2739 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2740 but not after initializing Perl.
2742 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2743 output them only if debugging. By default they are written stderr,
2744 apparently, which is not desirable.
2746 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2749 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2752 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2754 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2755 authenticated or an ident call has been made. Suppress the default
2756 values for $authenticated_id and $authenticated_sender (but permit -oMai
2757 and -oMas) when testing with -bh.
2759 PH/24 Re-jigged the order of the tests in the default configuration so that the
2760 tests for valid domains and recipients precede the DNS black list and CSA
2761 tests, on the grounds that those ones are more expensive.
2763 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2764 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2765 This bug exists in every version of Exim that I still have, right back to
2768 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2769 However, an attempt to turn on case-sensitivity in a regex key by
2770 including (?-i) didn't work because the subject string was already
2771 lowercased, and the effects were non-intuitive. It turns out that a
2772 one-line patch can be used to allow (?-i) to work as expected.
2778 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2779 other effects) broke the use of negated acl sub-conditions.
2781 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2784 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2785 "Deny" causes Exim to reject the incoming connection with a 554 error.
2786 Unfortunately, if there is a major crisis, such as a disk failure,
2787 tcp-wrappers gives "deny", whereas what one would like would be some
2788 kind of temporary error. A kludge has been added to help with this.
2789 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2790 554 error is used if errno is still zero or contains ENOENT (which occurs
2791 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2794 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2796 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2797 errors. Otherwise a message that provokes a temporary error (when other
2798 messages do not) can cause a whole host to time out.
2800 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2801 addresses were routed directly to files or pipes from a redirect router.
2802 File deliveries just didn't batch; pipe deliveries might have suffered
2805 PH/06 A failure to get a lock for a hints database would erroneously always say
2806 "Failed to get write lock", even when it was really a read lock.
2808 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2809 of 0600. This has been changed to use the value of the lockfile_mode
2810 option (which defaults to 0600).
2812 PH/08 Applied small patch from the Sieve maintainer.
2814 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2815 folder from quota calculations, a direct delivery into this folder messed
2816 up the contents of the maildirsize file. This was because the regex was
2817 used only to exclude .Trash (or whatever) when the size of the mailbox
2818 was calculated. There was no check that a delivery was happening into an
2819 excluded directory. This bug has been fixed by ignoring all quota
2820 processing for deliveries into excluded directories.
2822 PH/10 Added the maildirfolder_create_regex option to appendfile.
2828 PH/01 The code for finding all the local interface addresses on a FreeBSD
2829 system running IPv6 was broken. This may well have applied to all BSD
2830 systems, as well as to others that have similar system calls. The broken
2831 code found IPv4 interfaces correctly, but gave incorrect values for the
2832 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2833 that it would not match correctly against @[] and not recognize the IPv6
2836 PH/02 The ipliteral router was not recognizing addresses of the form user@
2837 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2839 PH/03 Added disable_ipv6.
2841 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2842 decoded form, because it is most often used to construct To: headers
2843 lines in autoreplies, and the decoded form may well be syntactically
2844 invalid. However, $reply_address has leading white space removed, and all
2845 newlines turned into spaces so that the autoreply transport does not
2848 PH/05 If group was specified without a user on a router, and no group or user
2849 was specified on a transport, the group from the router was ignored.
2851 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2852 for visible compile-time settings that can be used to change these
2853 numbers, for those that want even more. Backwards compatibility with old
2854 spool files has been maintained. However, going back to a previous Exim
2855 release will lost any variables that are in spool files.
2857 PH/07 Two small changes when running in the test harness: increase delay when
2858 passing a TCP/IP connection to a new process, in case the original
2859 process has to generate a bounce, and remove special handling of
2860 127.0.0.2 (sic), which is no longer necessary.
2862 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2863 be the same on different OS.
2865 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2868 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2869 whether --show-vars was specified or not
2871 JJ/02 exipick: Added support for new ACL variable spool format introduced
2874 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2875 syntactically invalid From: or Reply-to: line, and a filter used this to
2876 generate an autoreply, and therefore failed to obtain an address for the
2877 autoreply, Exim could try to deliver to a non-existent relative file
2878 name, causing unrelated and misleading errors. What now happens is that
2879 it logs this as a hard delivery error, but does not attempt to create a
2882 PH/11 The exinext utility has a -C option for testing purposes, but although
2883 the given file was scanned by exinext itself; it wasn't being passed on
2884 when Exim was called.
2886 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2887 an end-of-file indication when reading a command response.
2889 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2890 compiled. In many other places in Exim, IPv6 addresses are always
2891 recognized, so I have changed this. It also means that IPv4 domain
2892 literals of the form [IPV4:n.n.n.n] are now always recognized.
2894 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2895 used if the router is not running as root, for example, when verifying at
2896 ACL time, or when using -bh. The debugging output from this situation was
2897 non-existent - all you got was a failure to exec. I have made two
2900 (a) Failures to set uid/gid, the current directory, or a process leader
2901 in a subprocess such as that created by queryprogram now generate
2902 suitable debugging output when -d is set.
2904 (b) The queryprogram router detects when it is not running as root,
2905 outputs suitable debugging information if -d is set, and then runs
2906 the subprocess without attempting to change uid/gid.
2908 PH/15 Minor change to Makefile for building test_host (undocumented testing
2911 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2912 additional section of a DNS packet that returns MX or SRV records.
2913 Instead, it always explicitly searches for A/AAAA records. This avoids
2914 major problems that occur when a DNS server includes only records of one
2915 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2916 fixed another bug: if SRV records were looked up and the corresponding
2917 address records were *not* found in the additional section, the port
2918 values from the SRV records were lost.
2920 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2921 using the correct key (the original address) when searching the retry
2922 rules in order to find which one to use for generating the retry hint.
2924 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2925 adding the default one. Similarly, if it contains a Reply-To: header, the
2926 errors_reply_to option, if set, is not used.
2928 PH/19 When calculating a retry time, Exim used to measure the "time since
2929 failure" by looking at the "first failed" field in the retry record. Now
2930 it does not use this if it is later than than the arrival time of the
2931 message. Instead it uses the arrival time. This makes for better
2932 behaviour in cases where some deliveries succeed, thus re-setting the
2933 "first failed" field. An example is a quota failure for a huge message
2934 when small messages continue to be delivered. Without this change, the
2935 "time since failure" will always be short, possible causing more frequent
2936 delivery attempts for the huge message than are intended.
2937 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2939 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2940 $1, $2, $3) because the numerical variables can be reset during some
2941 expansion items (e.g. "match"), thereby losing the authentication data.
2943 PH/21 Make -bV show the size of off_t variables so that the test suite can
2944 decide whether to run tests for quotas > 2G.
2946 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2947 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2948 filecount value is greater than 2G or if a quota value is greater than 2G
2949 on a system where the size of off_t is not greater than 4, a panic error
2952 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2953 never match. The debug and -bh output now contains an explicit error
2954 message indicating a malformed IPv4 address or mask.
2956 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2957 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2958 PH/23 above applies.
2960 PH/25 Do not write to syslog when running in the test harness. The only
2961 occasion when this arises is a failure to open the main or panic logs
2962 (for which there is an explicit test).
2964 PH/26 Added the /no_tell option to "control=freeze".
2966 PH/27 If a host name lookup failed very early in a connection, for example, if
2967 the IP address matched host_lookup and the reverse lookup yielded a name
2968 that did not have a forward lookup, an error message of the form "no IP
2969 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2970 could be logged. Now it outputs the IP address instead of "NULL".
2972 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2973 allows the sender and the authenticated sender to be set when
2974 submitting a message from within Exim. Since child_open_exim() is
2975 documented for local_scan(), the new function should be too.
2977 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2978 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2979 results in an empty string is now treated as unset.
2981 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2983 PH/31 Added sender_verify_fail logging option.
2985 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2986 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2987 tidied the source and removed it altogether.
2989 PH/33 When a queue run was abandoned because the load average was too high, a
2990 log line was always written; now it is written only if the queue_run log
2991 selector is set. In addition, the log line for abandonment now contains
2992 information about the queue run such as the pid. This is always present
2993 in "start" and "stop" lines but was omitted from the "abandon" line.
2995 PH/34 Omit spaces between a header name and the colon in the error message that
2996 is given when verify = headers_syntax fails (if there are lots of them,
2997 the message gets confusing).
2999 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3000 names, as there are now some PTR records that contain slashes. This check
3001 is only to protect against broken name servers that fall over on strange
3002 characters, so the fact that it applies to all lookups doesn't matter.
3004 PH/36 Now that the new test suite is complete, we can remove some of the
3005 special code in Exim that was needed for the old test suite. For example,
3006 sorting DNS records because real resolvers return them in an arbitrary
3007 order. The new test suite's fake resolver always returns records in the
3010 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3011 bounces) except when queue_only is set, to avoid logging races between
3012 the different processes.
3014 PH/38 Panic-die if .include specifies a non-absolute path.
3016 PH/39 A tweak to the "H" retry rule from its user.
3018 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3019 a label. They prevented compilation on older perls.
3021 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3022 a warning to be raised on newish perls.
3024 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3025 on queue. Changes to match documented behaviour of showing count of
3026 messages matching specified criteria.
3028 PH/40 Changed the default ident timeout from 30s to 5s.
3030 PH/41 Added support for the use of login_cap features, on those BSD systems
3031 that have them, for controlling the resources used by pipe deliveries.
3033 PH/42 The content-scanning code uses fopen() to create files in which to put
3034 message data. Previously it was not paying any attention to the mode of
3035 the files. Exim runs with umask(0) because the rest of the code creates
3036 files with open(), and sets the required mode explicitly. Thus, these
3037 files were ending up world-writeable. This was not a big issue, because,
3038 being within the spool directory, they were not world-accessible. I have
3039 created a function called modefopen, which takes an additional mode
3040 argument. It sets umask(777), creates the file, chmods it to the required
3041 mode, then resets the umask. All the relevant calls to fopen() in the
3042 content scanning code have been changed to use this function.
3044 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3045 to 24 hours. This avoids potential overflow problems when processing G
3046 and H retry rules. I suspect nobody ever tinkers with this value.
3048 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3050 PH/45 When the plaintext authenticator is running as a client, the server's
3051 challenges are checked to ensure they are valid base64 strings. By
3052 default, the authentication attempt is cancelled if an invalid string is
3053 received. Setting client_ignore_invalid_base64 true ignores these errors.
3054 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3055 they are received. Thus, the responses can be made to depend on the
3056 challenges. If an invalid string is ignored, an empty string is placed in
3059 PH/46 Messages that are created by the autoreply transport now contains a
3060 References: header, in accordance with RFCs 2822 and 3834.
3062 PH/47 Added authenticated_sender_force to the smtp transport.
3064 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3066 PH/49 Installed latest patch from the Sieve maintainer.
3068 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3069 was also set, the appendfile transport was unnecessarily scanning a
3070 directory of message files (e.g. for maildir delivery) to find the count
3071 of files (along with the size), even though it did not need this
3072 information. It now does the scan only if it needs to find either the
3073 size of the count of files.
3075 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3077 PH/52 Two bugs concerned with error handling when the smtp transport is
3080 (i) Exim was not creating retry information for temporary errors given
3081 for individual recipients after the DATA command when the smtp transport
3082 was used in LMTP mode. This meant that they could be retried too
3083 frequently, and not timed out correctly.
3085 (ii) Exim was setting the flag that allows error details to be returned
3086 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3087 recipients that were returned after the DATA command.
3089 PH/53 This is related to PH/52, but is more general: for any failing address,
3090 when detailed error information was permitted to be returned to the
3091 sender, but the error was temporary, then after the final timeout, only
3092 "retry timeout exceeded" was returned. Now it returns the full error as
3093 well as "retry timeout exceeded".
3095 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3096 do this, and (what is worse) MTAs that accept it.
3098 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3099 will now be deprecated.
3101 PH/56 New os.c-cygwin from the Cygwin maintainer.
3103 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3104 formats (previously only available in exim formats via -bpr, -bpru,
3105 and -bpra. Now also available in native and exiqgrep formats)
3107 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3108 with very large, slow to parse queues
3110 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3112 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3114 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3115 responses to authentication challenges, though it was showing the
3116 challenges; (ii) I've removed the CR characters from the debug output for
3119 PH/58 Allow for the insertion of a newline as well as a space when a string
3120 is turned into more than one encoded-word during RFC 2047 encoding. The
3121 Sieve code now uses this.
3123 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3124 data_4xx, lost_connection, tls_required.
3126 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3127 message was being sent as an SMTP response.
3129 PH/61 Add -l and -k options to exicyclog.
3131 PH/62 When verifying, if an address was redirected to one new address, so that
3132 verification continued, and the new address failed or deferred after
3133 having set something in $address_data, the value of $address_data was not
3134 passed back to the ACL. This was different to the case when no
3135 redirection occurred. The value is now passed back in both cases.
3137 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3138 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3139 use login_cap.h, so on its own it isn't the distinguishing feature. The
3140 new name refers directly to the setclassresources() function.
3142 PH/65 Added configuration files for NetBSD3.
3144 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3146 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3147 is preferred over IPv4.
3149 PH/68 The bounce_return_message and bounce_return_body options were not being
3150 honoured for bounces generated during the reception of non-SMTP messages.
3151 In particular, this applied to messages rejected by the ACL. This bug has
3152 been fixed. However, if bounce_return_message is true and bounce_return_
3153 body is false, the headers that are returned for a non-SMTP message
3154 include only those that have been read before the error was detected.
3155 (In the case of an ACL rejection, they have all been read.)
3157 PH/69 The HTML version of the specification is now built in a directory called
3158 spec_html instead of spec.html, because the latter looks like a path with
3159 a MIME-type, and this confuses some software.
3161 PH/70 Catch two compiler warnings in sieve.c.
3163 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3164 function verify_get_ident() calls ip_connect() to connect a socket, but
3165 if the "connect()" function timed out, ip_connect() used to close the
3166 socket. However, verify_get_ident() also closes the socket later, and in
3167 between Exim writes to the log, which may get opened at this point. When
3168 the socket was closed in ip_connect(), the log could get the same file
3169 descriptor number as the socket. This naturally causes chaos. The fix is
3170 not to close the socket in ip_connect(); the socket should be closed by
3171 the function that creates it. There was only one place in the code where
3172 this was missing, in the iplookup router, which I don't think anybody now
3173 uses, but I've fixed it anyway.
3175 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3176 well as to direct DNS lookups. Otherwise the handling of names in host
3177 lists is inconsistent and therefore confusing.
3183 PH/01 Two changes to the default runtime configuration:
3185 (1) Move the checks for relay_from_hosts and authenticated clients from
3186 after to before the (commented out) DNS black list checks.
3188 (2) Add control=submission to the relay_from_hosts and authenticated
3189 clients checks, on the grounds that messages accepted by these
3190 statements are most likely to be submissions.
3192 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3194 (1) Generate an error if the third argument for the ${prvs expansion is
3197 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3200 (3) Reset the variables that are obtained from the first argument of
3201 ${prvscheck and used in the second argument before leaving the code,
3202 because their memory is reclaimed, so using them afterwards may do
3205 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3206 one (it's much easier than Tom thought :-).
3208 (5) Because of (4), we can now allow for the use of $prvscheck_result
3209 inside the third argument.
3211 PH/03 For some reason, the default setting of PATH when running a command from
3212 a pipe transport was just "/usr/bin". I have changed it to
3215 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3216 anything to be listed in the output from -bV.
3218 PH/05 When a filter generated an autoreply, the entire To: header line was
3219 quoted in the delivery log line, like this:
3221 => >A.N.Other <ano@some.domain> <original@ddress> ...
3223 This has been changed so that it extracts the operative address. There
3224 may be more than one such address. If so, they are comma-separated, like
3227 => >ano@some.domain,ona@other.domain <original@ddress> ...
3229 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3230 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3231 not being looked up in the rDNS to get a host name, Exim was showing the
3232 IP address twice in Received: lines, even though the IP addresses were
3233 identical. For example:
3235 Received: from [1.2.3.4] (helo=[1.2.3.4])
3237 However, if the real host name was known, it was omitting the HELO data
3238 if it matched the actual IP address. This has been tidied up so that it
3239 doesn't show the same IP address twice.
3241 PH/07 When both +timestamp and +memory debugging was on, the value given by
3242 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3243 called by the debug printing, thereby overwriting the timestamp buffer.
3244 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3246 PH/08 When the original message was included in an autoreply transport, it
3247 always said "this is a copy of the message, including all the headers",
3248 even if body_only or headers_only was set. It now gives an appropriate
3251 PH/09 Applied a patch from the Sieve maintainer which:
3253 o fixes some comments
3254 o adds the (disabled) notify extension core
3255 o adds some debug output for the result of if/elsif tests
3256 o points to the current vacation draft in the documentation
3257 and documents the missing references header update
3261 o fixes a bug in processing the envelope test (when testing
3262 multiple envelope elements, the last element determined the
3265 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3266 Electronic Mail") by including:
3268 Auto-submitted: auto-generated
3270 in the messages that it generates (bounce messages and others, such as
3271 warnings). In the case of bounce messages for non-SMTP messages, there was
3272 also a typo: it was using "Auto_submitted" (underscore instead of
3273 hyphen). Since every message generated by Exim is necessarily in response
3274 to another message, thes have all been changed to:
3276 Auto-Submitted: auto-replied
3278 in accordance with these statements in the RFC:
3280 The auto-replied keyword:
3282 - SHOULD be used on messages sent in direct response to another
3283 message by an automatic process,
3285 - MUST NOT be used on manually-generated messages,
3287 - MAY be used on Delivery Status Notifications (DSNs) and Message
3288 Disposition Notifications (MDNs),
3290 - MUST NOT be used on messages generated by automatic or periodic
3291 processes, except for messages which are automatic responses to
3294 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3295 to the default Received: header definition.
3297 PH/12 Added log selector acl_warn_skipped (default on).
3299 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3300 variables because (a) they are in the wrong storage pool and (b) even if
3301 they were copied, it wouldn't work properly because of the caching.
3303 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3304 checking when decoding. Apparently there are clients that generate
3305 overlong encoded strings. Why am I not surprised?
3307 PH/15 If the first argument of "${if match_address" was not empty, but did not
3308 contain an "@" character, Exim crashed. Now it writes a panic log message
3309 and treats the condition as false.
3311 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3313 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3314 extension "envelope-auth". The code is finished and in agreement with
3315 other implementations, but there is no documentation so far and in fact,
3316 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3317 not changing the active code.
3319 Print executed "if" and "elsif" statements when debugging is used. This
3320 helps a great deal to understand what a filter does.
3322 Document more things not specified clearly in RFC3028. I had all this
3323 sorted out, when out of a sudden new issues came to my mind. Oops."
3325 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3328 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3329 canonical form (as documented). However, after a host name lookup from
3330 the IP address, check_host() was doing a simple string comparison with
3331 addresses acquired from the DNS when checking that the found name did
3332 have the original IP as one of its addresses. Since any found IPv6
3333 addresses are likely to be in abbreviated form, the comparison could
3334 fail. Luckily, there already exists a function for doing the comparison
3335 by converting both addresses to binary, so now that is used instead of
3336 the text comparison.
3338 PH/20 There was another similar case to PH/19, when a complete host name was
3339 given in a host list; looking up its IP address could give an abbreviated
3340 form, whereas the current host's name might or might not be abbreviated.
3341 The same fix has been applied.
3347 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3348 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3351 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3352 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3354 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3356 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3357 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3358 Also, for "Auto-Submitted", treat anything other than "no" as
3359 non-personal, in accordance with RFC 3834. (Previously it treated
3360 anything starting "auto-" as non-personal.)
3362 TF/01 The control=submission/name=... option had a problem with syntax
3363 errors if the name included a slash character. The /name= option
3364 now slurps the rest of the string, so it can include any characters
3365 but it must come last in the list of options (after /sender_retain
3368 PH/05 Some modifications to the interface to the fake nameserver for the new
3376 TK/01 Added the "success_on_redirect" address verification option. See
3377 NewStuff for rationale and an example.
3379 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3381 PH/02 Patch to exigrep to allow it to work on syslog lines.
3383 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3384 fread() to skip over the body file's header line, because in Cygwin the
3385 header line is locked and is inaccessible.
3387 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3388 co-exist for some time) to make it clear that it is the Exim ID that is
3389 referenced, not the Message-ID: header line.
3391 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3392 string_format() function, because snprintf() does not exist on all
3395 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3396 ${stat: expansion item.
3398 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3399 protocol synchronization error", to keep the pedants happy.
3401 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3402 well as for IRIX systems, when gcc is being used. See the host.c source
3405 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3407 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3410 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3411 transport and to the smtp transport in LMTP mode.
3413 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3415 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3416 run a filter in a subprocess. This could lead to confusion in subsequent
3417 lookups in the parent process. There should also be a search_tidyup() at
3418 the end of the subprocess.
3420 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3421 only if the host matched helo_try_verify_hosts, which caused the
3422 verification to occur when the EHLO/HELO command was issued. The ACL just
3423 tested the remembered result. Now, if a previous verification attempt has
3424 not happened, "verify = helo" does it there and then.
3426 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3428 TK/03 Fix log output including CR from clamd.
3430 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3431 exist provoked a memory error which could cause a segfault.
3433 PH/15 Installed PCRE 6.2
3435 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3437 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3438 of the problem. Specifically, suggested +O2 rather than +O1 for the
3441 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3443 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3444 record, the port was not correctly logged when the outgoing_port log
3445 selector was set (it logged the transort's default port).
3447 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3448 fallback_hosts, and "hosts" in the smtp transport.
3450 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3451 host errors such as "Connection refused".
3453 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3454 authentication with radiusclient 0.4.9:
3456 - Error returned from rc_read_config was caught wrongly
3457 - Username/password not passed on to radius server due to wrong length.
3459 The presumption is that some radiusclient API changes for 4.51/PH/17
3460 were not taken care of correctly. The code is still untested by me (my
3461 Linux distribution still has 0.3.2 of radiusclient), but it was
3462 contributed by a Radius user.
3464 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3465 expanding the "port" option of the smtp transport.
3467 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3468 while reading a MIME header. Thanks to Tom Hughes for a patch.
3470 PH/24 Include config.h inside local_scan.h so that configuration settings are
3473 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3474 that in an ACL for RCPT (for example), you can examine exactly what was
3477 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3478 commands, but it was not correctly comparing the address with the actual
3479 client host address. Thus, it would show the EHLO address in Received:
3480 header lines when this was not necessary.
3482 PH/27 Added the % operator to ${eval:}.
3484 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3485 it should be ignoring failures (because with -C, for example, it has lost
3486 privilege). It wasn't ignoring creation failures other than "already
3489 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3492 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3493 because an input line was too long, either on its own, or by virtue of
3494 too many continuations, the temporary file was not being removed, and the
3495 return code was incorrect.
3497 PH/31 Missing "BOOL" in function definition in filtertest.c.
3499 PH/32 Applied Sieve patches from the maintainer.
3501 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3503 PH/33 Added "verify = not_blind".
3505 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3506 Local/Makefile (with some defaults set). These are used in built scripts
3507 such as exicyclog, but they have never been used in the exim_install
3508 script (though there are many overriding facilities there). I have
3509 arranged that the exim_install script now takes note of these two
3512 PH/35 Installed configuration files for Dragonfly.
3514 PH/36 When a locally submitted message by a trusted user did not contain a
3515 From: header, and the sender address was obtained from -f or from an SMTP
3516 MAIL command, and the trusted user did not use -F to supply a sender
3517 name, $originator_name was incorrectly used when constructing a From:
3518 header. Furthermore, $originator_name was used for submission mode
3519 messages from external hosts without From: headers in a similar way,
3520 which is clearly wrong.
3522 PH/37 Added control=suppress_local_fixups.
3524 PH/38 When log_selector = +received_sender was set, and the addition of the
3525 sender made the log line's construction buffer exactly full, or one byte
3526 less than full, an overflow happened when the terminating "\n" was
3529 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3530 when the result of a list match is failure because a DNS lookup failed.
3532 PH/40 RM_COMMAND is now used in the building process.
3534 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3535 the "build-* directories that it finds.
3537 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3538 address in a domain literal was a prefix of an interface address.
3540 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3541 when verifying a sender address, unless rewrite_headers is false.
3543 PH/44 Wrote a long comment about why errors_to addresses are verified as
3544 recipients, not senders.
3546 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3547 the ratelimit ACL was added.
3549 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3551 PH/46 Added extra information about PostgreSQL errors to the error string.
3553 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3554 avoiding the need to install special zones in a real server. This is
3555 backwards compatible; if it can't find the fake resolver, it drops back.
3556 Thus, both old and new test suites can be run.
3558 TF/02 Added util/ratelimit.pl
3560 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3561 clock is set back in time.
3563 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3564 Candler <B.Candler@pobox.com>.
3566 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3567 OK for addresses that are the result of redirections.
3569 PH/48 A number of further additions for the benefit of the new test suite,
3570 including a fake gethostbyname() that interfaces to the fake DNS resolver
3573 TF/06 The fix for widen_domains has also been applied to qualify_single and
3574 search_parents which are the other dnslookup options that can cause
3577 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3580 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3582 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3583 service" if it encountered an MX record with an empty target hostname.
3584 The message is now "an MX or SRV record indicated no SMTP service".
3586 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3587 if the DNS of the sending site is misconfigured. This is quite a
3588 common situation. This change restores the behaviour of treating a
3589 helo verification defer as a failure.
3591 PH/51 If self=fail was set on a router, the bounce message did not include the
3592 actual error message.
3598 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3600 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3601 command itself ended in error, the underlying message about the transport
3602 filter timeout was being overwritten with the pipe command error. Now the
3603 underlying error message should be appended to the second error message.
3605 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3607 PH/02 Reduce the amount of output that "make" produces by default. Full output
3608 can still be requested.
3610 PH/03 The warning log line about a condition test deferring for a "warn" verb
3611 was being output only once per connection, rather than after each
3612 occurrence (because it was using the same function as for successful
3613 "warn" verbs). This seems wrong, so I have changed it.
3615 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3616 it should not have, which might have caused a crash in the right
3617 circumstances, but probably never did.
3619 PH/04 Installed a modified version of Tony Finch's patch to make submission
3620 mode fix the return path as well as the Sender: header line, and to
3621 add a /name= option so that you can make the user's friendly name appear
3624 TF/03 Added the control = fakedefer ACL modifier.
3626 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3627 Mark Lowes for thorough testing.
3629 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3631 TK/03 Merged latest SRS patch from Miles Wilton.
3633 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3634 with the definition in sysexits.h (which is #included earlier).
3635 Fortunately, Exim does not actually use EX_OK. The code used to try to
3636 preserve the sysexits.h value, by assuming that macro definitions were
3637 scanned for macro replacements. I have been disabused of this notion,
3638 so now the code just undefines EX_OK before #including unistd.h.
3640 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3641 in the smtp transport. When a block could not be written in a single
3642 write() function, the timeout was being re-applied to each part-write.
3643 This seems wrong - if the receiver was accepting one byte at a time it
3644 would take for ever. The timeout is now adjusted when this happens. It
3645 doesn't have to be particularly precise.
3647 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3648 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3650 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3651 without a domain if the check to <postmaster@domain> fails.
3653 SC/01 Eximstats: added -xls and the ability to specify output files
3654 (patch written by Frank Heydlauf).
3656 SC/02 Eximstats: use FileHandles for outputting results.
3658 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3660 SC/04 Eximstats: fixed display of large numbers with -nvr option
3662 SC/05 Eximstats: fixed merging of reports with empty tables.
3664 SC/06 Eximstats: added the -include_original_destination flag
3666 SC/07 Eximstats: removed tabs and trailing whitespace.
3668 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3670 TK/06 MBOX spool code: Add real "From " MBOX separator line
3671 so the .eml file is really in mbox format (even though
3672 most programs do not really care). Patch from Alex Miller.
3674 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3675 The latter is generated from $received_to and is only set if the
3676 message has one envelope recipient. SA can use these headers,
3677 obviously out-of-the-box. Patch from Alex Miller.
3679 PH/08 The ${def test on a variable was returning false if the variable's
3680 value was "0", contrary to what the specification has always said!
3681 The result should be true unless the variable is empty.
3683 PH/09 The syntax error of a character other than { following "${if
3684 def:variable_name" (after optional whitespace) was not being diagnosed.
3685 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3686 accidental colon was present, for example, could give incorrect results.
3688 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3689 result is used (not including appendfile, where other changes are about
3692 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3693 This involved changing a lot of size variables from int to off_t. It
3694 should work with maildirs and everything.
3696 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3697 spamd dying while we are connected to it.
3699 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3702 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3703 function for BDB 4.3.
3705 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3707 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3708 the sources for PCRE have been re-arranged and more files are now
3711 PH/15 The code I had for printing potentially long long variables in PH/11
3712 above was not the best (it lost precision). The length of off_t variables
3713 is now inspected at build time, and an appropriate printing format (%ld
3714 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3715 to be "long long int" or "long int". This is needed for the internal
3716 formatting function string_vformat().
3718 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3719 the configuration file to be ":syslog", then the script "guesses" where
3720 the logs files are, rather than using the compiled in default. In our
3721 case the guess is not the same as the compiled default, so the script
3722 suddenly stopped working when I started to use syslog. The patch checks
3723 to see if log_file_path is "". If so, it attempts to read it from exim
3724 with no configuration file to get the compiled in version, before it
3725 falls back to the previous guessing code."
3727 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3728 implementing BATV in an Exim configuration. See NewStuff for the gory
3731 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3732 Makefile that are specific to HP-UX.
3734 PH/18 If the "use_postmaster" option was set for a recipient callout together
3735 with the "random" option, the postmaster address was used as the MAIL
3736 FROM address for the random test, but not for the subsequent recipient
3737 test. It is now used for both.
3739 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3740 patch removes a few documentation additions to RFC 3028, because the
3741 latest draft now contains them. It adds the new en;ascii-case comparator
3742 and a new error check for 8bit text in MIME parts. Comparator and
3743 require names are now matched exactly. I enabled the subaddress
3744 extension, but it is not well tested yet (read: it works for me)."
3746 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3747 rework some of the code of TK/09 above to avoid the hardwired use of
3748 "%lld" and "long long". Replaced the call to snprintf() with a call to
3751 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3752 records point to non-existent hosts", "retry timeout exceeded", and
3753 "retry time not reached for any host after a long failure period".
3755 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3756 experimental DomainKeys support:
3758 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3759 (2) On an error such as an illegally used "control", the wrong name for
3760 the control was given.
3762 These problems did NOT occur unless DomainKeys support was compiled.
3764 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3766 PH/24 Added ${if match_ip condition.
3768 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3769 databases so that it will be absolutely obvious if a crash occurs in the
3770 DB library. This is a regular occurrence (often caused by mis-matched
3773 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3774 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3775 picked up on a user's system that detects such things. There doesn't seem
3776 to be a gcc warning option for this - only an attribute that has to be
3777 put on the function's prototype. It seems that in Fedora Core 4 they have
3778 set this on a number of new functions. No doubt there will be more in due
3781 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3782 specify a transport. However, if an address that was verified by such a
3783 router was the subject of a callout, Exim crashed because it tried to
3784 read the rcpt_include_affixes from the non-existent transport. Now it
3785 just assumes that the setting of that option is false. This bug was
3786 introduced by 4.51/PH/31.
3788 PH/28 Changed -d+all to exclude +memory, because that information is very
3789 rarely of interest, but it makes the output a lot bigger. People tend to
3790 do -d+all out of habit.
3792 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3793 code in os-type was giving problems when libc.so lives in lib64, like on
3796 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3797 aren't the modern standard, and it seems that some systems' include files
3798 don't always have them. Exim was already checking for some of the newer
3799 ones like T_AAAA, and defining it itself. I've added checks for all the
3800 record types that Exim uses.
3802 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3803 not automatically generating a new one, as it is supposed to. This
3804 prevented TLS from working. If the file did exist, but contained invalid
3805 data, a new version was generated, as expected. It was only the case of a
3806 non-existent file that was broken.
3808 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3809 with a change in libdomainkeys > 0.64.
3811 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3812 from DNS. If the selector record carries the flag, it now has
3813 precedence over the domain-wide flag.
3815 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3817 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3818 the use of a port name that isn't defined in /etc/services) occurred, the
3819 message was deferred as in a normal delivery, and thus remained on the
3820 spool, instead of being failed because of the mua_wrapper setting. This
3821 is now fixed, and I tidied up some of the mua_wrapper messages at the
3824 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3825 the messages in a hash of arrays rather than using individual hashes.
3826 This is a bit cleaner and results in dramatic memory savings, albeit
3827 at a slight CPU cost.
3829 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3830 as requested by Marc Sherman.
3832 SC/10 Eximstats: added histograms for user specified patterns as requested
3835 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3837 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3838 fopen() in the content-scanning modules that did not already have it.
3844 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3845 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3847 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3849 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3851 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3852 to test Sieve filters that use "vacation".
3854 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3855 that changes the way the GnuTLS parameters are stored in the cache file.
3856 The new format can be generated externally. For backward compatibility,
3857 if the data in the cache doesn't make sense, Exim assumes it has read an
3858 old-format file, and it generates new data and writes a new file. This
3859 means that you can't go back to an older release without removing the
3862 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3863 work if there are any delivery delays because "one_time" forces the
3864 parent to be marked "delivered", so its unseen clone is never tried
3865 again. For this reason, Exim now forbids the simultaneous setting of
3868 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3869 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3870 entry if you want to know the details. The fix, however, neglected to
3871 consider the case when local delivery batching is involved. The test for
3872 "previously delivered" was not happening when checking to see if an
3873 address could be batched with a previous (undelivered) one; under
3874 certain circumstances this could lead to multiple deliveries to the same
3877 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3878 in its include files, and this causes problems building Exim.
3880 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3881 header_syntax) but Exim was just ignoring anything given after a slash.
3882 In particular, this caused confusion with an attempt to use "verify =
3883 reverse_host_lookup/defer_ok". An error is now given when options are
3884 supplied for verify items that do not have them. (Maybe reverse_host_
3885 lookup should have a defer_ok option, but that's a different point.)
3887 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3888 defined by RFC 821) to 2048, because there were problems with some AUTH
3889 commands, and RFC 1869 says the size should be increased for extended
3890 SMTP commands that take arguments.
3892 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3895 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3896 "unknown" error; now it says that the functionality isn't in the binary.
3898 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3899 an address' error message when a string expansion fails (syntax or
3900 whatever). Otherwise the password may appear in the log. Following change
3901 PH/42 below, there is no longer a chance of it appearing in a bounce
3904 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3906 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3907 of its addresses was ever tried. (Bugzilla bug #2.)
3909 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3910 the result incorrectly in the debug output. (It correctly added a newline
3911 to what was transported.)
3913 TF/01 Added $received_time.
3915 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3916 commented out examples of how to interface to a virus scanner and to
3917 SpamAssassin. Also added commented examples of av_scanner and
3918 spamd_address settings.
3920 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3921 and controls are allowed in which ACLs. There were a couple of minor
3922 errors. Some of the entries in the conditions table (which is a table of
3923 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3924 negation of where the condition IS allowed.
3926 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3928 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3929 header file does not have a version number, so I've had to invent a new
3930 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3931 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3932 radiusclient), but it was contributed by a Radius user.
3934 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3935 files or directories, for OpenSSL.
3937 PH/19 When an Exim process that is running as root has to create an Exim log
3938 file, it does so in a subprocess that runs as exim:exim so as to get the
3939 ownership right at creation (otherwise, other Exim processes might see
3940 the file with the wrong ownership). There was no test for failure of this
3941 fork() call, which would lead to the process getting stuck as it waited
3942 for a non-existent subprocess. Forks do occasionally fail when resources
3943 run out. I reviewed all the other calls to fork(); they all seem to check
3946 PH/20 When checking for unexpected SMTP input at connect time (before writing
3947 the banner), Exim was not dealing correctly with a non-positive return
3948 from the read() function. If the client had disconnected by this time,
3949 the result was a log entry for a synchronization error with an empty
3950 string after "input=" when read() returned zero. If read() returned -1
3951 (an event I could not check), uninitialized data bytes were printed.
3952 There were reports of junk text (parts of files, etc) appearing after
3955 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3957 PH/22 Added support for macro redefinition, and (re)definition in between
3958 driver and ACL definitions.
3960 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3961 forgetting to use the resulting value; it was using the unexpanded value.
3963 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3964 hadn't been configured. The fix is from Juergen Kreileder, who
3965 understands it better than I do:
3967 "Here's what I see happening with three configured cyrus_sasl
3968 authenticators configured (plain, login, cram-md5):
3970 On startup auth_cyrus_sasl_init() gets called for each of these.
3971 This means three calls to sasl_listmech() without a specified mech_list.
3972 => SASL tests which mechs of all available mechs actually work
3973 => three warnings about OTP not working
3974 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3976 With the patch, sasl_listmech() also gets called three times. But now
3977 SASL's mech_list option is set to the server_mech specified in the the
3978 authenticator. Or in other words, the answer from sasl_listmech()
3979 gets limited to just the mech you're testing for (which is different
3981 => the return list contains just 'plain' or 'login', 'cram-md5' or
3982 nothing depending on the value of ob->server_mech.
3984 I've just tested the patch: Authentication still works fine,
3985 unavailable mechs specified in the exim configuration are still
3986 caught, and the auth.log warnings about OTP are gone."
3988 PH/25 When debugging is enabled, the contents of the command line are added
3989 to the debugging output, even when log_selector=+arguments is not
3992 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3993 answer is "GNU", and only if the return is "GNU/something" is the answer
3996 PH/27 $acl_verify_message is now set immediately after the failure of a
3997 verification in an ACL, and so is available in subsequent modifiers. In
3998 particular, the message can be preserved by coding like this:
4000 warn !verify = sender
4001 set acl_m0 = $acl_verify_message
4003 Previously, $acl_verify_message was set only while expanding "message"
4004 and "log_message" when a very denied access.
4006 PH/28 Modified OS/os.c-Linux with
4008 -#ifndef OS_LOAD_AVERAGE
4009 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4011 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4012 nomenclature these days.)
4014 PH/29 Installed patch from the Sieve maintainer that adds the options
4015 sieve_useraddress and sieve_subaddress to the redirect router.
4017 PH/30 In these circumstances:
4018 . Two addresses routed to the same list of hosts;
4019 . First host does not offer TLS;
4020 . First host accepts first address;
4021 . First host gives temporary error to second address;
4022 . Second host offers TLS and a TLS session is established;
4023 . Second host accepts second address.
4024 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4025 and peerdn, if requested) that were in fact used only for the second
4028 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4029 attention to any local part prefix or suffix that was matched by the
4030 router that accepted the address. It now behaves in the same way as it
4031 does for delivery: the affixes are removed from the local part unless
4032 rcpt_include_affixes is set on the transport.
4034 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4035 timeout during the DATA phase of an incoming message.
4037 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4038 applied a patch sanctioned by the Sieve maintainer.
4040 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4041 the uid or gid is negative. A case of a negative gid caused this to be
4042 noticed. The fix allows for either to be negative.
4044 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4045 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4048 PH/36 The change PH/12 above was broken. Fixed it.
4050 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4051 the grounds that routing the same address twice would always produce the
4052 same answer. This might have been true once, but it is certainly no
4053 longer true now. Routing a child address may depend on the previous
4054 routing that produced that child. Some complicated redirection strategies
4055 went wrong when messages had multiple recipients, and made Exim's
4056 behaviour dependent on the order in which the addresses were given.
4058 I have moved the duplicate checking until after the routing is complete.
4059 Exim scans the addresses that are assigned to local and remote
4060 transports, and removes any duplicates. This means that more work will be
4061 done, as duplicates will always all be routed, but duplicates are
4062 presumably rare, so I don't expect this is of any significance.
4064 For deliveries to pipes, files, and autoreplies, the duplicate checking
4065 still happens during the routing process, since they are not going to be
4068 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4069 It corrects a timeout issue with spamd. This is Ian's comment: "The
4070 background is that sometimes spamd either never reads data from a
4071 connection it has accepted, or it never writes response data. The exiscan
4072 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4073 blindly assumes that writes won't block so it may never time out."
4075 PH/39 Allow G after quota size as well as K and M.
4077 PH/40 The value set for $authenticated_id in an authenticator may not contain
4078 binary zeroes or newlines because the value is written to log lines and
4079 to spool files. There was no check on this. Now the value is run through
4080 the string_printing() function so that such characters are converted to
4081 printable escape sequences.
4083 PH/41 $message_linecount is a new variable that contains the total number of
4084 lines in the message. Compare $body_linecount, which is the count for the
4087 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4088 bounce and delay warning messages, except in certain special cases, which
4091 (a) An SMTP error message from a remote host;
4092 (b) A message specified in a :fail: redirection;
4093 (c) A message specified in a "fail" command in a system filter;
4094 (d) A message specified in a FAIL return from the queryprogram router;
4095 (e) A message specified by the cannot_route_message router option.
4097 In these cases only, Exim does include the error details in bounce and
4098 warning messages. There are also a few cases where bland messages such
4099 as "unrouteable address" or "local delivery error" are given.
4101 PH/43 $value is now also set for the "else" part of a ${run expansion.
4103 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4104 being worked on, but at least Exim now implements the latest version to
4107 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4108 process to complete was treated as a delivery failure, a timeout while
4109 writing the message to the pipe was logged, but erroneously treated as a
4110 successful delivery. Such timeouts include transport filter timeouts. For
4111 consistency with the overall process timeout, these timeouts are now
4112 treated as errors, giving rise to delivery failures by default. However,
4113 there is now a new Boolean option for the pipe transport called
4114 timeout_defer, which, if set TRUE, converts the failures into defers for
4115 both kinds of timeout. A transport filter timeout is now identified in
4118 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4119 systems where "make" and "gmake" are different, calling "gmake" at top
4120 level broke things. I've arranged for the value of $(MAKE) to be passed
4121 from the Makefile to this script so that it can call the same version of
4125 A note about Exim versions 4.44 and 4.50
4126 ----------------------------------------
4128 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4129 changes of various kinds. As a consequence, a big documentation update was
4130 needed. This delayed the release for rather longer than seemed good, especially
4131 in the light of a couple of (minor) security issues. Therefore, the changes
4132 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4133 release. So 4.44 and 4.50 are in effect two different branches that both start
4136 I have left the 4.50 change log unchanged; it contains all the changes since
4137 4.43. The change log for 4.44 is below; many of its items are identical to
4138 those for 4.50. This seems to be the most sensible way to preserve the
4139 historical information.
4145 1. Minor wording change to the doc/README.SIEVE file.
4147 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4148 computation of the current number of files was incorrect.
4150 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4151 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4154 4. Give more explanation in the error message when the command for a transport
4155 filter fails to execute.
4157 5. There are several places where Exim runs a non-Exim command in a
4158 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4159 was being done only for the command run by the queryprogram router. It is
4160 now done for all such subprocesses. The other cases are: ${run, transport
4161 filters, and the commands run by the lmtp and pipe transports.
4163 6. Added CONFIGURE_GROUP build-time option.
4165 7. Some older OS have a limit of 256 on the maximum number of file
4166 descriptors. Exim was using setrlimit() to set 1000 as a large value
4167 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4168 systems. I've change it so that if it can't get 1000, it tries for 256.
4170 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4171 was an oversight, and furthermore, ever since the addition of extra
4172 controls (e.g. 4.43/32), the checks on when to allow different forms of
4173 "control" were broken. There should now be diagnostics for all cases when a
4174 control that does not make sense is encountered.
4176 9. Added the /retain_sender option to "control=submission".
4178 10. $recipients is now available in the predata ACL (oversight).
4180 11. Tidy the search cache before the fork to do a delivery from a message
4181 received from the command line. Otherwise the child will trigger a lookup
4182 failure and thereby defer the delivery if it tries to use (for example) a
4183 cached ldap connection that the parent has called unbind on.
4185 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4186 of $address_data from the recipient verification was clobbered by the
4187 sender verification.
4189 13. The value of address_data from a sender verification is now available in
4190 $sender_address_data in subsequent conditions in the ACL statement.
4192 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4194 15. Added a new option "connect=<time>" to callout options, to set a different
4197 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4198 was its contents. (It was OK if the option was not defined at all.)
4200 17. A "Completed" log line is now written for messages that are removed from
4201 the spool by the -Mrm option.
4203 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4204 information about exactly what failed.
4206 19. Added -dd to debug only the daemon process.
4208 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4209 handles timeouts, both on the server side and network timeouts. Renamed the
4210 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4212 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4213 It is now set to "smtps".
4215 22. $host_address is now set to the target address during the checking of
4216 ignore_target_hosts.
4218 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4219 being passed; this would have caused $sender_host_name to have been used if
4220 matching the list had actually called for a host name (not very likely,
4221 since this list is usually IP addresses). A host name is now passed as
4224 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4225 code that specifies a non-restarting handler (typically sigaction() in
4226 modern systems) in an attempt to fix a rare and obscure crash bug.
4228 25. Narrowed the window for a race in the daemon that could cause it to ignore
4229 SIGCHLD signals. This is not a major problem, because they are used only to
4230 wake it up if nothing else does.
4232 26. A malformed maildirsize file could cause Exim to calculate negative values
4233 for the mailbox size or file count. Odd effects could occur as a result.
4234 The maildirsize information is now recalculated if the size or filecount
4237 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4238 support for a long time. Removed HAVE_SYS_VFS_H.
4240 28. Installed the latest version of exipick from John Jetmore.
4242 29. In an address list, if the pattern was not a regular expression, an empty
4243 subject address (from a bounce message) matched only if the pattern was an
4244 empty string. Non-empty patterns were not even tested. This was the wrong
4245 because it is perfectly reasonable to use an empty address as part of a
4246 database query. An empty address is now tested by patterns that are
4247 lookups. However, all the other forms of pattern expect the subject to
4248 contain a local part and a domain, and therefore, for them, an empty
4249 address still always fails if the pattern is not itself empty.
4251 30. Exim went into a mad DNS loop when attempting to do a callout where the
4252 host was specified on an smtp transport, and looking it up yielded more
4253 than one IP address.
4255 31. Re-factored the code for checking spool and log partition space into a
4256 function that finds that data and another that does the check. The former
4257 is then used to implement four new variables: $spool_space, $log_space,
4258 $spool_inodes, and $log_inodes.
4260 32. The RFC2047 encoding function was originally intended for short strings
4261 such as real names; it was not keeping to the 75-character limit for
4262 encoded words that the RFC imposes. It now respects the limit, and
4263 generates multiple encoded words if necessary. To be on the safe side, I
4264 have increased the buffer size for the ${rfc2047: expansion operator from
4267 33. It is now permitted to omit both strings after an "if" condition; if the
4268 condition is true, the result is "true". As before, when the second string
4269 is omitted, a false condition yields an empty string. This makes it less
4270 cumbersome to write custom ACL and router conditions.
4272 34. Failure to deliver a bounce message always caused it to be frozen, even if
4273 there was an errors_to setting on the router. The errors_to setting is now
4276 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4277 canonical form (fully expanded) before being placed in
4278 $sender_host_address.
4280 36. The table in the code that translates DNS record types into text (T_A to
4281 "A" for instance) was missing entries for NS and CNAME. It is just possible
4282 that this could have caused confusion if both these types were looked up
4283 for the same domain, because the text type is used as part of Exim's
4284 per-process caching. But the chance of anyone hitting this buglet seems
4287 37. The dnsdb lookup has been extended in a number of ways.
4289 (1) There is a new type, "zns", which walks up the domain tree until it
4290 finds some nameserver records. It should be used with care.
4292 (2) There is a new type, "mxh", which is like "mx" except that it returns
4293 just the host names, not the priorities.
4295 (3) It is now possible to give a list of domains (or IP addresses) to be
4296 looked up. The behaviour when one of the lookups defers can be
4297 controlled by a keyword.
4299 (4) It is now possible to specify the separator character for use when
4300 multiple records are returned.
4302 38. The dnslists ACL condition has been extended: it is now possible to supply
4303 a list of IP addresses and/or domains to be looked up in a particular DNS
4306 39. Added log_selector=+queue_time_overall.
4308 40. When running the queue in the test harness, wait just a tad after forking a
4309 delivery process, to get repeatability of debugging output.
4311 41. Include certificate and key file names in error message when GnuTLS fails
4312 to set them up, because the GnuTLS error message doesn't include the name
4313 of the failing file when there is a problem reading it.
4315 42. Allow both -bf and -bF in the same test run.
4317 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4319 44. The "Exiscan patch" is now merged into the mainline Exim source.
4321 45. Sometimes the final signoff response after QUIT could fail to get
4322 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4323 < 0 before doing a fflush(). This bug looks as though it goes back to the
4324 introduction of TLS in release 3.20, but "sometimes" must have been rare
4325 because the tests only now provoked it.
4327 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4328 (this can affect the format of dates).
4330 47. exim_tidydb, when checking for the continued existence of a message for
4331 which it has found a message-specific retry record, was not finding
4332 messages that were in split spool directories. Consequently, it was
4333 deleting retry records that should have stayed in existence.
4335 48. Steve fixed some bugs in eximstats.
4337 49. The SPA authentication driver was not abandoning authentication and moving
4338 on to the next authenticator when an expansion was forced to fail,
4339 contradicting the general specification for all authenticators. Instead it
4340 was generating a temporary error. It now behaves as specified.
4342 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4343 (the order specifies the preference for clients). The order is now AES256,
4344 AES128, 3DES, ARCFOUR128.
4346 51. Small patch to Sieve code - explicitly set From: when generating an
4349 52. Exim crashed if a remote delivery caused a very long error message to be
4350 recorded - for instance if somebody sent an entire SpamAssassin report back
4351 as a large number of 550 error lines. This bug was coincidentally fixed by
4352 increasing the size of one of Exim's internal buffers (big_buffer) that
4353 happened as part of the Exiscan merge. However, to be on the safe side, I
4354 have made the code more robust (and fixed the comments that describe what
4357 53. Now that there can be additional text after "Completed" in log lines (if
4358 the queue_time_overall log selector is set), a one-byte patch to exigrep
4359 was needed to allow it to recognize "Completed" as not the last thing in
4362 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4363 patch that reportedly fixes this has been added. I am not expert enough to
4364 create a test for it. This is what the patch creator wrote:
4366 "I found a little strange behaviour of ldap code when working with
4367 Windows 2003 AD Domain, where users was placed in more than one
4368 Organization Units. When I tried to give exim partial DN, the exit code
4369 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4370 But simultaneously result of request was absolutely normal ldap result,
4371 so I produce this patch..."
4373 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4374 REFERENCE, so I have modified the code to exclude the patch when that macro
4377 55. Some experimental protocols are using DNS PTR records for new purposes. The
4378 keys for these records are domain names, not reversed IP addresses. The
4379 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4380 leaves it alone. Component reversal etc. now happens only for IP addresses.
4383 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4385 57. Double the size of the debug message buffer (to 2048) so that more of very
4386 long debug lines gets shown.
4388 58. The exicyclog utility now does better if the number of log files to keep
4389 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4391 59. Two changes related to the smtp_active_hostname option:
4393 (1) $smtp_active_hostname is now available as a variable.
4394 (2) The default for smtp_banner uses $smtp_active_hostname instead
4395 of $primary_hostname.
4397 60. The host_aton() function is supposed to be passed a string that is known
4398 to be a valid IP address. However, in the case of IPv6 addresses, it was
4399 not checking this. This is a hostage to fortune. Exim now panics and dies
4400 if the condition is not met. A case was found where this could be provoked
4401 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4402 components; fortuitously, this particular loophole had already been fixed
4403 by change 4.50/55 above.
4405 If there are any other similar loopholes, the new check in host_aton()
4406 itself should stop them being exploited. The report I received stated that
4407 data on the command line could provoke the exploit when Exim was running as
4408 exim, but did not say which command line option was involved. All I could
4409 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4410 running as the user.
4413 61. There was a buffer overflow vulnerability in the SPA authentication code
4414 (which came originally from the Samba project). I have added a test to the
4415 spa_base64_to_bits() function which I hope fixes it.
4418 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4419 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4421 63. The daemon start-up calls getloadavg() while still root for those OS that
4422 need the first call to be done as root, but it missed one case: when
4423 deliver_queue_load_max is set with deliver_drop_privilege. This is
4424 necessary for the benefit of the queue runner, because there is no re-exec
4425 when deliver_drop_privilege is set.
4427 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4428 This has been fixed.
4430 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4431 was in use, was not putting the data itself into the right store pool;
4432 consequently, it could be overwritten for a subsequent message in the same
4433 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4436 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4438 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4439 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4440 the function were treating the return as a boolean value, which happened to
4441 work because 0=false and not-0=true, but is not correct code.
4443 68. The host_aton() function was not handling scoped IPv6 addresses (those
4444 with, for example, "%eth0" on the end) correctly.
4446 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4447 negated items (that is, ~something) in unsigned ints. Some compilers
4448 apparently mutter when there is no cast.
4450 70. If an address verification called from an ACL failed, and did not produce a
4451 user-specific message (i.e. there was only a "system" message), nothing was
4452 put in $acl_verify_message. In this situation, it now puts the system
4455 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4456 session; change 4.31/43 added the unwanted input to the log line - except
4457 that it did not do this in the start of session case. It now does.
4459 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4460 This is wrong and can cause the other end to generate a synchronization
4461 error if it is another Exim or anything else that does the synchronization
4462 check. A QUIT command is no longer sent after a timeout.
4464 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4465 during host lookups.
4467 74. The defer_ok option of callout verification was not working if it was used
4468 when verifying addresses in header lines, that is, for this case:
4470 verify = header_sender/callout=defer_ok
4472 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4473 those file descriptors could be used for SMTP connections. If anything
4474 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4475 could be sent to the SMTP client, causing chaos. The daemon now opens
4476 stdin, stdout, and stderr to /dev/null when it puts itself into the
4479 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4480 log by default. The user can override this with suitable Perl magic.
4482 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4483 the log message when discard triggers, was not working for the DATA ACL or
4484 for the non-SMTP ACL.
4486 78. Error message wording change in sieve.c.
4488 79. If smtp_accept_max_per_host was set, the number of connections could be
4489 restricted to fewer than expected, because the daemon was trying to set up
4490 a new connection before checking whether the processes handling previous
4491 connections had finished. The check for completed processes is now done
4492 earlier. On busy systems, this bug wouldn't be noticed because something
4493 else would have woken the daemon, and it would have reaped the completed
4496 80. If a message was submitted locally by a user whose login name contained one
4497 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4498 It caused a spool format error. I have fixed the spool reading code. A
4499 related problem was that the "from" clause in the Received: line became
4500 illegal because of the space(s). It is now covered by ${quote_local_part.
4502 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4505 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4507 83. Patch from Sieve maintainer to fix unterminated string problem in
4508 "vacation" handling.
4510 84. Some minor changes to the Linux configuration files to help with other
4511 OS variants using glibc.
4513 85. One more patch for Sieve to update vacation handling to latest spec.
4516 ----------------------------------------------------
4517 See the note above about the 4.44 and 4.50 releases.
4518 ----------------------------------------------------
4524 1. Change 4.43/35 introduced a bug that caused file counts to be
4525 incorrectly computed when quota_filecount was set in an appendfile
4528 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4529 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4532 3. Give more explanation in the error message when the command for a transport
4533 filter fails to execute.
4535 4. There are several places where Exim runs a non-Exim command in a
4536 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4537 was being done only for the command run by the queryprogram router. It is
4538 now done for all such subprocesses. The other cases are: ${run, transport
4539 filters, and the commands run by the lmtp and pipe transports.
4541 5. Some older OS have a limit of 256 on the maximum number of file
4542 descriptors. Exim was using setrlimit() to set 1000 as a large value
4543 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4544 systems. I've change it so that if it can't get 1000, it tries for 256.
4546 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4547 was an oversight, and furthermore, ever since the addition of extra
4548 controls (e.g. 4.43/32), the checks on when to allow different forms of
4549 "control" were broken. There should now be diagnostics for all cases when a
4550 control that does not make sense is encountered.
4552 7. $recipients is now available in the predata ACL (oversight).
4554 8. Tidy the search cache before the fork to do a delivery from a message
4555 received from the command line. Otherwise the child will trigger a lookup
4556 failure and thereby defer the delivery if it tries to use (for example) a
4557 cached ldap connection that the parent has called unbind on.
4559 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4560 of $address_data from the recipient verification was clobbered by the
4561 sender verification.
4563 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4564 was its contents. (It was OK if the option was not defined at all.)
4566 11. A "Completed" log line is now written for messages that are removed from
4567 the spool by the -Mrm option.
4569 12. $host_address is now set to the target address during the checking of
4570 ignore_target_hosts.
4572 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4573 being passed; this would have caused $sender_host_name to have been used if
4574 matching the list had actually called for a host name (not very likely,
4575 since this list is usually IP addresses). A host name is now passed as
4578 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4579 code that specifies a non-restarting handler (typically sigaction() in
4580 modern systems) in an attempt to fix a rare and obscure crash bug.
4582 15. Narrowed the window for a race in the daemon that could cause it to ignore
4583 SIGCHLD signals. This is not a major problem, because they are used only to
4584 wake it up if nothing else does.
4586 16. A malformed maildirsize file could cause Exim to calculate negative values
4587 for the mailbox size or file count. Odd effects could occur as a result.
4588 The maildirsize information is now recalculated if the size or filecount
4591 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4592 support for a long time. Removed HAVE_SYS_VFS_H.
4594 18. Updated exipick to current release from John Jetmore.
4596 19. Allow an empty sender to be matched against a lookup in an address list.
4597 Previously the only cases considered were a regular expression, or an
4600 20. Exim went into a mad DNS lookup loop when doing a callout where the
4601 host was specified on the transport, if the DNS lookup yielded more than
4604 21. The RFC2047 encoding function was originally intended for short strings
4605 such as real names; it was not keeping to the 75-character limit for
4606 encoded words that the RFC imposes. It now respects the limit, and
4607 generates multiple encoded words if necessary. To be on the safe side, I
4608 have increased the buffer size for the ${rfc2047: expansion operator from
4611 22. Failure to deliver a bounce message always caused it to be frozen, even if
4612 there was an errors_to setting on the router. The errors_to setting is now
4615 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4616 canonical form (fully expanded) before being placed in
4617 $sender_host_address.
4619 24. Updated eximstats to version 1.33
4621 25. Include certificate and key file names in error message when GnuTLS fails
4622 to set them up, because the GnuTLS error message doesn't include the name
4623 of the failing file when there is a problem reading it.
4625 26. Expand error message when OpenSSL has problems setting up cert/key files.
4628 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4629 (this can affect the format of dates).
4631 28. exim_tidydb, when checking for the continued existence of a message for
4632 which it has found a message-specific retry record, was not finding
4633 messages that were in split spool directories. Consequently, it was
4634 deleting retry records that should have stayed in existence.
4636 29. eximstats updated to version 1.35
4637 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4638 1.35 - bugfix such that pie charts by volume are generated correctly
4640 30. The SPA authentication driver was not abandoning authentication and moving
4641 on to the next authenticator when an expansion was forced to fail,
4642 contradicting the general specification for all authenticators. Instead it
4643 was generating a temporary error. It now behaves as specified.
4645 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4646 (the order specifies the preference for clients). The order is now AES256,
4647 AES128, 3DES, ARCFOUR128.
4649 31. Small patch to Sieve code - explicitly set From: when generating an
4652 32. Exim crashed if a remote delivery caused a very long error message to be
4653 recorded - for instance if somebody sent an entire SpamAssassin report back
4654 as a large number of 550 error lines. This bug was coincidentally fixed by
4655 increasing the size of one of Exim's internal buffers (big_buffer) that
4656 happened as part of the Exiscan merge. However, to be on the safe side, I
4657 have made the code more robust (and fixed the comments that describe what
4660 33. Some experimental protocols are using DNS PTR records for new purposes. The
4661 keys for these records are domain names, not reversed IP addresses. The
4662 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4663 leaves it alone. Component reversal etc. now happens only for IP addresses.
4666 34. The host_aton() function is supposed to be passed a string that is known
4667 to be a valid IP address. However, in the case of IPv6 addresses, it was
4668 not checking this. This is a hostage to fortune. Exim now panics and dies
4669 if the condition is not met. A case was found where this could be provoked
4670 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4671 components; fortuitously, this particular loophole had already been fixed
4672 by change 4.50/55 or 4.44/33 above.
4674 If there are any other similar loopholes, the new check in host_aton()
4675 itself should stop them being exploited. The report I received stated that
4676 data on the command line could provoke the exploit when Exim was running as
4677 exim, but did not say which command line option was involved. All I could
4678 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4679 running as the user.
4682 35. There was a buffer overflow vulnerability in the SPA authentication code
4683 (which came originally from the Samba project). I have added a test to the
4684 spa_base64_to_bits() function which I hope fixes it.
4687 36. The daemon start-up calls getloadavg() while still root for those OS that
4688 need the first call to be done as root, but it missed one case: when
4689 deliver_queue_load_max is set with deliver_drop_privilege. This is
4690 necessary for the benefit of the queue runner, because there is no re-exec
4691 when deliver_drop_privilege is set.
4693 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4694 was in use, was not putting the data itself into the right store pool;
4695 consequently, it could be overwritten for a subsequent message in the same
4696 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4699 38. Sometimes the final signoff response after QUIT could fail to get
4700 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4701 < 0 before doing a fflush(). This bug looks as though it goes back to the
4702 introduction of TLS in release 3.20, but "sometimes" must have been rare
4703 because the tests only now provoked it.
4709 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4710 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4711 Now it returns an integer. A number of calls were still expecting a T/F
4712 return. Fortuitously, in all cases, the tests worked in OK situations,
4713 which is the norm. However, things would have gone wrong on any write
4714 failures on the smtp file descriptor. This function is used when sending
4715 messages over SMTP and also when doing verify callouts.
4717 2. When Exim is called to do synchronous delivery of a locally submitted
4718 message (the -odf or -odi options), it no longer closes stderr before doing
4721 3. Implemented the mua_wrapper option.
4723 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4725 5. Implemented the functions header_remove(), header_testname(),
4726 header_add_at_position(), and receive_remove_recipient(), and exported them
4729 6. If an ACL "warn" statement specified the addition of headers, Exim already
4730 inserted X-ACL-Warn: at the start if there was no header name. However, it
4731 was not making this test for the second and subsequent header lines if
4732 there were newlines in the string. This meant that an invalid header could
4733 be inserted if Exim was badly configured.
4735 7. Allow an ACL "warn" statement to add header lines at the start or after all
4736 the Received: headers, as well as at the end.
4738 8. Added the rcpt_4xx retry error code.
4740 9. Added postmaster_mailfrom=xxx to callout verification option.
4742 10. Added mailfrom=xxxx to the callout verification option, for verify=
4745 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4746 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4748 12. Inserted some casts to stop certain compilers warning when using pointer
4749 differences as field lengths or precisions in printf-type calls (mostly
4750 affecting debugging statements).
4752 13. Added optional readline() support for -be (dynamically loaded).
4754 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4755 same clock tick as a message's arrival, so that its received time was the
4756 same as the "first fail" time on the retry record, and that message
4757 remained on the queue past the ultimate address timeout, every queue runner
4758 would try a delivery (because it was past the ultimate address timeout) but
4759 after another failure, the ultimate address timeout, which should have then
4760 bounced the address, did not kick in. This was a "< instead of <=" error;
4761 in most cases the first failure would have been in the next clock tick
4762 after the received time, and all would be well.
4764 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4765 being recognized when the domain list was tested by the match_domain
4766 condition in an expansion string.
4768 16. Added the ${str2b64: operator.
4770 17. Exim was always calling setrlimit() to set a large limit for the number of
4771 processes, without checking whether the existing limit was already
4772 adequate. (It did check for the limit on file descriptors.) Furthermore,
4773 errors from getrlimit() and setrlimit() were being ignored. Now they are
4774 logged to the main and panic logs, but Exim does carry on, to try to do its
4775 job under whatever limits there are.
4777 18. Imported PCRE 5.0.
4779 19. Trivial typo in log message " temporarily refused connection" (the leading
4782 20. If the log selector return_path_on_delivery was set and an address was
4783 redirected to /dev/null, the delivery process crashed because it assumed
4784 that a return path would always be set for a "successful" delivery. In this
4785 case, the whole delivery is bypassed as an optimization, and therefore no
4788 21. Internal re-arrangement: the function for sending a challenge and reading
4789 a response while authentication was assuming a zero-terminated challenge
4790 string. It's now changed to take a pointer and a length, to allow for
4791 binary data in such strings.
4793 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4795 23. Exim was not respecting finduser_retries when seeking the login of the
4796 uid under which it was called; it was always trying 10 times. (The default
4797 setting of finduser_retries is zero.) Also, it was sleeping after the final
4798 failure, which is pointless.
4800 24. Implemented tls_on_connect_ports.
4802 25. Implemented acl_smtp_predata.
4804 26. If the domain in control=submission is set empty, Exim assumes that the
4805 authenticated id is a complete email address when it generates From: or
4806 Sender: header lines.
4808 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4809 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4810 chown and chgrp in /bin and hostname in /usr/bin.
4812 28. Exim was keeping the "process log" file open after each use, just as it
4813 does for the main log. This opens the possibility of it remaining open for
4814 long periods when the USR1 signal hits a daemon. Occasional processlog
4815 errors were reported, that could have been caused by this. Anyway, it seems
4816 much more sensible not to leave this file open at all, so that is what now
4819 29. The long-running daemon process does not normally write to the log once it
4820 has entered its main loop, and it closes the log before doing so. This is
4821 so that log files can straightforwardly be renamed and moved. However,
4822 there are a couple of unusual error situations where the daemon does write
4823 log entries, and I had neglected to close the log afterwards.
4825 30. The text of an SMTP error response that was received during a remote
4826 delivery was being truncated at 512 bytes. This is too short for some of
4827 the long messages that one sometimes sees. I've increased the limit to
4830 31. It is now possible to make retry rules that apply only when a message has a
4831 specific sender, in particular, an empty sender.
4833 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4834 it possible to be selective about when SMTP synchronization is enforced.
4836 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4838 32. Implemented hosts_connection_nolog.
4840 33. Added an ACL for QUIT.
4842 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4845 35. Added mailbox_size and mailbox_filecount to appendfile.
4847 36. Added control = no_multiline_responses to ACLs.
4849 37. There was a bug in the logic of the code that waits for the clock to tick
4850 in the case where the clock went backwards by a substantial amount such
4851 that the microsecond fraction of "now" was more than the microsecond
4852 fraction of "then" (but the whole seconds number was less).
4854 38. Added support for the libradius Radius client library this is found on
4855 FreeBSD (previously only the radiusclient library was supported).
4861 1. When certain lookups returned multiple values in the form name=value, the
4862 quoting of the values was not always being done properly. Specifically:
4863 (a) If the value started with a double quote, but contained no whitespace,
4865 (b) If the value contained whitespace other than a space character (i.e.
4866 tabs or newlines or carriage returns) it was not quoted.
4867 This fix has been applied to the mysql and pgsql lookups by writing a
4868 separate quoting function and calling it from the lookup code. The fix
4869 should probably also be applied to nisplus, ibase and oracle lookups, but
4870 since I cannot test any of those, I have not disturbed their existing code.
4872 2. A hit in the callout cache for a specific address caused a log line with no
4873 reason for rejecting RCPT. Now it says "Previous (cached) callout
4874 verification failure".
4876 3. There was an off-by-one bug in the queryprogram router. An over-long
4877 return line was truncated at 256 instead of 255 characters, thereby
4878 overflowing its buffer with the terminating zero. As well as fixing this, I
4879 have increased the buffer size to 1024 (and made a note to document this).
4881 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4882 when Exim is waiting for an SMTP response from a remote server, Exim
4883 restarts its select() call on the socket, thereby resetting its timeout.
4884 This is not a problem when such interrupts are rare. Somebody set up a cron
4885 job to run exiwhat every 2 minutes, which is less than the normal select()
4886 timeout (5 or 10 minutes). This meant that the select() timeout never
4887 kicked in because it was always reset. I have fixed this by comparing the
4888 time when an interrupt arrives with the time at the start of the first call
4889 to select(). If more time than the timeout has elapsed, the interrupt is
4890 treated as a timeout.
4892 5. Some internal re-factoring in preparation for the addition of Sieve
4893 extensions (by MH). In particular, the "personal" test is moved to a
4894 separate function, and given an option for scanning Cc: and Bcc: (which is
4895 not set for Exim filters).
4897 6. When Exim created an email address using the login of the caller as the
4898 local part (e.g. when creating a From: or Sender: header line), it was not
4899 quoting the local part when it contained special characters such as @.
4901 7. Installed new OpenBSD configuration files.
4903 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4904 try to make them clearer.
4906 9. Callout options, other than the timeout value, were being ignored when
4907 verifying sender addresses in header lines. For example, when using
4909 verify = header_sender/callout=no_cache
4911 the cache was (incorrectly) being used.
4913 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4914 only the Cygwin environment.
4916 11. When return_path_on_delivery was set as a log selector, if different remote
4917 addresses in the same message used different return paths and parallel
4918 remote delivery occurred, the wrong values would sometimes be logged.
4919 (Whenever a remote delivery process finished, the return path value from
4920 the most recently started remote delivery process was logged.)
4922 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4923 lines when AUTH and/or TLS are in use. This is the "received protocol"
4924 field. Exim used to use "asmtp" for authenticated SMTP, without any
4925 indication (in the protocol name) for TLS use. Now it follows the RFC and
4926 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4927 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4928 names appear in log lines as well as in Received: header lines.
4930 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4931 extensions, and comparison tests, and to fix some bugs.
4933 14. Changes to the "personal" filter test:
4935 (1) The test was buggy in that it was just doing the equivalent of
4936 "contains" tests on header lines. For example, if a user's address was
4937 anne@some.where, the "personal" test would incorrectly be true for
4939 To: susanne@some.where
4941 This test is now done by extracting each address from the header in turn,
4942 and checking the entire address. Other tests that are part of "personal"
4943 are now done using regular expressions (for example, to check local parts
4944 of addresses in From: header lines).
4946 (2) The list of non-personal local parts in From: addresses has been
4947 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4948 taken from the Sieve specification recommendations.
4950 (3) If the message contains any header line starting with "List-" it is
4951 treated as non-personal.
4953 (4) The test for "circular" in the Subject: header line has been removed
4954 because it now seems ill-conceived.
4956 15. Minor typos in src/EDITME comments corrected.
4958 16. Installed latest exipick from John Jetmore.
4960 17. If headers_add on a router specified a text string that was too long for
4961 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4962 of string_sprintf() is now avoided.
4964 18. $message_body_size was not set (it was always zero) when running the DATA
4965 ACL and the local_scan() function.
4967 19. For the "mail" command in an Exim filter, no default was being set for
4968 the once_repeat time, causing a random time value to be used if "once" was
4969 specified. (If the value happened to be <= 0, no repeat happened.) The
4970 default is now 0s, meaning "never repeat". The "vacation" command was OK
4971 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4972 (I found it when inspecting the code).
4974 20. There is now an overall timeout for performing a callout verification. It
4975 defaults to 4 times the callout timeout, which applies to individual SMTP
4976 commands during the callout. The overall timeout applies when there is more
4977 than one host that can be tried. The timeout is checked before trying the
4978 next host. This prevents very long delays if there are a large number of
4979 hosts and all are timing out (e.g. when the network connections are timing
4980 out). The value of the overall timeout can be changed by specifying an
4981 additional sub-option for "callout", called "maxwait". For example:
4983 verify = sender/callout=5s,maxwait=20s
4985 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4986 to the end before writing, but this should make it even safer).
4988 22. Exim was forgetting that it had advertised PIPELINING for the second and
4989 subsequent messages on an SMTP connection. It was also not resetting its
4990 memory on STARTTLS and an internal HELO.
4992 23. When Exim logs an SMTP synchronization error within a session, it now
4993 records whether PIPELINING has been advertised or not.
4995 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4996 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4997 rather than long int.
4999 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5001 26. Added the never_mail option to autoreply.
5007 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5008 crash if the getsockname() call failed; this can happen if a connection is
5009 closed very soon after it is established. The problem was simply in the
5010 order in which certain operations were done, causing Exim to try to write
5011 to the SMTP stream before it had set up the file descriptor. The bug has
5012 been fixed by making things happen in the correct order.
5018 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5019 before the connection was closed, thus losing the rejection response.
5021 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5022 some early Solaris releases, but causes trouble in current releases where
5023 socklen_t is defined.
5025 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5028 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5031 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5032 that releasing the top of it at the end releases what was used for sub-
5033 expansions (unless the block got too big). However, discard this block if
5034 the first thing is a variable or header, so that we can use its block when
5035 it is dynamic (useful for very large $message_headers, for example).
5037 6. Lookups now cache *every* query, not just the most recent. A new, separate
5038 store pool is used for this. It can be recovered when all lookup caches are
5039 flushed. Lookups now release memory at the end of their result strings.
5040 This has involved some general refactoring of the lookup sources.
5042 7. Some code has been added to the store_xxx() functions to reduce the amount
5043 of flapping under certain conditions.
5045 8. log_incoming_interface used to affect only the <= reception log lines. Now
5046 it causes the local interface and port to be added to several more SMTP log
5047 lines, for example "SMTP connection from", and rejection lines.
5049 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5051 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5053 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5054 could be overwritten at the end of the current message (or the start of a
5055 new message if it was set in a HELO ACL). The value is now preserved for
5056 the duration of the SMTP connection.
5058 12. If a transport had a headers_rewrite setting, and a matching header line
5059 contained an unqualified address, that address was qualified, even if it
5060 did not match any rewriting rules. The underlying bug was that the values
5061 of the flags that permit the existence of unqualified sender and recipient
5062 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5063 non-local messages, and by -bnq for local messages) were not being
5064 preserved with the message after it was received.
5066 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5067 "next input=" as part of the text comprising the host identity instead of
5068 the correct text. The code was using the same buffer for two different
5069 strings. However, depending on which order the printing function evaluated
5070 its arguments, the bug did not always show up. Under Linux, for example, my
5071 test suite worked just fine.
5073 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5074 doesn't work with some older versions of Perl. It has been changed to "my",
5075 which in any case is probably the better facility to use.
5077 15. A really picky compiler found some instances of statements for creating
5078 error messages that either had too many or two few arguments for the format
5081 16. The size of the buffer for calls to the DNS resolver has been increased
5082 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5083 for addresses that have a lot of PTR records. This alleviates a problem; it
5084 does not fully solve it.
5086 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5087 buffer now truncates the list and logs the incident, which is the same
5088 action as happens when Exim is looking up a host name and its aliases.
5089 Previously in this situation something unpredictable would happen;
5090 sometimes it was "internal error: store_reset failed".
5092 18. If a server dropped the connection unexpectedly when an Exim client was
5093 using GnuTLS and trying to read a response, the client delivery process
5094 crashed while trying to generate an error log message.
5096 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5097 string, for example:
5099 warn message = H1: something\nH2: something
5101 the text was added as a single header line from Exim's point of view
5102 though it ended up OK in the delivered message. However, searching for the
5103 second and subsequent header lines using $h_h2: did not work. This has been
5104 fixed. Similarly, if a system filter added multiple headers in this way,
5105 the routers could not see them.
5107 20. Expanded the error message when iplsearch is called with an invalid key to
5108 suggest using net-iplsearch in a host list.
5110 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5111 ACLs are no longer actually imposed (and a message to that effect is
5114 22. If a "gecos" field in a passwd entry contained escaped characters, in
5115 particular, if it contained a \" sequence, Exim got it wrong when building
5116 a From: or a Sender: header from that name. A second bug also caused
5117 incorrect handling when an unquoted " was present following a character
5118 that needed quoting.
5120 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5121 was not being matched caselessly.
5123 24. Arranged for all hyphens in the exim.8 source to be escaped with
5126 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5127 properly. Recipient callout cache records were still being keyed to include
5128 the sender, even when use_sender was set false. This led to far more
5129 callouts that were necessary. The sender is no longer included in the key
5130 when use_sender is false.
5132 26. Added "control = submission" modifier to ACLs.
5134 27. Added the ${base62d: operator to decode base 62 numbers.
5136 28. dnsdb lookups can now access SRV records.
5138 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5139 the configuration file.
5141 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5142 (-v) mode. This makes the output for a verbose queue run more intelligible.
5144 31. Added a use_postmaster feature to recipient callouts.
5146 32. Added the $body_zerocount variable, containing the number of binary zero
5147 bytes in the message body.
5149 33. The time of last modification of the "new" subdirectory is now used as the
5150 "mailbox time last read" when there is a quota error for a maildir
5153 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5155 35. Added +ignore_unknown as a special item in host lists.
5157 36. Code for decoding IPv6 addresses in host lists is now included, even if
5158 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5159 address was recognized as an IP address, but was then not correctly decoded
5160 into binary, causing unexpected and incorrect effects when compared with
5167 1. Very minor rewording of debugging text in manualroute to say "list of
5168 hosts" instead of "hostlist".
5170 2. If verify=header_syntax was set, and a header line with an unqualified
5171 address (no domain) and a large number of spaces between the end of the
5172 name and the colon was received, the reception process suffered a buffer
5173 overflow, and (when I tested it) crashed. This was caused by some obsolete
5174 code that should have been removed. The fix is to remove it!
5176 3. When running in the test harness, delay a bit after writing a bounce
5177 message to get a bit more predictability in the log output.
5179 4. Added a call to search_tidyup() just before forking a reception process. In
5180 theory, someone could use a lookup in the expansion of smtp_accept_max_
5181 per_host which, without the tidyup, could leave open a database connection.
5183 5. Added the variables $recipient_data and $sender_data which get set from a
5184 lookup success in an ACL "recipients" or "senders" condition, or a router
5185 "senders" option, similar to $domain_data and $local_part_data.
5187 6. Moved the writing of debug_print from before to after the "senders" test
5190 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5191 problems for message scanning, either using a data ACL, or using
5192 local_scan() because the Received: header was not generated till after they
5193 were called (in order to set the time as the time of reception completion).
5194 I have revised the way this works. The header is now generated after the
5195 body is received, but before the ACL or local_scan() are called. After they
5196 are run, the timestamp in the header is updated.
5202 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5203 before starting a queue runner without re-exec. This happened only when
5204 deliver_drop_privilege was set or when the Exim user was set to root. The
5205 effect of the bug was that timeouts during subsequent deliveries caused
5206 crashes instead of being properly handled. The handler is now left at its
5207 default (and expected) setting.
5209 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5210 message, again when deliver_drop_privilege is set or Exim is run as root.
5211 The bug described in (1) was not present in this case, but the tidying up
5212 of the other signals was missing. I have made the two cases consistent.
5214 3. The ignore_target_hosts setting on a manualroute router was being ignored
5215 for hosts that were looked up using the /MX notation.
5217 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5220 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5221 operated on the sender address. After changing the $sender_address to <>
5222 for the sender address verify, Exim was re-instated it as the original
5223 (before rewriting) address, but remembering that it had rewritten it, so it
5224 wasn't rewriting it again. This bug also had the effect of breaking the
5225 sender address verification caching when the sender address was rewritten.
5227 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5228 This has been changed so that if the ip literal address matches
5229 ignore_target_hosts, the router declines.
5231 7. Added expansion conditions match_domain, match_address, and match_local_
5232 part (NOT match_host).
5234 8. The placeholder for the Received: header didn't have a length field set.
5236 9. Added code to Exim itself and to exim_lock to test for a specific race
5237 condition that could lead to file corruption when using MBX delivery. The
5238 issue is with the lockfile that is created in /tmp. If this file is removed
5239 after a process has opened it but before that process has acquired a lock,
5240 there is the potential for a second process to recreate the file and also
5241 acquire a lock. This could lead to two Exim processes writing to the file
5242 at the same time. The added code performs the same test as UW imapd; it
5243 checks after acquiring the lock that its file descriptor still refers to
5244 the same named file.
5246 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5247 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5250 11. Added the smtp_active_hostname option. If used, this will typically be made
5251 to depend on the incoming interface address. Because $interface_address is
5252 not set up until the daemon has forked a reception process, error responses
5253 that can happen earlier (such as "too many connections") no longer contain
5256 12. If an expansion in a condition on a "warn" statement fails because a lookup
5257 defers, the "warn" statement is abandoned, and the next ACL statement is
5258 processed. Previously this caused the whole ACL to be aborted.
5260 13. Added the iplsearch lookup type.
5262 14. Added ident_timeout as a log selector.
5264 15. Added tls_certificate_verified as a log selector.
5266 16. Added a global option tls_require_ciphers (compare the smtp transport
5267 option of the same name). This controls incoming TLS connections.
5269 17. I finally figured out how to make tls_require_ciphers do a similar thing
5270 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5271 before starting the TLS session.
5273 18. Tabs are now shown as \t in -bP output.
5275 19. If the log selector return_path_on_delivery was set, Exim crashed when
5276 bouncing a message because it had too many Received: header lines.
5278 20. If two routers both had headers_remove settings, and the first one included
5279 a superfluous trailing colon, the final name in the first list and the
5280 first name in the second list were incorrectly joined into one item (with a
5281 colon in the middle).
5287 1. Added -C and -D options to the exinext utility, mainly to make it easier
5288 to include in the automated testing, but these could be helpful when
5289 multiple configurations are in use.
5291 2. The exinext utility was not formatting the output nicely when there was
5292 an alternate port involved in the retry record key, nor when there was a
5293 message id as well (for retries that were specific to a specific message
5294 and a specific host). It was also confused by IPv6 addresses, because of
5295 the additional colons they contain. I have fixed the IPv4 problem, and
5296 patched it up to do a reasonable job for IPv6.
5298 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5299 delivery, the log line now contains "pipelined" if PIPELINING was used.
5301 4. An SMTP transport process used to panic and die if the bind() call to set
5302 an explicit outgoing interface failed. This has been changed; it is now
5303 treated in the same way as a connect() failure.
5305 5. A reference to $sender_host_name in the part of a conditional expansion
5306 that was being skipped was still causing a DNS lookup. This no longer
5309 6. The def: expansion condition was not recognizing references to header lines
5310 that used bh_ and bheader_.
5312 7. Added the _cache feature to named lists.
5314 8. The code for checking quota_filecount in the appendfile transport was
5315 allowing one more file than it should have been.
5317 9. For compatibility with Sendmail, the command line option
5325 and sets the incoming protocol and host name (for trusted callers). The
5326 host name and its colon can be omitted when only the protocol is to be set.
5327 Note the Exim already has two private options, -pd and -ps, that refer to
5328 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5329 "s", but I don't think that's a major issue.
5331 10. A number of refactoring changes to the code, none of which should affect
5334 (a) The number of logging options was getting close to filling up the
5335 32-bit word that was used as a bit map. I have split them into two classes:
5336 those that are passed in the argument to log_write(), and those that are
5337 only ever tested independently outside of that function. These are now in
5338 separate 32-bit words, so there is plenty of room for expansion again.
5339 There is no change in the user interface or the logging behaviour.
5341 (b) When building, for example, log lines, the code previously used a
5342 macro that called string_cat() twice, in order to add two strings. This is
5343 not really sufficiently general. Furthermore, there was one instance where
5344 it was actually wrong because one of the argument was used twice, and in
5345 one call a function was used. (As it happened, calling the function twice
5346 did not affect the overall behaviour.) The macro has been replaced by a
5347 function that can join an arbitrary number of extra strings onto a growing
5350 (c) The code for expansion conditions now uses a table and a binary chop
5351 instead of a serial search (which was left over from when there were very
5352 few conditions). Also, it now recognizes conditions like "pam" even when
5353 the relevant support is not compiled in: a suitably worded error message is
5354 given if an attempt is made to use such a condition.
5356 11. Added ${time_interval:xxxxx}.
5358 12. A bug was causing one of the ddress fields not to be passed back correctly
5359 from remote delivery subprocesses. The field in question was not being
5360 subsequently used, so this caused to problems in practice.
5362 13. Added new log selectors queue_time and deliver_time.
5364 14. Might have fixed a bug in maildirsizefile handling that threw up
5365 "unexpected character" debug warnings, and recalculated the data
5366 unnecessarily. In any case, I expanded the warning message to give more
5369 15. Added the message "Restricted characters in address" to the statements in
5370 the default ACL that block characters like @ and % in local parts.
5372 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5373 Three changes have been made:
5375 (a) There was a serious bug; a negative response to MAIL caused the whole
5376 recipient domain to be cached as invalid, thereby blocking all messages
5377 to all local parts at the same domain, from all senders. This bug has
5378 been fixed. The domain is no longer cached after a negative response to
5379 MAIL if the sender used is not empty.
5381 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5384 (c) A new callout option, "use_sender" has been added for people who want
5385 the modified behaviour.
5391 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5394 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5395 indeed breaks things for older releases.
5397 3. Added additional logging to the case where there is a problem reading data
5398 from a filter that is running in a subprocess using a pipe, in order to
5399 try to track down a specific problem.
5401 4. Testing facility fudge: when running in the test harness and attempting
5402 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5403 getting "No route to host". Convert this to a timeout.
5405 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5408 6. Some OS don't have socklen_t but use size_t instead. This affects the
5409 fifth argument of getsockopt() amongst other things. This is now
5410 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5411 can be set for individual OS. I have set it for SunOS5, OSF1, and
5412 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5413 some earlier ones do not.
5415 7. Change 4.30/15 was not doing the test caselessly.
5417 8. The standard form for an IPv6 address literal was being rejected by address
5418 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5419 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5420 this, as well as the form without the "IPv6" on the front (but only when
5421 address literals are enabled, of course).
5423 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5425 10. Exim crashed if a message with an empty sender address specified by -f
5426 encountered a router with an errors_to setting. This could be provoked only
5427 by a command such as
5431 where an empty string was supplied; "<>" did not hit this bug.
5433 11. Installed PCRE release 4.5.
5435 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5436 remained set. It is now erased.
5438 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5439 times from message ids (which are base 36 rather than the normal 62).
5441 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5442 were being counted as actual protocol errors, and logged if the log
5443 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5444 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5445 and DATA following a set of rejected RCPTs do not count as protocol errors.
5446 In other words, Exim assumes they were pipelined, though this may not
5447 actually be the case. Of course, in all cases the client gets an
5448 appropriate error code.
5450 15. If a lookup fails in an ACL condition, a message about the failure may
5451 be available; it is used if testing the ACL cannot continue, because most
5452 such messages specify what the cause of the deferral is. However, some
5453 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5454 that caused an old message to be retained and used if a later statement
5455 caused a defer, replacing the real cause of the deferral.
5457 16. If an IP address had so many PTR records that the DNS lookup buffer
5458 was not large enough to hold them, Exim could crash while trying to process
5459 the truncated data. It now detects and logs this case.
5461 17. Further to 4.21/58, another change has been made: if (and only if) the
5462 first line of a message (the first header line) ends with CRLF, a bare LF
5463 in a subsequent header line has a space inserted after it, so as not to
5464 terminate the header.
5466 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5467 unnecessarily, used atoi() instead of strtol(), and didn't check the
5468 termination when getting file sizes from file names by regex.
5470 19. Completely re-implemented the support for maildirsize files, in the light
5471 of a number of problems with the previous contributed implementation
5472 (4.30/29). In particular:
5474 . If the quota is zero, the maildirsize file is maintained, but no quota is
5477 . If the maildir directory does not exist, it is created before any attempt
5478 to write a maildirsize file.
5480 . The quota value in the file is just a cache; if the quota is changed in
5481 the transport, the new value overrides.
5483 . A regular expression is available for excluding directories from the
5486 20. The autoreply transport checks the characters in options that define the
5487 message's headers; it allows continued headers, but it was checking with
5488 isspace() after an embedded newline instead of explicitly looking for a
5491 21. If all the "regular" hosts to which an address was routed had passed their
5492 expiry times, and had not reached their retry times, the address was
5493 bounced, even if fallback hosts were defined. Now Exim should go on to try
5496 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5497 equivalent code in the SMTP transport. Some hosts send humungous responses
5498 to HELO/EHLO, more than 1024 it seems.
5500 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5501 gives compiler warnings in some environments. I've now done it "properly",
5504 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5505 (because of problems with the built-in one) was declared to return uschar *
5506 instead of char *, causing compiler failure.
5508 25. Fixed a file descriptor leak when processing alias/forward files.
5510 26. Fixed a minor format string issue in dbfn.c.
5512 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5514 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5515 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5518 29. When a sender address is verified, it is cached, to save repeating the test
5519 when there is more than one recipient in a message. However, when the
5520 verification involves a callout, it is possible for different callout
5521 options to be set for different recipients. It is too complicated to keep
5522 track of this in the cache, so now Exim always runs a verification when a
5523 callout is required, relying on the callout cache for the optimization.
5524 The overhead is duplication of the address routing, but this should not be
5527 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5528 to be verified with callout=postmaster, and the main callout worked but the
5529 postmaster check failed, the verification correctly failed. However, if a
5530 subsequent RCPT command asked for sender verification *without* the
5531 postmaster check, incorrect caching caused this verification also to fail,
5534 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5535 it was not caching the DNS options (qualify_single, search_parents) that
5536 were used when the lookup failed. A subsequent lookup with different
5537 options therefore always gave the same answer, though there were cases
5538 where it should not have. (Example: a "domains = !$mx_any" option on a
5539 dnslookup router: the "domains" option is always processed without any
5540 widening, but the router might have qualify_single set.) Now Exim uses the
5541 cached value only when the same options are set.
5543 32. Added John Jetmore's "exipick" utility to the distribution.
5545 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5546 than a timeout (e.g. a certificate is required, and is not provided), an
5547 Exim server now closes the connection immediately. Previously it waited for
5548 the client to close - but if the client is SSL, it seems that they each
5549 wait for each other, leading to a delay before one of them times out.
5551 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5552 maintained 0.8.x compatibility because I don't think many are using it, and
5553 it is clearly obsolete.
5555 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5558 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5559 client certificate was expired. A simple patch fixes this, though I don't
5560 understand the full logic of why the verify callback is called multiple
5563 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5564 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5565 which causes problems with some clients (such as the Certicom SSL Plus
5566 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5567 disables the coutermeasure allowing Eudora to connect."
5569 38. Exim was not checking that a write() to a log file succeeded. This could
5570 lead to Bad Things if a log got too big, in particular if it hit a file
5571 size limit. Exim now panics and dies if it cannot write to a log file, just
5572 as it does if it cannot open a log file.
5574 39. Modified OS/Makefile-Linux so that it now contains
5576 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5578 The two -D definitions ensure that Exim is compiled with large file
5579 support, which makes it possible to handle log files that are bigger than
5582 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5583 instance) a domain was checked against a named list that involved a lookup,
5584 causing $domain_data to be set, then another domain was checked against the
5585 same list, then the first domain was re-checked, the value of $domain_data
5586 after the final check could be wrong. In particular, if the second check
5587 failed, it could be set empty. This bug probably also applied to
5590 41. The strip_trailing_dot option was not being applied to the address given
5591 with the -f command-line option.
5593 42. The code for reading a message's header from the spool was incrementing
5594 $received_count, but never initializing it. This meant that the value was
5595 incorrect (doubled) while delivering a message in the same process in which
5596 it was received. In the most common configuration of Exim, this never
5597 happens - a fresh exec is done - but it can happen when
5598 deliver_drop_privilege is set.
5600 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5601 it now includes up to 150 characters of the unexpected data in the log
5604 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5605 and building data strings. The size of both of these buffers was 10 000
5606 bytes - far larger than anybody would *ever* want, thought I. Needless to
5607 say, somebody hit the limit. I have increased the maximum line length to
5608 20 000 and the maximum data length of concatenated lines to 100 000. I have
5609 also fixed two bugs, because there was no checking on these buffers. Tsk,
5610 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5611 buffer is too small.
5613 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5614 lsearch lookups. Now it does.
5616 46. When parsing a route_list item in a manualroute router, a fixed-length
5617 buffer was used for the list of hosts. I made this 1024 bytes long,
5618 thinking that nobody would ever have a list of hosts that long. Wrong.
5619 Somebody had a whole pile of complicated expansion conditions, and the
5620 string was silently truncated, leading to an expansion error. It turns out
5621 that it is easier to change to an unlimited length (owing to other changes
5622 that have happened since this code was originally written) than to build
5623 structure for giving a limitation error. The length of the item that
5624 expands into the list of hosts is now unlimited.
5626 47. The lsearch lookup could not handle data where the length of text line was
5627 more than 4095 characters. Such lines were truncated, leading to shortened
5628 data being returned. It should now handle lines of any length.
5630 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5631 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5634 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5635 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5636 no longer happens in comment lines. A list of replacements is now placed
5637 at the head of all of the source files, except those whose only change is
5638 to replace PERL_COMMAND in the very first #! line.
5640 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5641 messages on the queue, with a bottom-up merge sort, using code contributed
5642 by Michael Haardt. This should make operations like -bp somewhat faster on
5643 large queues. It won't affect queue runners, except when queue_run_in_order
5646 51. Installed eximstats 1.31 in the distribution.
5648 52. Added support for SRV lookups to the dnslookup router.
5650 53. If an ACL referred to $message_body or $message_body_end, the value was not
5651 reset for any messages that followed in the same SMTP session.
5653 54. The store-handling optimization for building very long strings was not
5654 differentiating between the different store pools. I don't think this
5655 actually made any difference in practice, but I've tidied it.
5657 55. While running the routers to verify a sender address, $sender_address
5658 was still set to the sender address. This is wrong, because when routing to
5659 send a bounce to the sender, it would be empty. Therefore, I have changed
5660 it so that, while verifying a sender address, $sender_address is set to <>.
5661 (There is no change to what happens when verifying a recipient address.)
5663 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5664 target A or AAAA records (if not already returned) without resetting the
5665 qualify_single or search_parents options of the DNS resolver. These are
5666 inappropriate in this case because the targets of MX and SRV records must
5667 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5668 target that, when qualified, matched something in the local domain. These
5669 two options are now turned off when doing these lookups.
5671 57. It seems that at least some releases of Reiserfs (which does not have the
5672 concept of a fixed number of inodes) returns zero and not -1 for the
5673 number of available inodes. This interacted badly with check_spool_inodes,
5674 which assumed that -1 was the "no such thing" setting. What I have done is
5675 to check that the total number of inodes is greater than zero before doing
5676 the test of how many are available.
5678 58. When a "warn" ACL statement has a log_message modifier, the message is
5679 remembered, and not repeated. This is to avoid a lot of repetition when a
5680 message has many recipients that cause the same warning to be written.
5681 However, Exim was preserving the list of already written lines for an
5682 entire SMTP session, which doesn't seem right. The memory is now reset if a
5683 new message is started.
5685 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5686 debugging output unless log_rewrite was also set.
5688 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5689 of (int)(handle) when we know that handle contains (void *)(-1).
5691 61. The Exim daemon panic-logs an error return when it closes the incoming
5692 connection. However "connection reset by peer" seems to be common, and
5693 isn't really an error worthy of noting specially, so that particular error
5696 62. When Exim is trying to find all the local interfaces, it used to panic and
5697 die if the ioctl to get the interface flags failed. However, it seems that
5698 on at least one OS (Solaris 9) it is possible to have an interface that is
5699 included in the list of interfaces, but for which you get a failure error
5700 for this call. This happens when the interface is not "plumbed" into a
5701 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5702 failure of the "get flags" call assumes that the interface is down.
5704 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5705 makes life easier for people who are doing arithmetic on fields extracted
5706 from dates, where you often get leading zeros that should not be
5707 interpreted as octal.
5709 64. Added qualify_domain to the redirect router, to override the global
5712 65. If a pathologically long header line contained very many addresses (the
5713 report of this problem mentioned 10 000) and each of them was rewritten,
5714 Exim could use up a very large amount of memory. (It kept on making new
5715 copies of the header line as it rewrote, and never released the old ones.)
5716 At the expense of a bit more processing, the header rewriting function has
5717 been changed so that it no longer eats memory in this way.
5719 66. The generation of the Received: header has been moved from the time that a
5720 message starts to be received, to the time that it finishes. The timestamp
5721 in the Received: header should now be very close to that of the <= log
5722 line. There are two side-effects of this change:
5724 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5725 logged header lines no longer include the local Received: line, because
5726 it has not yet been created. The same applies to a copy of the message
5727 that is returned to a non-SMTP sender when a message is rejected.
5729 (b) When a filter file is tested using -bf, no additional Received: header
5730 is added to the test message. After some thought, I decided that this
5733 This change does not affect the value of $received_for. It is still set
5734 after address rewriting, but before local_scan() is called.
5736 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5738 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5739 gave an unhelpful panic error message, and a defer error. I have managed to
5740 change this behaviour so that it now rejects any supplied certificate,
5741 which seems right, as the list of acceptable certificates is empty.
5743 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5744 gave an unhelpful defer error. I have not managed to make this reject any
5745 supplied certificates, but the error message it gives is "no certificate
5746 supplied", which is not helpful.
5748 70. exigrep's output now also includes lines that are not associated with any
5749 message, but which match the given pattern. Implemented by a patch from
5750 Martin Sluka, which also tidied up the Perl a bit.
5752 71. Recipient callout verification, like sender verification, was using <> in
5753 the MAIL FROM command. This isn't really the right thing, since the actual
5754 sender may affect whether the remote host accepts the recipient or not. I
5755 have changed it to use the actual sender in the callout; this means that
5756 the cache record is now keyed on a recipient/sender pair, not just the
5757 recipient address. There doesn't seem to be a real danger of callout loops,
5758 since a callout by the remote host to check the sender would use <>.
5759 [SEE ABOVE: changed after hitting problems.]
5761 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5762 temporary errors. However, in the case of such a code being given after
5763 the end of a data transmission (i.e. after ".") Exim was failing to write
5764 a retry record for the message. (Yes, there was some broken host that was
5765 actually sending 8xx at this point.)
5767 73. An unknown lookup type in a host list could cause Exim to panic-die when
5768 the list was checked. (An example that provoked this was putting <; in the
5769 middle of a list instead of at the start.) If this happened during a DATA
5770 ACL check, a -D file could be left lying around. This kind of configuration
5771 error no longer causes Exim to die; instead it causes a defer error. The
5772 incident is still logged to the main and panic logs.
5774 74. Buglet left over from Exim 3 conversion. The message "too many messages
5775 in one connection" was written to the rejectlog but not the mainlog, except
5776 when address rewriting (yes!) was being logged.
5778 75. Added write_rejectlog option.
5780 76. When a system filter was run not as root (that is, when system_filter_user
5781 was set), the values of the $n variables were not being returned to the
5782 main process; thus, they were not subsequently available in the $sn
5785 77. Added +return_path_on_delivery log selector.
5787 78. A connection timeout was being treated differently from recipients deferred
5788 when testing hosts_max_try with a message that was older than the host's
5789 retry timeout. (The host should not be counted, thus allowing all hosts to
5790 be tried at least once before bouncing.) This may have been the cause of an
5791 occasionally reported bug whereby a message would remain on the queue
5792 longer than the retry timeout, but would be bounced if a delivery was
5793 forced. I say "may" because I never totally pinned down the problem;
5794 setting up timeout/retry tests is difficult. See also the next item.
5796 79. The ultimate address timeout was not being applied to errors that involved
5797 a combination of host plus message (for example, a timeout on a MAIL
5798 command). When an address resolved to a number of possible hosts, and they
5799 were not all tried for each delivery (e.g. because of hosts_max_try), a
5800 message could remain on the queue longer than the retry timeout.
5802 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5805 81. Fixed an obscure SMTP outgoing bug which required at least the following
5806 conditions: (a) there was another message waiting for the same server;
5807 (b) the server returned 5xx to all RCPT commands in the first message so
5808 that the message was not completed; (c) the server dropped the connection
5809 or gave a negative response to the RSET that Exim sends to abort the
5810 transaction. The observed case was a dropped connection after DATA that had
5811 been sent in pipelining mode. That is, the server had advertised PIPELINING
5812 but was not implementing it correctly. The effect of the bug was incorrect
5813 behaviour, such as trying another host, and this could lead to a crash.
5819 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5820 and daemon.c were passed as pointers to ints; they should have been
5821 pointers to socklen_t variables (which are typically unsigned ints).
5823 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5826 3. Fixed a really odd bug that affected only the testing scheme; patching a
5827 certain fixed string in the binary changed the value of another string that
5828 happened to be identical to the end of the original first string.
5830 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5831 name", it returns that address as the IP address. On some operating
5832 systems (e.g. Solaris), it also passes back the IP address string as the
5833 "host name". However, on others (e.g. Linux), it passes back an empty
5834 string. Exim wasn't checking for this, and was changing the host name to an
5835 empty string, assuming it had been canonicalized.
5837 5. Although rare, it is permitted to have more than one PTR record for a given
5838 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5839 all the names associated with an address, because they do in Solaris.
5840 However, it seems that they do not in Linux for data that comes from the
5841 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5842 I found this out when I moved to a new Linux workstation and tried to run
5843 the Exim test suite.
5845 To get round this problem I have changed the code so that it now does its
5846 own call to the DNS to look up PTR records when searching for a host name.
5847 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5848 addresses that are only in /etc/hosts are still found.
5850 This behaviour is, however, controlled by an option called host_lookup_
5851 order, which defaults to "bydns:byaddr". If people want to use the other
5852 order, or indeed, just use one or the other means of lookup, they can
5853 specify it in this variable.
5855 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5856 some operating systems, this comes back from gethostbyaddr() as an empty
5857 string, and this is what Exim used to test for. However, it seems that in
5858 other systems, "." is yielded. Exim now tests for this case too.
5860 7. The values of check_spool_space and check_log_space are now held internally
5861 as a number of kilobytes instead of an absolute number of bytes. If a
5862 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5863 kilobyte. This means that much larger values can be stored.
5865 8. Exim monitor: an attempt to get the action menu when not actually pointing
5866 at a message produces an empty menu entitled "No message selected". This
5867 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5868 no entries in it ("Shell widget menu has zero width and/or height"). So I
5869 have added a single, blank menu entry in this case.
5871 9. Added ${quote_local_part.
5873 10. MIME decoding is now applied to the contents of Subject: header lines when
5876 11. Now that a reference to $sender_host_address automatically causes a reverse
5877 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5878 host lookup before query-style lookups in lists that might use this
5879 variable. This has therefore been abolished, and the "net-" prefix is no
5880 longer necessary for query-style lookups.
5882 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5883 have been a typo for LFLAGS, so it has been changed.
5885 13. The install script calls Exim with "-C /dev/null" in order to find the
5886 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5887 to be output. However, since Exim outputs its version number before the
5888 error, it didn't break the script. It just looked ugly. I fixed this by
5889 always allowing "-C /dev/null" if the caller is root.
5891 14. Ignore overlarge ACL variable number when reading spool file - insurance
5892 against a later release with more variables having written the file.
5894 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5895 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5896 this, as well as the form without the "IPv6" on the front.
5898 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5899 OS/Makefile-Darwin file.
5901 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5902 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5904 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5905 variable. However, some OS use ints for inodes. I've added cast to long int
5906 to get rid of the compiler warning.
5908 19. I had forgotten to lock out "/../" in configuration file names when
5909 ALT_CONFIG_PREFIX was set.
5911 20. Routers used for verification do not need to specify transports. However,
5912 if such a router generated a host list, and callout was configured, Exim
5913 crashed, because it could not find a port number from the (non-existent)
5914 transport. It now assumes port 25 in this circumstance.
5916 21. Added the -t option to exigrep.
5918 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5919 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5920 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5921 them is set, LOOKUP_LSEARCH is forced.
5923 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5925 24. Added sender and host information to the "rejected by local_scan()" log
5926 line; previously there was no indication of these.
5928 25. Added .include_if_exists.
5930 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5931 Linux. It turns out that not all file systems support this. Apparently some
5932 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5933 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5934 not supported on the file descriptor, is now ignored when Exim is trying to
5935 sync a directory. This applies only to Linux.
5937 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5939 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5940 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5943 29. Added support for maildirsize files from supplied patch (modified a bit).
5945 30. The use of :fail: followed by an empty string could lead Exim to respond to
5946 sender verification failures with (e.g.):
5948 550 Verification failed for <xxx>
5949 550 Sender verify failed
5951 where the first response line was missing the '-' that indicates it is not
5952 the final line of the response.
5954 31. The loop for finding the name of the user that called Exim had a hardwired
5955 limit of 10; it now uses the value of finduser_retries, which is used for
5956 all other user lookups.
5958 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5961 33. Exim was neglecting to zero errno before one call of strtol() when
5962 expanding a string and expecting an integer value. On some systems this
5963 resulted in spurious "integer overflow" errors. Also, it was casting the
5964 result into an int without checking.
5966 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5967 did not work. The code looks as if it has *never* worked, though it appears
5968 to have been documented since at least release 1.62. I have made it work.
5970 35. The "timeout_DNS" error in retry rules, also documented since at least
5971 1.62, also never worked. As it isn't clear exactly what this means, and
5972 clearly it isn't a major issue, I have abolished the feature by treating it
5973 as "timeout", and writing a warning to the main and panic logs.
5975 36. The display of retry rules for -brt wasn't always showing the error code
5978 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5979 timeout_connect_A, timeout_connect_MX.
5981 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5982 to the empty sender.
5984 39. The daemon was not analysing the content of -oX till after it had closed
5985 stderr and disconnected from the controlling terminal. This meant that any
5986 syntax errors were only noted on the panic log, and the return code from
5987 the command was 0. By re-arranging the code a little, I've made the
5988 decoding happen first, so such errors now appear on stderr, and the return
5989 code is 1. However, the actual setting up of the sockets still happens in
5990 the disconnected process, so errors there are still only recorded on the
5993 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5994 connections (as happens on some IP stacks) was logged at start up time as
5995 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5996 it from "IPv6 and IPv4", which means that two separate sockets are being
5999 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6000 says whether AF_INET or AF_INET6 was passed as an argument.
6002 42. Exiwhat output was messed up when time zones were included in log
6005 43. Exiwhat now gives more information about the daemon's listening ports,
6006 and whether -tls-on-connect was used.
6008 44. The "port" option of the smtp transport is now expanded.
6010 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6011 silently ignored. Now an error message is written to the main and panic
6014 46. There's a new ACL modifier called "logwrite" which writes to a log file
6015 as soon as it is encountered.
6017 47. Added $local_user_uid and $local_user_gid at routing time.
6019 48. Exim crashed when trying to verify a sender address that was being
6022 49. Exim was recognizing only a space character after ".include". It now also
6023 recognizes a tab character.
6025 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6026 extracting the relevant information from the specification. The man page no
6027 longer contains scrambled data for the -d option, and I've added a section
6028 at the front about calling Exim under different names.
6030 51. Added "extra_headers" argument to the "mail" command in filter files.
6032 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6035 53. Installed eximstats 1.29.
6037 54. Added transport_filter_timeout as a generic transport option.
6039 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6040 Cc: header lines. This was required by RFC 822, but it not required by RFC
6043 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6044 incoming messages that did not have them. Now it does so only if the
6045 message originates locally, that is, if there is no associated remote host
6046 address. When Resent- header lines are present, this applies to the Resent-
6047 lines rather than the non-Resent- lines.
6049 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6050 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6052 58. Messages for configuration errors now include the name of the main
6053 configuration file - useful now that there may be more than one file in a
6054 list (.included file names were always shown).
6056 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6057 for those rare installations that do not start the daemon as root or run it
6058 setuid root. I've cut out the call to initgroups() if the daemon is not
6061 60. The Exim user and group can now be bound into the binary as text strings
6062 that are looked up at the start of Exim's processing.
6064 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6066 62. Added $mailstore_basename variable.
6068 63. Installed patch to sieve.c from Michael Haardt.
6070 64. When Exim failed to open the panic log after failing to open the main log,
6071 the original message it was trying to log was written to stderr and debug
6072 output, but if they were not available (the usual case in production), it
6073 was lost. Now it is written to syslog before the two lines that record the
6074 failures to open the logs.
6076 65. Users' Exim filters run in subprocesses under the user's uid. It is
6077 possible for a "deliver" command or an alias in a "personal" command to
6078 provoke an address rewrite. If logging of address rewriting is configured,
6079 this fails because the process is not running as root or exim. There may be
6080 a better way of dealing with this, but for the moment (because 4.30 needs
6081 to be released), I have disabled address rewrite logging when running a
6082 filter in a non-root, non-exim process.
6088 1. The buildconfig auxiliary program wasn't quoting the value set for
6089 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6090 not defined. This bug was masked in 4.22 by the effect that was fixed in
6093 2. Some messages that were rejected after a message id was allocated were
6094 shown as "incomplete" by exigrep. It no longer does this for messages that
6095 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6097 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6098 have allow_domain_literals set, the ID did not get logged in the <= line.
6099 Domain literals are now always recognized in Message-ID: header lines.
6101 4. The first argument for a ${extract expansion item is the key name or field
6102 number. Leading and trailing spaces in this item were not being ignored,
6103 causing some misleading effects.
6105 5. When deliver_drop_privilege was set, single queue runner processes started
6106 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6107 same command in the process it spins off) were not dropping privilege.
6109 6. When the daemon running as "exim" started a queue runner, it always
6110 re-executed Exim in the spun-off process. This is a waste of effort when
6111 deliver_drop_privilege is set. The new process now just calls the
6112 queue-runner function directly.
6118 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6121 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6122 ignored. Though the use of -oP was forcing the writing of a pid file, it
6123 was always written to the default place.
6125 3. If the message "no IP address found for host xxxx" is generated during
6126 incoming verification, it is now followed by identification of the incoming
6127 connection (so you can more easily find what provoked it).
6129 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6131 5. Added some features to "harden" Exim a bit more against certain attacks:
6133 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6134 be put in Local/Makefile. This is like the never_users runtime option,
6135 but it cannot be overridden. The default setting is "root".
6137 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6138 prefix string with which any file named in a -C command line option
6141 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6142 is retained for -C and -D only if the caller of Exim is root. Without
6143 it, the exim user may also use -C and -D and retain privilege.
6145 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6146 command line option is disabled.
6148 6. Macro names set by the -D option must start with an upper case letter, just
6149 like macro names defined in the configuration file.
6151 7. Added "dereference=" facility to LDAP.
6153 8. Two instances of the typo "uknown" in the source files are fixed.
6155 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6156 the Configure-Makefile script screwed up while processing it.
6158 10. Incorporated PCRE 4.4.
6160 11. The SMTP synchronization check was not operating right at the start of an
6161 SMTP session. For example, it could not catch a HELO sent before the client
6162 waited for the greeting. There is now a check for outstanding input at the
6163 point when the greeting is written. Because of the duplex, asynchronous
6164 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6165 way, but not yet received, when the check is performed.
6167 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6168 on TCP/IP sockets, because this apparently causes some broken clients to
6171 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6172 unchanged) from the Cygwin maintainer.
6174 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6175 was supported instead of testing for mbx. Effectively a typo.
6177 15. The spa authenticator server code was not checking that the input it
6178 received was valid base64.
6180 16. The debug output line for the "set" modifier in ACLs was not showing the
6181 name of the variable that was being set.
6183 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6185 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6186 lookup on its own. Something else was needed to trigger the lookup. For
6187 example, a match in host_lookup or the need for a host name in a host list.
6188 Now, if $sender_host_name is referenced and the host name has not yet been
6189 looked up, a lookup is performed. If the lookup fails, the variable remains
6190 empty, and $host_lookup_failed is set to "1".
6192 19. Added "eqi" as a case-independent comparison operator.
6194 20. The saslauthd authentication condition could segfault if neither service
6195 nor realm was specified.
6197 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6198 error message that was logged was misleading, and incoming SMTP
6199 connections were dropped. The message is now more accurate, and temporary
6200 errors are given to SMTP connections.
6202 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6203 (except RSET) with a 421 error, until QUIT is received. However, it was
6204 failing to send a response to QUIT.
6206 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6207 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6208 ACL is now run after the helo_try_verify_hosts code.
6210 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6211 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6212 case-independent, so other case variants are also recognized). Apparently
6213 some systems use these upper case variants.
6215 25. If more than two messages were waiting for the same host, and a transport
6216 filter was specified for the transport, Exim sent two messages over the
6217 same TCP/IP connection, and then failed with "socket operation on non-
6218 socket" when it tried to send the third.
6220 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6222 27. The extern definition of crypt16() in expand.c was not being excluded when
6223 the OS had its own crypt16() function.
6225 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6226 as a preferred synonym for return_size_limit, both as an option and as an
6229 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6231 30. Changed the default configuration ACL to relax the local part checking rule
6232 for addresses that are not in any local domains. For these addresses,
6233 slashes and pipe symbols are allowed within local parts, but the sequence
6234 /../ is explicitly forbidden.
6236 31. SPA server authentication was not clearing the challenge buffer before
6239 32. log_message in a "warn" ACL statement was writing to the reject log as
6240 well as to the main log, which contradicts the documentation and doesn't
6241 seem right (because no rejection is happening). So I have stopped it.
6243 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6244 However, I am unable to do any testing of this.
6246 34. Fixed an infelicity in the appendfile transport. When checking directories
6247 for a mailbox, to see if any needed to be created, it was accidentally
6248 using path names with one or more superfluous leading slashes; tracing
6249 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6251 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6252 contents are added to the log line that is written for every discarded
6253 recipient. (Previously a log_message setting was ignored.)
6255 36. The ${quote: operator now quotes the string if it is empty.
6257 37. The install script runs exim in order to find its version number. If for
6258 some reason other than non-existence or emptiness, which it checks, it
6259 could not run './exim', it was installing it with an empty version number,
6260 i.e. as "exim-". This error state is now caught, and the installation is
6263 38. An argument was missing from the function that creates an error message
6264 when Exim fails to connect to the socket for saslauthd authentication.
6265 This could cause Exim to crash, or give a corrupted message.
6267 39. Added isip, isip4, and isip6 to ${if conditions.
6269 40. The ACL variables $acl_xx are now saved with the message, and can be
6270 accessed later in routers, transports, and filters.
6272 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6273 strings in the file are not string-expanded.
6275 42. If a MAIL command specified a SIZE value that was too large to fit into an
6276 int variable, the check against message_size_limit failed. Such values are
6277 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6278 day this will have to be increased, but I don't think I want to be around
6279 when emails are that large.
6286 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6287 iconv() is not standard in FreeBSD.
6289 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6290 IPv6 enabled. The observed symptom was a segmentation fault on return from
6291 the function os_common_find_running_interfaces() in src/os.c.
6293 3. In the check_special_case() function in daemon.c I had used "errno" as an
6294 argument name, which causes warnings on some systems. This was basically a
6295 typo, since it was named "eno" in the comments!
6297 4. The code that waits for the clock to tick (at a resolution of some fraction
6298 of a second) so as to ensure message-id uniqueness was always waiting for
6299 at least one whole tick, when it could have waited for less. [This is
6300 almost certainly not relevant at current processor speeds, where it is
6301 unlikely to ever wait at all. But we try to future-proof.]
6303 5. The function that sleeps for a time interval that includes fractions of a
6304 second contained a race. It did not block SIGALRM between setting the
6305 timer, and suspending (a couple of lines later). If the interval was short
6306 and the sigsuspend() was delayed until after it had expired, the suspension
6307 never ended. On busy systems this could lead to processes getting stuck for
6310 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6311 process, before it forks any delivery processes. The open lookup caching
6312 mechanism meant that the open file or database connection was passed into
6313 the delivery process. The problem was that delivery processes always tidy
6314 up cached lookup data. This could cause a problem for the next delivery
6315 process started by the queue runner, because the external queue runner
6316 process does not know about the closure. So the next delivery process
6317 still has data in the lookup cache. In the case of a file lookup, there was
6318 no problem because closing a file descriptor in a subprocess doesn't affect
6319 the parent. However, if the lookup was caching a connection to a database,
6320 the connection was closed, and the second delivery process was likely to
6321 see errors such as "PGSQL: query failed: server closed the connection
6322 unexpectedly". The problem has been fixed by closing all cached lookups
6323 in a queue runner before running a delivery process.
6325 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6326 seem to have the "const" qualifier which it has on other OS. I've
6329 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6330 *of the same type* (client or server) with the same public name that an
6331 error should be diagnosed.
6333 9. When Exim looked up a host name for an IP address, but failed to find the
6334 original IP address when looking up the host name (a safety check), it
6335 output the message "<ip address> does not match any IP for NULL", which was
6336 confusing, to say the least. The bug was that the host name should have
6337 appeared instead of "NULL".
6339 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6340 user that is built into the binary, and the -C or -D options is used, root
6341 privilege is dropped before the configuration file is read. In addition,
6342 logging is switched to stderr instead of the normal log files. If the
6343 configuration then re-defines the Exim user, the unprivileged environment
6344 is probably not what is expected, so Exim logs a panic warning message (but
6347 However, if deliver_drop_privilege is set, the unprivileged state may well
6348 be exactly what is intended, so the warning has been cut out in that case,
6349 and Exim is allowed to try to write to its normal log files.
6355 1. smtp_return_error_details was not giving details for temporary sender
6356 or receiver verification errors.
6358 2. Diagnose a configuration error if two authenticators have the same public
6361 3. Exim used not to create the message log file for a message until the first
6362 delivery attempt. This could be confusing when incoming messages were held
6363 for policy or load reasons. The message log file is now created at the time
6364 the message is received, and an initial "Received" line is written to it.
6366 4. The automatically generated man page for command line options had a minor
6367 bug that caused no ill effects; however, a more serious problem was that
6368 the procedure for building the man page automatically didn't always
6369 operate. Consequently, release 4.20 contains an out-of-date version. This
6370 shouldn't happen again.
6372 5. When building Exim with embedded Perl support, the script that builds the
6373 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6374 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6376 6. The freeze_tell option was not being used for messages that were frozen on
6377 arrival, either by an ACL or by local_scan().
6379 7. Added the smtp_incomplete_transaction log selector.
6381 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6382 was accepting AUTH without a new EHLO.
6384 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6385 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6388 10. Logging of TCP/IP connections (when configured) now happens in the main
6389 daemon process instead of the child process, so that the TCP/IP connection
6390 count is more accurate (but it can never be perfect).
6392 11. The use of "drop" in a nested ACL was not being handled correctly in the
6393 outer ACL. Now, if condition failure induced by the nested "drop" causes
6394 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6395 or "require"), the connection is dropped.
6397 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6398 that yield "discard" can now be used with an "accept" or a "discard" verb,
6399 but an error is generated for any others (because I can't see a useful way
6400 to define what should happen).
6402 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6403 are now processed in the same way as lines in the Exim configuration file.
6404 In particular, continuation lines are supported.
6406 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6408 15. Added -ti meaning -t -i.
6410 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6411 domains, and warn by logging if others are found.
6413 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6414 returned from GIFCONF when Exim is trying to find the list of interfaces on
6415 a host. The code in os.c has been modified to copy each ifreq to an aligned
6416 structure in all cases.
6418 Also, in some cases, the returned ifreq's were being copied to a 'struct
6419 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6420 means the last couple of bytes of an IPv6 address could be chopped if the
6421 ifreq contained only a normal sockaddr (14 bytes storage).
6423 18. Named domain lists were not supported in the hosts_treat_as_local option.
6424 An entry such as +xxxx was not recognized, and was treated as a literal
6427 19. Ensure that header lines added by a DATA ACL are included in the reject log
6428 if the ACL subsequently rejects the message.
6430 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6431 MD5 (which is deprecated).
6433 21. When testing a filter file using -bf, Exim was writing a message when it
6434 took the sender from a "From " line in the message, but it was not doing so
6435 when it took $return_path from a Return-Path: header line. It now does.
6437 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6438 with a valid header line field name (a series of printing characters
6439 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6441 23. Changed "disc" in the source to "disk" to conform to the documentation and
6442 the book and for uniformity.
6444 24. Ignore Sendmail's -Ooption=value command line item.
6446 25. When execve() failed while trying to run a command in a pipe transport,
6447 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6448 could be confused with a return value of 69 from the command itself. This
6449 has been changed to 127, the value the shell returns if it is asked to run
6450 a non-existent command. The wording for the related log line suggests a
6451 non-existent command as the problem.
6453 26. If received_header_text expands to an empty string, do not add a Received:
6454 header line to the message. (Well, it adds a token one on the spool, but
6455 marks it "old" so that it doesn't get used or transmitted.)
6457 27. Installed eximstats 1.28 (addition of -nt option).
6459 28. There was no check for failure on the call to getsockname() in the daemon
6460 code. This can fail if there is a shortage of resources on the system, with
6461 ENOMEM, for example. A temporary error is now given on failure.
6463 29. Contrary to the C standard, it seems that in some environments, the
6464 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6465 program. Exim now does this explicitly; it affects the formatting of
6466 timestamps using strftime().
6468 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6469 complaints. I've now initialized all the variables, to avoid this.
6471 32. Header lines added by a system filter were not being "seen" during
6472 transport-time rewrites.
6474 33. The info_callback() function passed to OpenSSL is set up with type void
6475 (*)(SSL *, int, int), as described somewhere. However, when calling the
6476 function (actually a macro) that sets it up, the type void(*)() is
6477 expected. I've put in a cast to prevent warnings from picky compilers.
6479 34. If a DNS black list lookup found a CNAME record, but there were no A
6480 records associated with the domain it pointed at, Exim crashed.
6482 35. If a DNS black list lookup returned more than one A record, Exim ignored
6483 all but the first. It now scans all returned addresses if a particular IP
6484 value is being sought. In this situation, the contents of the
6485 $dnslist_value variable are a list of all the addresses, separated by a
6488 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6489 to accept a host name and all its aliases if the forward lookup for any of
6490 them yielded the IP address of the incoming connection. Now it accepts only
6491 those names whose forward lookup yields the correct IP address. Any other
6492 names are discarded. This closes a loophole whereby a rogue DNS
6493 administrator could create reverse DNS records to break through a
6494 wildcarded host restriction in an ACL.
6496 37. If a user filter or a system filter that ran in a subprocess used any of
6497 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6498 the wrong values were passed to the pipe command ($thisaddress had the
6499 value of $0, $0 had the value of $1, etc). This bug was introduced by
6500 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6502 38. Improved the line breaking for long SMTP error messages from ACLs.
6503 Previously, if there was no break point between 40 and 75 characters, Exim
6504 left the rest of the message alone. Two changes have been made: (a) I've
6505 reduced the minimum length to 35 characters; (b) if it can't find a break
6506 point between 35 and 75 characters, it looks ahead and uses the first one
6507 that it finds. This may give the occasional overlong line, but at least the
6508 remaining text gets split now.
6510 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6511 file descriptors might be low, and that setting 1000 would always raise it.
6512 It turns out that in some environments, the limit is already over 1000 and
6513 that lowering it causes trouble. So now Exim takes care not to decrease it.
6515 40. When delivering a message, the value of $return_path is set to $sender_
6516 address at the start of routing (routers may change the value). By an
6517 oversight, this default was not being set up when an address was tested by
6518 -bt or -bv, which affected the outcome if any router or filter referred to
6521 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6522 log only when "message" or "log_message" are set. However, if one of the
6523 conditions was an address verification, or a call to a nested ACL, the
6524 messages generated by the underlying test were being passed through. This
6525 no longer happens. The underlying message is available in $acl_verify_
6526 message for both "message" and "log_message" expansions, so it can be
6527 passed through if needed.
6529 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6530 new expansion $bh_ to give the encoded byte string without charset
6531 translation. Translation happens only if iconv() is available; HAVE_ICONV
6532 indicates this at build time. HEADERS_CHARSET gives the charset to
6533 translate to; headers_charset can change it in the configuration, and
6534 "headers charset" can change it in an individual filter file.
6536 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6537 that creates RFC 2047 encoded "words" labels them as that charset instead
6538 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6539 expansion operator; (ii) when Exim creates a From: line for a local
6540 message; (iii) when a header line is rewritten to include a "phrase" part.
6542 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6543 buggy, causing it to skip the first lines of messages whose message ID
6544 ended in 'D'. This would not have bitten before Exim release 4.14, because
6545 message IDs were unlikely to end in 'D' before then. The effect was to have
6546 incorrect size information for certain domains.
6548 45. #include "config.h" was missing at the start of the crypt16.c module. This
6549 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6552 46. If there was a timeout during a "random" callout check, Exim treated it as
6553 a failure of the random address, and carried on sending RSET and the real
6554 address. If the delay was just some slowness somewhere, the response to the
6555 original RCPT would be taken as a response to RSET and so on, causing
6556 mayhem of various kinds.
6558 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6559 when I implemented it. It didn't allow for the fact that some option values
6560 may legitimately be negative (e.g. size_addition), and it didn't even do
6561 the right test for positive values.
6563 48. Domain names in DNS records are case-independent. Exim always looks them up
6564 in lower case. Some resolvers return domain names in exactly the case they
6565 appear in the zone file, that is, they may contain uppercase letters. Not
6566 all resolvers do this - some return always lower case. Exim was treating a
6567 change of case by a resolver as a change of domain, similar to a widening
6568 of a domain abbreviation. This triggered its re-routing code and so it was
6569 trying to route what was effectively the same domain again. This normally
6570 caused routing to fail (because the router wouldn't handle the domain
6571 twice). Now Exim checks for this case specially, and just changes the
6572 casing of the domain that it ultimately uses when it transmits the message
6575 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6578 50. If a filter generated a file delivery with a non-absolute name (possible if
6579 no home directory exists for the router), the forbid_file option was not
6582 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6583 the existing equality matching.
6585 52. Exim was using ints instead of ino_t variables in some places where it was
6586 dealing with inode numbers.
6588 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6589 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6590 called TMPDIR, and if it finds it is different, it changes its value.
6592 54. The smtp_printf() function is now made available to local_scan() so
6593 additional output lines can be written before returning. There is also an
6594 smtp_fflush() function to enable the detection of a dropped connection.
6595 The variables smtp_input and smtp_batched_input are exported to
6598 55. Changed the default runtime configuration: the message "Unknown user"
6599 has been removed from the ACL, and instead placed on the localuser router,
6600 using the cannot_route_message feature. This means that any verification
6601 failures that generate their own messages won't get overridden. Similarly,
6602 the "Unrouteable address" message that was in the ACL for unverifiable
6603 relay addresses has also been removed.
6605 56. Added hosts_avoid_esmtp to the smtp transport.
6607 57. The exicyclog script was not checking for the esoteric option
6608 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6609 will work only if exicyclog is run under the appropriate euid.
6611 58. Following a discussion on the list, the rules by which Exim recognises line
6612 endings on incoming messages have been changed. The -dropcr and drop_cr
6613 options are now no-ops, retained only for backwards compatibility. The
6614 following line terminators are recognized: LF CRLF CR. However, special
6615 processing applies to CR:
6617 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6618 nor a local message in the state where . is a terminator.
6620 (ii) If a bare CR is encountered in a header line, an extra space is added
6621 after the line terminator so as not to end the header. The reasoning
6622 behind this is that bare CRs in header lines are most likely either
6623 to be mistakes, or people trying to play silly games.
6625 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6626 was being incorrectly given as 18 bytes larger than it should have been.
6627 This is a VOB (very old bug).
6629 60. This may never have affected anything current, but just in case it has:
6630 When the local host is found other than at the start of a list of hosts,
6631 the local host, those with the same MX, and any that follow, are discarded.
6632 When the list in question was part of a longer list of hosts, the following
6633 hosts (not currently being processed) were also being discarded. This no
6634 longer happens. I'm not sure if this situation could ever has previously
6637 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6640 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6641 header. This is something that is recommended in a new Internet Draft, and
6642 is something that is documented as being done by Sendmail. There are two
6643 possible values. For messages generated by the autoreply transport, Exim
6646 Auto-Submitted: auto-replied
6648 whereas for all other generated messages (e.g. bounces) it adds
6650 Auto-Submitted: auto-generated
6652 63. The "personal" condition in filters now includes a test for the
6653 Auto-Submitted: header. If it contains the string "auto-" the message it
6654 not considered personal.
6656 64. Added rcpt_include_affixes as a generic transport option.
6658 65. Added queue_only_override (default true).
6660 66. Added the syslog_duplication option.
6662 67. If what should have been the first header line of a message consisted of
6663 a space followed by a colon, Exim was mis-interpreting it as a header line.
6664 It isn't of course - it is syntactically invalid and should therefore be
6665 treated as the start of the message body. The misbehaviour could have
6666 caused a number of strange effects, including loss of data in subsequent
6667 header lines, and spool format errors.
6669 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6670 client host had authenticated. This control can now be exercised by an ACL
6671 for more flexibility.
6673 69. By default, callouts do not happen when testing with -bh. There is now a
6674 variant, -bhc, which does actually run the callout code, including
6675 consulting and updating the callout cache.
6677 70. Added support for saslauthd authentication, courtesy of Alexander
6680 71. If statvfs() failed on the spool or log directories while checking their
6681 size for availability, Exim confusingly gave the error "space shortage".
6682 Furthermore, in debugging mode it crashed with a floating point exception.
6683 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6684 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6685 serious problem, Exim now writes to the main and panic logs when this
6686 happens, with details of the failure. It then refuses to accept the
6687 incoming message, giving the message "spool directory problem" or "log
6688 directory problem" with a 421 code for SMTP messages.
6690 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6691 0, 1, and 2 exist, because some OS complain for execs without them (see
6692 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6693 descriptors. However, the code omitted to check that the open succeeded,
6694 causing mysterious errors if for some reason the permissions on /dev/null
6695 got screwed. Now Exim writes a message to the main and panic logs, and
6696 bombs out if it can't open /dev/null.
6698 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6699 interact so that it is all more flexible. It is supposed to remain
6700 backwards compatible. Also added extra_local_interfaces.
6702 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6703 to bomb out with an assertion failure - to the client this appears as a
6704 connection drop. This problem occurs in the part of the code that was taken
6705 from the Samba project. Fortunately, the assertion is in a very simple
6706 function, so I have fixed this by reproducing the function inline in the
6707 one place where it is called, and arranging for authentication to fail
6708 instead of killing the process with assert().
6710 75. The SPA client code was not working when the server requested OEM rather
6711 than Unicode encoding.
6713 76. Added code to make require_files with a specific uid setting more usable in
6714 the case where statting the file as root fails - usually a non-root-mounted
6715 NFS file system. When this happens and the failure is EACCES, Exim now
6716 forks a subprocess and does the per-uid checking as the relevant uid.
6718 77. Added process_log_path.
6720 78. If log_file_path was not explicitly set, a setting of check_log_space or
6721 check_log_inodes was ignored.
6723 79. If a space check for the spool or log partitions fails, the incident is now
6724 logged. Of course, in the latter case the data may get lost...
6726 80. Added the %p formatting code to string_format() so that it can be used to
6727 print addresses in debug_print(). Adjusted all the address printing in the
6728 debugging in store.c to use %p rather than %d.
6730 81. There was a concern that a line of code in smtp_in.c could overflow a
6731 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6732 initially expressed, the concern was not well-founded, because trailing
6733 spaces are removed early. However, if the trailing spaces were followed by
6734 a NULL, they did not get removed, so the overflow was possible. Two fixes
6737 (a) I re-wrote the offending code in a cleaner fashion.
6738 (b) If an incoming SMTP command contains a NULL character, it is rejected
6741 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6742 runs initgroups(), so that if the Exim user is in any additional groups,
6743 they will be used during message reception.
6749 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob