1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
44 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
45 than -2003 did; needs libidn2 in addition to libidn.
47 JH/02 The path option on a pipe transport is now expanded before use.
49 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
50 Patch provided by "Björn", documentation fix added too.
52 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
53 missing a wire-to-host endian conversion.
55 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
56 close after a BDAT command line could be taken as a following command,
57 giving a synch failure. Fix by only checking for synch immediately
58 before acknowledging the chunk.
60 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
61 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
62 macro. Patches provided by Josh Soref.
64 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
65 Previously we did not; the RFC seems ambiguous and VRFY is not listed
66 by IANA as a service extension. However, John Klensin suggests that we
69 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
70 the dkim code may be unix-mode line endings rather than smtp wire-format
71 CRLF, so prepend a CR to any bare LF.
73 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
74 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
76 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
77 the first were themselves being wrongly included in the feed into dkim
78 processing; with most chunk sizes in use this resulted in an incorrect
79 body hash calculated value.
81 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
82 DKIM signature block, for verification. Although advised against by
83 standards it is specifically not ruled illegal.
85 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
87 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
88 missing a body hash (the bh= tag).
90 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
91 It seems that HAProxy sends the Proxy Protocol information in clear and
92 only then does a TLS startup, so do the same.
94 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
95 TCP connections (such as for Spamd) unless the daemon successfully set
96 Fast Open mode on its listening sockets. This fixes breakage seen on
97 too-old kernels or those not configured for Fast Open, at the cost of
98 requiring both directions being enabled for TFO, and TFO never being used
99 by non-daemon-related Exim processes.
101 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
102 endings, at least on the first header line. Try to canonify any that get
103 past that check, despite the cost.
105 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
106 now limited to an arbitrary five deep, while parsing addresses with the
107 strip_excess_angle_brackets option enabled.
109 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
110 instead leave the unprompted TLS handshake in socket buffer for the
111 TLS library to consume.
113 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
115 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
117 JH/16 Drop variables when they go out of scope. Memory management drops a whole
118 region in one operation, for speed, and this leaves assigned pointers
119 dangling. Add checks run only under the testsuite which checks all
120 variables at a store-reset and panics on a dangling pointer; add code
121 explicitly nulling out all the variables discovered. Fixes one known
122 bug: a transport crash, where a dangling pointer for $sending_ip_address
123 originally assigned in a verify callout, is re-used.
125 PP/06 Drop '.' from @INC in various Perl scripts.
127 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
129 PP/08 Reduce a number of compilation warnings under clang; building with
130 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
131 should be warning-free.
133 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
135 HS/01 Fix portability problems introduced by PP/08 for platforms where
136 realloc(NULL) is not equivalent to malloc() [SunOS et al].
138 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
139 chunk. This allows us to accept broken chunked messages. We need a more
140 general solution here.
142 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
143 already-broken messages in the queue.
145 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
147 JH/19 Fix reference counting bug in routing-generated-address tracking.
153 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
154 supports it and a size is available (ie. the sending peer gave us one).
156 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
157 years of being deprecated). The replacements are the ACLs
158 acl_smtp_mime and acl_not_smtp_mime.
160 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
161 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
162 or even an in-clear connection were permitted. Now, if the host lookup
163 was dnssec and dane was requested then the host is only used if the
164 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
165 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
166 if one fails this test.
167 This means that a poorly-configured remote DNS will make it incommunicado;
168 but it protects against a DNS-interception attack on it.
170 JH/04 Bug 1810: make continued-use of an open smtp transport connection
171 non-noisy when a race steals the message being considered.
173 JH/05 If main configuration option tls_certificate is unset, generate a
174 self-signed certificate for inbound TLS connections.
176 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
177 in rewrites and routers.
179 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
180 and logged a warning sing 4.83; now they are a configuration file error.
182 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
183 (lacking @domain). Apply the same qualification processing as RCPT.
185 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
187 JH/10 Support ${sha256:} applied to a string (as well as the previous
190 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
191 a cutthrough deliver is pending, as we always want to make a connection.
192 This also avoids re-routing the message when later placing the cutthrough
193 connection after a verify cache hit.
194 Do not update it with the verify result either.
196 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
197 when routing results in more than one destination address.
199 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
200 signing (which inhibits the cutthrough capability). Previously only
201 the presence of an option was tested; now an expansion evaluating as
202 empty is permissible (obviously it should depend only on data available
203 when the cutthrough connection is made).
205 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
206 the relevant preceding SMTP command did not note the pipelining mode.
208 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
209 Previously they were not counted.
211 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
212 as one having no matching records. Previously we deferred the message
213 that needed the lookup.
215 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
216 distinguished as "(=".
218 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
219 for missing MX records. Previously it only worked for missing A records.
221 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
223 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
224 after the data-go-ahead and data-ack. Patch from Jason Betts.
226 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
227 even for a "none" policy. Patch from Tony Meyer.
229 JH/22 Fix continued use of a connection for further deliveries. If a port was
230 specified by a router, it must also match for the delivery to be
233 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
234 When one of the recipients of a message was unsuitable for the connection
235 (has no matching addresses), we lost track of needing to mark it
236 deferred. As a result mail would be lost.
238 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
240 JH/25 Decoding ACL controls is now done using a binary search; the source code
241 takes up less space and should be simpler to maintain. Merge the ACL
242 condition decode tables also, with similar effect.
244 JH/26 Fix problem with one_time used on a redirect router which returned the
245 parent address unchanged. A retry would see the parent address marked as
246 delivered, so not attempt the (identical) child. As a result mail would
249 JH/27 Fix a possible security hole, wherein a process operating with the Exim
250 UID can gain a root shell. Credit to http://www.halfdog.net/ for
251 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
254 JH/28 Enable {spool,log} filesystem space and inode checks as default.
255 Main config options check_{log,spool}_{inodes,space} are now
256 100 inodes, 10MB unless set otherwise in the configuration.
258 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
259 Previously it only applied to the main-section connection policy
262 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
264 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
265 by me. Added RFC7919 DH primes as an alternative.
267 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
268 are not in the system include path.
270 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
271 GnuTLS, when a session startup failed (eg because the client disconnected)
272 Exim did stdio operations after fclose. This was exposed by a recent
273 change which nulled out the file handle after the fclose.
275 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
276 signed directly by the cert-signing cert, rather than an intermediate
277 OCSP-signing cert. This is the model used by LetsEncrypt.
279 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
281 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
282 an incoming connection.
284 HS/02 Bug 1802: Do not half-close the connection after sending a request
287 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
288 fallback to "prime256v1".
290 JH/34 SECURITY: Use proper copy of DATA command in error message.
291 Could leak key material. Remotely exploitable. CVE-2016-9963.
297 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
298 and 3.4.4 - once the server is enabled to respond to an OCSP request
299 it does even when not requested, resulting in a stapling non-aware
300 client dropping the TLS connection.
302 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
303 support variable-length bit vectors. No functional change.
305 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
306 The I= interface field on outgoing lines is now after the H= remote
307 host field, same as incoming lines. There is a separate
308 outgoing_interface log selector which allows you to disable the
311 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
312 If not running log_selector +smtp_connection the mainlog would be held
313 open indefinitely after a "too many connections" event, including to a
314 deleted file after a log rotate. Leave the per net connection logging
315 leaving it open for efficiency as that will be quickly detected by the
316 check on the next write.
318 HS/01 Bug 1671: Fix post transport crash.
319 Processing the wait-<transport> messages could crash the delivery
320 process if the message IDs didn't exist for some reason. When
321 using 'split_spool_directory=yes' the construction of the spool
322 file name failed already, exposing the same netto behaviour.
324 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
325 mime_regex ACL conditions.
327 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
328 to DSN fail messages (bounces): remote IP, remote greeting, remote response
329 to HELO, local diagnostic string.
331 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
332 log line to debug. Even when configured with a tls authenticator many
333 client connections are expected to not authenticate in this way, so
334 an authenticate fail is not an error.
336 HS/02 Add the Exim version string to the process info. This way exiwhat
337 gives some more detail about the running daemon.
339 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
340 matter for fast-change records such as DNSBLs.
342 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
343 retry record, even if constant. There may be multiple transports with
344 different interface settings and the retry behaviour needs to be kept
347 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
349 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
351 JH/10 Bug 840: fix log_defer_output option of pipe transport
353 JH/11 Bug 830: use same host for all RCPTS of a message, even under
354 hosts_randomize. This matters a lot when combined with mua_wrapper.
356 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
357 ${quote_pgsql:<string>} operator.
359 JH/13 Bug 1708: avoid misaligned access in cached lookup.
361 JH/14 Change header file name for freeradius-client. Relevant if compiling
362 with Radius support; from the Gentoo tree and checked under Fedora.
364 JH/15 Bug 1712: Introduce $prdr_requested flag variable
366 JH/16 Bug 1714: Permit an empty string as expansion result for transport
367 option transport_filter, meaning no filtering.
369 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
371 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
372 defaults to "*" (all hosts). The variable is now available when not built
373 with TLS, default unset, mainly to enable keeping the testsuite sane.
374 If a server certificate is not supplied (via tls_certificate) an error is
375 logged, and clients will find TLS connections fail on startup. Presumably
376 they will retry in-clear.
377 Packagers of Exim are strongly encouraged to create a server certificate
378 at installation time.
380 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
381 with the $config_file variable.
383 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
384 in transport context, after the attempt, and per-recipient. The latter type
385 is per host attempted. The event data is the error message, and the errno
386 information encodes the lookup type (A vs. MX) used for the (first) host,
387 and the trailing two digits of the smtp 4xx response.
389 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
390 to write to mainlog (or rejectlog, paniclog) in the window between file
391 creation and permissions/ownership being changed. Particularly affects
392 installations where exicyclog is run as root, rather than exim user;
393 result is that the running daemon panics and dies.
395 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
397 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
398 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
399 "pri" and "weight". Note that the previous implicit priority given by the
400 list order is no longer honoured.
402 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
405 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
406 by defining SUPPORT_SOCKS.
408 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
409 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
410 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
411 variables are renamed to proxy_{local,external}_{address,port}.
413 JH/27 Move Internationalisation support from Experimental to mainline, enabled
414 for a build by defining SUPPORT_I18N
416 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
417 of the query string, and make ${quote_redis:} do that quoting.
419 JH/29 Move Events support from Experimental to mainline, enabled by default
420 and removable for a build by defining DISABLE_EVENT.
422 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
424 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
425 cached by the daemon.
427 JH/32 Move Redis support from Experimental to mainline, enabled for a build
428 by defining LOOKUP_REDIS. The libhiredis library is required.
430 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
431 keys are given for lookup.
433 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
434 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
435 only supported when built with TLS support. The PolarSSL SHA routines
436 are still used when the TLS library is too old for convenient support.
438 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
439 openssl_options), for security. OpenSSL forces this from version 1.1.0
440 server-side so match that on older versions.
442 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
443 allocation for $value could be released as the expansion processing
444 concluded, but leaving the global pointer active for it.
446 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
447 and to use the domains and local_parts ACL conditions.
449 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
450 incorrectly not doubled on cutthrough transmission, hence seen as a
451 body-termination at the receiving system - resulting in truncated mails.
452 Commonly the sender saw a TCP-level error, and retransmitted the message
453 via the normal store-and-forward channel. This could result in duplicates
454 received - but deduplicating mailstores were liable to retain only the
455 initial truncated version.
457 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
459 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
461 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
462 we're in there, support oversigning also; bug 1309.
464 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
466 HS/04 Add support for keep_environment and add_environment options.
468 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
469 either intentional arithmetic overflow during PRNG, or testing config-
472 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
473 delivery resulted in actual delivery. Cancel cutthrough before DATA
476 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
477 rejects a recipient: pass the reject to the originator.
479 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
480 Many were false-positives and ignorable, but it's worth fixing the
483 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
484 for the new environment-manipulation done at startup. Move the routines
485 from being local to tls.c to being global via the os.c file.
487 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
488 an extract embedded as result-arg for a map, the first arg for extract
489 is unavailable so we cannot tell if this is a numbered or keyed
490 extraction. Accept either.
496 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
499 JH/02 The smtp transport option "multi_domain" is now expanded.
501 JH/03 The smtp transport now requests PRDR by default, if the server offers
504 JH/04 Certificate name checking on server certificates, when exim is a client,
505 is now done by default. The transport option tls_verify_cert_hostnames
506 can be used to disable this per-host. The build option
507 EXPERIMENTAL_CERTNAMES is withdrawn.
509 JH/05 The value of the tls_verify_certificates smtp transport and main options
510 default to the word "system" to access the system default CA bundle.
511 For GnuTLS, only version 3.0.20 or later.
513 JH/06 Verification of the server certificate for a TLS connection is now tried
514 (but not required) by default. The verification status is now logged by
515 default, for both outbound TLS and client-certificate supplying inbound
518 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
521 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
522 Status Notification (bounce) messages are now MIME format per RFC 3464.
523 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
524 under the control of the dsn_advertise_hosts option, and routers may
525 have a dsn_lasthop option.
527 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
528 default, modifiable by a malware= option. The list separator for
529 the options can now be changed in the usual way. Bug 68.
531 JH/10 The smtp_receive_timeout main option is now expanded before use.
533 JH/11 The incoming_interface log option now also enables logging of the
534 local interface on delivery outgoing connections.
536 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
537 if the interface and destination host and port all match.
539 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
542 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
543 Patch from Andrew Lewis.
545 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
546 now supports optional time-restrictions, weighting, and priority
547 modifiers per server. Patch originally by <rommer@active.by>.
549 JH/16 The spamd_address main option now supports a mixed list of local
550 and remote servers. Remote servers can be IPv6 addresses, and
551 specify a port-range.
553 JH/17 Bug 68: The spamd_address main option now supports an optional
554 timeout value per server.
556 JH/18 Bug 1581: Router and transport options headers_add/remove can
557 now have the list separator specified.
559 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
562 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
565 JH/21 Support for the A6 type of dns record is withdrawn.
567 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
568 rather than the verbs used.
570 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
571 from 255 to 1024 chars.
573 JH/24 Verification callouts now attempt to use TLS by default.
575 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
576 are generic router options now. The defaults didn't change.
578 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
579 Original patch from Alexander Shikoff, worked over by JH.
581 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
582 files if ZCAT_COMMAND is not executable.
584 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
586 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
588 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
589 Normally benign, it bites when the pair was led to by a CNAME;
590 modern usage is to not canonicalize the domain to a CNAME target
591 (and we were inconsistent anyway for A-only vs AAAA+A).
593 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
595 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
596 when evaluating $sender_host_dnssec.
598 JH/31 Check the HELO verification lookup for DNSSEC, adding new
599 $sender_helo_dnssec variable.
601 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
603 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
605 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
607 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
608 documented as working, but never had. Support all but $spam_report.
610 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
611 added for tls authenticator.
613 HS/03 Add perl_taintmode main config option
619 TL/01 When running the test suite, the README says that variables such as
620 no_msglog_check are global and can be placed anywhere in a specific
621 test's script, however it was observed that placement needed to be near
622 the beginning for it to behave that way. Changed the runtest perl
623 script to read through the entire script once to detect and set these
624 variables, reset to the beginning of the script, and then run through
625 the script parsing/test process like normal.
627 TL/02 The BSD's have an arc4random API. One of the functions to induce
628 adding randomness was arc4random_stir(), but it has been removed in
629 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
630 function when detected.
632 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
633 cause callback expansion.
635 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
636 syntax errors in an expansion can be treated as a string instead of
637 logging or causing an error, due to the internal use of bool_lax
638 instead of bool when processing it.
640 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
641 server certificates when making smtp deliveries.
643 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
645 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
647 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
649 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
650 Merged patch from Sebastian Wiedenroth.
652 JH/05 Fix results-pipe from transport process. Several recipients, combined
653 with certificate use, exposed issues where response data items split
654 over buffer boundaries were not parsed properly. This eventually
655 resulted in duplicates being sent. This issue only became common enough
656 to notice due to the introduction of connection certificate information,
657 the item size being so much larger. Found and fixed by Wolfgang Breyha.
659 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
660 size buffer was used, resulting in syntax errors when an expansion
663 JH/07 Add support for directories of certificates when compiled with a GnuTLS
664 version 3.3.6 or later.
666 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
667 is EXPERIMENTAL_EVENT, the main-configuration and transport options
668 both become "event_action", the variables become $event_name, $event_data
669 and $event_defer_errno. There is a new variable $verify_mode, usable in
670 routers, transports and related events. The tls:cert event is now also
671 raised for inbound connections, if the main configuration event_action
674 TL/06 In test suite, disable OCSP for old versions of openssl which contained
675 early OCSP support, but no stapling (appears to be less than 1.0.0).
677 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
678 server certificate names available under the smtp transport option
679 "tls_verify_cert_hostname" now do not permit multi-component wildcard
682 JH/10 Time-related extraction expansions from certificates now use the main
683 option "timezone" setting for output formatting, and are consistent
684 between OpenSSL and GnuTLS compilations. Bug 1541.
686 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
687 encoded parameter in the incoming message. Bug 1558.
689 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
690 include certificate info, eximon was claiming there were spoolfile
693 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
695 JH/14 Log delivery-related information more consistently, using the sequence
696 "H=<name> [<ip>]" wherever possible.
698 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
699 are problematic for Debian distribution, omit them from the release
702 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
704 JH/16 Fix string representation of time values on 64bit time_t architectures.
707 JH/17 Fix a null-indirection in certextract expansions when a nondefault
708 output list separator was used.
713 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
714 checkers that were complaining about end of non-void function with no
717 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
718 This was a regression introduced in 4.83 by another bugfix.
720 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
722 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
723 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
729 TF/01 Correctly close the server side of TLS when forking for delivery.
731 When a message was received over SMTP with TLS, Exim failed to clear up
732 the incoming connection properly after forking off the child process to
733 deliver the message. In some situations the subsequent outgoing
734 delivery connection happened to have the same fd number as the incoming
735 connection previously had. Exim would try to use TLS and fail, logging
736 a "Bad file descriptor" error.
738 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
739 utilities have not been installed.
741 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
742 temporary space as the ACL may create new global variables.
744 TL/01 LDAP support uses per connection or global context settings, depending
745 upon the detected version of the libraries at build time.
747 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
748 to extract and use the src ip:port in logging and expansions as if it
749 were a direct connection from the outside internet. PPv2 support was
750 updated based on HAProxy spec change in May 2014.
752 JH/02 Add ${listextract {number}{list}{success}{fail}}.
754 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
755 Properly escape header and check for NULL return.
757 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
760 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
762 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
763 characters in header names, implemented as a verify condition.
764 Contributed by Michael Fischer v. Mollard.
766 TL/05 Rename SPF condition results err_perm and err_temp to standardized
767 results permerror and temperror. Previous values are deprecated but
768 still accepted. In a future release, err_perm and err_temp will be
769 completely removed, which will be a backward incompatibility if the
770 ACL tests for either of these two old results. Patch contributed by
771 user bes-internal on the mailing list.
773 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
775 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
776 selectors, in both main and reject logs.
778 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
781 JH/07 Add malware type "sock" for talking to simple daemon.
783 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
785 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
786 routers/transports under cutthrough routing.
788 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
789 numbers. Touch up "bool" conditional to keep the same definition.
791 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
793 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
795 JH/12 Expand items in router/transport headers_add or headers_remove lists
796 individually rather than the list as a whole. Bug 1452.
798 Required for reasonable handling of multiple headers_ options when
799 they may be empty; requires that headers_remove items with embedded
800 colons must have them doubled (or the list-separator changed).
802 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
803 view the policy declared in the DMARC record. Currently, $dmarc_status
804 is a combined value of both the record presence and the result of the
807 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
809 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
810 dnslookup router and the smtp transport (applying to the forward
813 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
814 of ldap servers used for a specific lookup. Patch provided by Heiko
817 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
818 New variable $lookup_dnssec_authenticated for observability.
820 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
821 Patch submitted by Lars Timman.
823 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
825 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
826 Requires trusted mode and valid format message id, aborts otherwise.
827 Patch contributed by Heiko Schlichting.
829 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
830 certextract with support for various fields. Bug 1358.
832 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
833 is requested by default, modifiable by smtp transport option
836 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
837 operate on certificate variables to give certificate fingerprints
838 Also new ${sha256:cert_variable}.
840 JH/23 The PRDR feature is moved from being Experimental into the mainline.
842 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
843 Christian Aistleitner.
845 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
847 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
848 file. Patch from Wolfgang Breyha.
850 JH/25 Expand the coverage of the delivery $host and $host_address to
851 client authenticators run in verify callout. Bug 1476.
853 JH/26 Port service names are now accepted for tls_on_connect_ports, to
854 align with daemon_smtp_ports. Bug 72.
856 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
857 support and error reporting did not work properly.
859 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
860 and is readable. Patch from Andrew Colin Kissa.
862 TL/14 Enhance documentation of ${run expansion and how it parses the
863 commandline after expansion, particularly in the case when an
864 unquoted variable expansion results in an empty value.
866 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
868 PP/02 Fix internal collision of T_APL on systems which support RFC3123
869 by renaming away from it. Addresses GH issue 15, reported by
872 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
874 TL/15 SECURITY: prevent double expansion in math comparison functions
875 (can expand unsanitized data). Not remotely exploitable.
882 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
884 PP/02 Make -n do something, by making it not do something.
885 When combined with -bP, the name of an option is not output.
887 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
890 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
891 $sender_host_name and config options to manage this, and basic check
894 PP/05 DSCP support for outbound connections and control modifier for inbound.
896 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
897 (Only plugin which currently uses this is kerberos4, which nobody should
898 be using, but we should make it available and other future plugins might
899 conceivably use it, even though it would break NAT; stuff *should* be
900 using channel bindings instead).
902 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
903 name; added for Sendmail compatibility; requires admin caller.
904 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
905 just ignore it); requires trusted caller.
906 Also parse but ignore: -Ac -Am -X<logfile>
909 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
911 TL/02 Add +smtp_confirmation as a default logging option.
913 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
914 Patch by Magnus Holmgren from 2007-02-20.
916 TL/04 Bugzilla 1281 - Spec typo.
917 Bugzilla 1283 - Spec typo.
918 Bugzilla 1290 - Spec grammar fixes.
920 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
922 TL/06 Add Experimental DMARC support using libopendmarc libraries.
924 TL/07 Fix an out of order global option causing a segfault. Reported to dev
925 mailing list by by Dmitry Isaikin.
927 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
929 JH/02 Support "G" suffix to numbers in ${if comparisons.
931 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
933 NM/01 Bugzilla 1197 - Spec typo
934 Bugzilla 1196 - Spec examples corrections
936 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
938 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
939 gnutls_enable_pkcs11, but renamed to more accurately indicate its
942 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
943 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
945 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
946 "acl {{name}{arg}...}", and optional args on acl condition
949 JH/05 Permit multiple router/transport headers_add/remove lines.
951 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
953 JH/07 Avoid using a waiting database for a single-message-only transport.
954 Performance patch from Paul Fisher. Bugzilla 1262.
956 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
959 JH/09 Add $headers_added variable, with content from use of ACL modifier
960 add_header (but not yet added to the message). Bugzilla 199.
962 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
963 Pulled from Bugzilla 817 by Wolfgang Breyha.
965 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
967 (nb: this is the same fix as in Exim 4.80.1)
969 JH/11 Add A= logging on delivery lines, and a client_set_id option on
972 JH/12 Add optional authenticated_sender logging to A= and a log_selector
975 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
977 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
978 advertise SMTP AUTH mechanism to us, instead of a generic
979 protocol violation error. Also, make Exim more robust to bad
980 data from the Dovecot auth socket.
982 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
984 When a queue runner is handling a message, Exim first routes the
985 recipient addresses, during which it prunes them based on the retry
986 hints database. After that it attempts to deliver the message to
987 any remaining recipients. It then updates the hints database using
990 So if a recipient address works intermittently, it can get repeatedly
991 deferred at routing time. The retry hints record remains fresh so the
992 address never reaches the final cutoff time.
994 This is a fairly common occurrence when a user is bumping up against
995 their storage quota. Exim had some logic in its local delivery code
996 to deal with this. However it did not apply to per-recipient defers
997 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
999 This change adds a proper retry rule check during routing so that the
1000 final cutoff time is checked against the message's age. We only do
1001 this check if there is an address retry record and there is not a
1002 domain retry record; this implies that previous attempts to handle
1003 the address had the retry_use_local_parts option turned on. We use
1004 this as an approximation for the destination being like a local
1005 delivery, as in LMTP.
1007 I suspect this new check makes the old local delivery cutoff check
1008 redundant, but I have not verified this so I left the code in place.
1010 TF/02 Correct gecos expansion when From: is a prefix of the username.
1012 Test 0254 submits a message to Exim with the header
1016 When I ran the test suite under the user fanf2, Exim expanded
1017 the header to contain my full name, whereas it should have added
1018 a Resent-Sender: header. It erroneously treats any prefix of the
1019 username as equal to the username.
1021 This change corrects that bug.
1023 GF/01 DCC debug and logging tidyup
1024 Error conditions log to paniclog rather than rejectlog.
1025 Debug lines prefixed by "DCC: " to remove any ambiguity.
1027 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1029 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1030 Bug spotted by Jeremy Harris; was flawed since initial commit.
1031 Would have resulted in OCSP responses post-SNI triggering an Exim
1032 NULL dereference and crash.
1034 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1036 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1037 Bug detection, analysis and fix by Samuel Thibault.
1038 Bugzilla 1331, Debian bug #698092.
1040 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1042 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1043 Server implementation by Todd Lyons, client by JH.
1044 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1045 config variable "prdr_enable" controls whether the server
1046 advertises the facility. If the client requests PRDR a new
1047 acl_data_smtp_prdr ACL is called once for each recipient, after
1048 the body content is received and before the acl_smtp_data ACL.
1049 The client is controlled by both of: a hosts_try_prdr option
1050 on the smtp transport, and the server advertisement.
1051 Default client logging of deliveries and rejections involving
1052 PRDR are flagged with the string "PRDR".
1054 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1055 fclose(). Diagnosis by Todd Lyons.
1057 PP/17 Update configure.default to handle IPv6 localhost better.
1058 Patch by Alain Williams (plus minor tweaks).
1061 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1062 This is now consistent with GnuTLS, and is now documented: the
1063 previous undocumented portable approach to treating the option as
1064 unset was to force an expansion failure. That still works, and
1065 an empty string is now equivalent.
1067 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1068 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1069 not performing validation itself.
1071 PP/20 Added force_command boolean option to pipe transport.
1072 Patch from Nick Koston, of cPanel Inc.
1074 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1077 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1079 PP/21 Fix eximon continuous updating with timestamped log-files.
1080 Broken in a format-string cleanup in 4.80, missed when I repaired the
1081 other false fix of the same issue.
1082 Report and fix from Heiko Schlichting.
1085 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1086 Report from Prashanth Katuri.
1088 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1089 It's SecureTransport, so affects any MacOS clients which use the
1090 system-integrated TLS libraries, including email clients.
1092 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1093 using a MIME ACL for non-SMTP local injection.
1094 Report and assistance in diagnosis by Warren Baker.
1096 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1098 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1100 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1101 last id that failed so it may be referenced in subsequent ACL's.
1103 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1106 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1107 ldap library initialization, allowing self-signed CA's to be
1108 used. Also properly sets require_cert option later in code by
1109 using NULL (global ldap config) instead of ldap handle (per
1110 session). Bug diagnosis and testing by alxgomz.
1112 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1113 the src/util/ subdirectory.
1115 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1116 renamed to Transport Post Delivery Action by Jeremy Harris, as
1119 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1120 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1121 redis_servers = needs to be configured which will be used by the redis
1122 lookup. Patch from Warren Baker, of The Packet Hub.
1124 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1126 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1127 hostname or reverse DNS when processing a host list. Used suggestions
1128 from multiple comments on this bug.
1130 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1132 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1133 Missed a few lines, added it to make the runtest require no keyboard
1136 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1137 contains upper case chars. Make router use caseful_local_part.
1139 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1140 support when GnuTLS has been built with p11-kit.
1146 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1148 This, or similar/improved, will also be change PP/11 of 4.82.
1154 PP/01 Handle short writes when writing local log-files.
1155 In practice, only affects FreeBSD (8 onwards).
1156 Bugzilla 1053, with thanks to Dmitry Isaikin.
1158 NM/01 Bugzilla 949 - Documentation tweak
1160 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1163 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1165 PP/02 Implemented gsasl authenticator.
1167 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1169 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1170 `pkg-config foo` for cflags/libs.
1172 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1173 with rest of GSASL and with heimdal_gssapi.
1175 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1176 `pkg-config foo` for cflags/libs for the TLS implementation.
1178 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1179 properties get this fed in as external SSF. A number of robustness
1180 and debugging improvements to the cyrus_sasl authenticator.
1182 PP/08 cyrus_sasl server now expands the server_realm option.
1184 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1185 Patch by Jeremy Harris.
1187 PP/10 Added dbmjz lookup type.
1189 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1191 PP/12 MAIL args handles TAB as well as SP, for better interop with
1192 non-compliant senders.
1193 Analysis and variant patch by Todd Lyons.
1195 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1196 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1197 Patch from Dirk Mueller <dmueller@suse.com>
1199 PP/13 tls_peerdn now print-escaped for spool files.
1200 Observed some $tls_peerdn in wild which contained \n, which resulted
1201 in spool file corruption.
1203 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1204 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1205 or write after TLS renegotiation, which otherwise led to messages
1208 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1209 as a tracking header (ie: a signed header comes before the signature).
1210 Patch from Wolfgang Breyha.
1212 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1213 comma-sep list; embedded commas doubled.
1215 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1217 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1219 Report and patch from Dmitry Banschikov.
1221 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1222 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1223 failures. We appear to now support TLS1.1+ with Exim.
1225 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1226 lets Exim select keys and certificates based upon TLS SNI from client.
1227 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1228 before an outbound SMTP session. New log_selector, +tls_sni.
1230 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1231 NULL dereference. Report and patch from Alun Jones.
1233 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1234 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1235 Not seeing resolver debug output on NetBSD, but suspect this is a
1236 resolver implementation change.
1238 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1239 Left warnings. Added "eximon gdb" invocation mode.
1241 PP/21 Defaulting "accept_8bitmime" to true, not false.
1243 PP/22 Added -bw for inetd wait mode support.
1245 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1246 locate the relevant includes and libraries. Made this the default.
1248 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1249 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1251 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1252 This may cause build issues on older platforms.
1254 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1255 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1256 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1257 Added SNI support via GnuTLS too.
1258 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1260 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1262 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1263 Applied second patch from Janne, implementing suggestion to default
1264 multiple-strings-in-record handling to match SPF spec.
1266 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1268 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1269 read-only, out of scope).
1270 Patch from Wolfgang Breyha, report from Stuart Northfield.
1272 PP/29 Fix three issues highlighted by clang analyser static analysis.
1273 Only crash-plausible issue would require the Cambridge-specific
1274 iplookup router and a misconfiguration.
1275 Report from Marcin Mirosław.
1277 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1279 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1280 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1281 As part of this, removing so much warning spew let me fix some minor
1282 real issues in debug logging.
1284 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1285 assignment on my part. Fixed.
1287 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1288 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1289 Janne Snabb (who went above and beyond: thank you).
1291 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1292 string otherwise requires a connection and a bunch more work and it's
1293 relatively easy to get wrong. Should also expose TLS library linkage
1296 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1297 64-bit ${eval} (JH/03).
1299 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1300 GNU libc to support some of the 64-bit stuff, should not lead to
1301 conflicts. Defined before os.h is pulled in, so if a given platform
1302 needs to override this, it can.
1304 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1305 protection layer was required, which is not implemented.
1306 Bugzilla 1254, patch from Wolfgang Breyha.
1308 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1309 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1310 tls_dhparam take prime identifiers. Also unbreak combination of
1311 OpenSSL+DH_params+TLSSNI.
1313 PP/39 Disable SSLv2 by default in OpenSSL support.
1319 PP/01 Solaris build fix for Oracle's LDAP libraries.
1320 Bugzilla 1109, patch from Stephen Usher.
1322 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1324 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1327 TF/02 Fix a couple more cases where we did not log the error message
1328 when unlink() failed. See also change 4.74-TF/03.
1330 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1331 lock up or crash if it happened to be inside a call to libc when it
1332 got a SIGUSR1 from exiwhat.
1334 The SIGUSR1 handler appends the current process status to the process
1335 log which is later printed by exiwhat. It used to use the general
1336 purpose logging code to do this, but several functions it calls are
1337 not safe for signals.
1339 The new output code in the SIGUSR1 handler is specific to the process
1340 log, and simple enough that it's easy to inspect for signal safety.
1341 Removing some special cases also simplifies the general logging code.
1342 Removing the spurious timestamps from the process log simplifies
1345 TF/04 Improved ratelimit ACL condition.
1347 The /noupdate option has been deprecated in favour of /readonly which
1348 has clearer semantics. The /leaky, /strict, and /readonly update modes
1349 are mutually exclusive. The update mode is no longer included in the
1350 database key; it just determines when the database is updated. (This
1351 means that when you upgrade Exim will forget old rate measurements.)
1353 Exim now checks that the per_* options are used with an update mode that
1354 makes sense for the current ACL. For example, when Exim is processing a
1355 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1356 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1357 must specify per_mail/readonly. If you omit the update mode it defaults to
1358 /leaky where that makes sense (as before) or /readonly where required.
1360 The /noupdate option is now undocumented but still supported for
1361 backwards compatibility. It is equivalent to /readonly except that in
1362 ACLs where /readonly is required you may specify /leaky/noupdate or
1363 /strict/noupdate which are treated the same as /readonly.
1365 A useful new feature is the /count= option. This is a generalization
1366 of the per_byte option, so that you can measure the throughput of other
1367 aggregate values. For example, the per_byte option is now equivalent
1368 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1370 The per_rcpt option has been generalized using the /count= mechanism
1371 (though it's more complicated than the per_byte equivalence). When it is
1372 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1373 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1374 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1375 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1376 Note that using per_rcpt with a non-readonly update mode in more than
1377 one ACL will cause the recipients to be double-counted. (The per_mail
1378 and per_byte options don't have this problem.)
1380 The handling of very low rates has changed slightly. If the computed rate
1381 is less than the event's count (usually one) then this event is the first
1382 after a long gap. In this case the rate is set to the same as this event's
1383 count, so that the first message of a spam run is counted properly.
1385 The major new feature is a mechanism for counting the rate of unique
1386 events. The new per_addr option counts the number of different
1387 recipients that someone has sent messages to in the last time period. It
1388 behaves like per_rcpt if all the recipient addresses are different, but
1389 duplicate recipient addresses do not increase the measured rate. Like
1390 the /count= option this is a general mechanism, so the per_addr option
1391 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1392 example, measure the rate that a client uses different sender addresses
1393 with the options per_mail/unique=$sender_address. There are further
1394 details in the main documentation.
1396 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1398 TF/06 Removed a few PCRE remnants.
1400 TF/07 Automatically extract Exim's version number from tags in the git
1401 repository when doing development or release builds.
1403 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1404 Bugzilla 879. Patch from Paul Fisher.
1406 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1407 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1410 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1412 PP/05 Variable $av_failed, true if the AV scanner deferred.
1413 Bugzilla 1078. Patch from John Horne.
1415 PP/06 Stop make process more reliably on build failure.
1416 Bugzilla 1087. Patch from Heiko Schlittermann.
1418 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1419 Bugzilla 1089. Patch from Heiko Schlittermann.
1421 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1422 Bugzilla 1131. Patch from Holger Weiß.
1424 PP/09 Handle IPv6 addresses with SPF.
1425 Bugzilla 860. Patch from Wolfgang Breyha.
1427 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1429 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1432 PP/11 match_* no longer expand right-hand-side by default.
1433 New compile-time build option, EXPAND_LISTMATCH_RHS.
1434 New expansion conditions, "inlist", "inlisti".
1436 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1438 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1440 PP/14 fix log_write() format string regression from TF/03.
1441 Bugzilla 1152. Patch from Dmitry Isaikin.
1447 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1449 PP/02 Harmonised TLS library version reporting; only show if debugging.
1450 Layout now matches that introduced for other libraries in 4.74 PP/03.
1452 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1454 PP/04 New "dns_use_edns0" global option.
1456 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1459 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1460 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1462 TK/01 Updated PolarSSL code to 0.14.2.
1463 Bugzilla 1097. Patch from Andreas Metzler.
1465 PP/07 Catch divide-by-zero in ${eval:...}.
1466 Fixes bugzilla 1102.
1468 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1471 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1472 format-string attack -- SECURITY: remote arbitrary code execution.
1474 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1475 time unintentionally subject to list matching rules, letting the header
1476 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1477 arbitrary string expansion). This allowed for information disclosure.
1479 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1480 INT_MIN/-1 -- value coerced to INT_MAX.
1486 NM/01 Workaround for PCRE version dependency in version reporting
1489 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1490 This fixes portability to compilers other than gcc, notably
1491 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1493 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1494 makefiles for portability to HP-UX and POSIX correctness.
1496 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1497 Also via indented variable definition in the Makefile.
1498 (Debugging by Oliver Heesakkers).
1500 PP/02 Restore caching of spamd results with expanded spamd_address.
1501 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1503 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1504 Improves build reliability. Fix from: Frank Elsner
1506 NM/02 Fix wide character breakage in the rfc2047 coding
1507 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1509 NM/03 Allow underscore in dnslist lookups
1510 Fixes bug 1026. Patch from Graeme Fowler
1512 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1513 Code patches from Adam Ciarcinski of NetBSD.
1515 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1518 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1519 is logged, to avoid truncation. Patch from John Horne.
1521 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1522 Patch from Jakob Hirsch.
1524 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1525 SQL string expansion failure details.
1526 Patch from Andrey Oktyabrski.
1528 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1529 Patch from Simon Arlott.
1531 PP/09 New lookups functionality failed to compile on old gcc which rejects
1532 extern declarations in function scope.
1533 Patch from Oliver Fleischmann
1535 PP/10 Use sig_atomic_t for flags set from signal handlers.
1536 Check getgroups() return and improve debugging.
1537 Fixed developed for diagnosis in bug 927 (which turned out to be
1540 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1541 Patch from Mark Zealey.
1543 PP/12 Bugzilla 1056: Improved spamd server selection.
1544 Patch from Mark Zealey.
1546 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1547 Based on patch from Heiko Schlittermann.
1549 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1550 Patch from Uwe Doering, sign-off by Michael Haardt.
1552 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1553 variable declaration deep within a block. Bug and patch from
1556 PP/15 lookups-Makefile IRIX compatibility coercion.
1558 PP/16 Make DISABLE_DKIM build knob functional.
1560 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1561 Patch by Simon Arlott
1563 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1564 variable argument macros. Our copy now differs from upstream.
1570 TF/01 Failure to get a lock on a hints database can have serious
1571 consequences so log it to the panic log.
1573 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1574 controlled using the smtp_confirmation log selector.
1576 TF/03 Include the error message when we fail to unlink a spool file.
1578 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1579 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1580 for maintaining out-of-tree patches for some time.
1582 PP/01 Bugzilla 139: Documentation and portability issues.
1583 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1584 Handle per-OS dynamic-module compilation flags.
1586 PP/02 Let /dev/null have normal permissions.
1587 The 4.73 fixes were a little too stringent and complained about the
1588 permissions on /dev/null. Exempt it from some checks.
1589 Reported by Andreas M. Kirchwitz.
1591 PP/03 Report version information for many libraries, including
1592 Exim version information for dynamically loaded libraries. Created
1593 version.h, now support a version extension string for distributors
1594 who patch heavily. Dynamic module ABI change.
1596 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1597 privilege escalation vulnerability whereby the Exim run-time user
1598 can cause root to append content of the attacker's choosing to
1601 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1604 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1605 If dropping privileges for untrusted macros, we disabled normal logging
1606 on the basis that it would fail; for the Exim run-time user, this is not
1607 the case, and it resulted in successful deliveries going unlogged.
1608 Fixed. Reported by Andreas Metzler.
1614 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1615 only prepend for the Resent-* case. Fixes regression introduced in
1616 Exim 4.70 by NM/22 for Bugzilla 607.
1618 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1619 increasing numbers of administrators be bitten by this.
1621 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1623 PP/03 Bugzilla 994: added openssl_options main configuration option.
1625 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1627 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1629 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1631 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1632 without a peer certificate, leading to a segfault because of an
1633 assumption that peers always have certificates. Be a little more
1634 paranoid. Problem reported by Martin Tscholak.
1636 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1637 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1638 NB: ClamAV planning to remove STREAM in "middle of 2010".
1639 CL also introduces -bmalware, various -d+acl logging additions and
1640 more caution in buffer sizes.
1642 PP/09 Implemented reverse_ip expansion operator.
1644 PP/10 Bugzilla 937: provide a "debug" ACL control.
1646 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1648 PP/12 Bugzilla 973: Implement --version.
1650 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1652 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1654 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1656 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1657 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1658 ignore trailing whitespace.
1660 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1662 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1665 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1666 Notification from Dr Andrew Aitchison.
1668 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1669 ExtendedDetectionInfo response format.
1670 Notification from John Horne.
1672 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1675 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1676 XSL and documented dependency on system catalogs, with examples of how
1679 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1682 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1683 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1684 configuration file which is writeable by the Exim user or group.
1686 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1687 of configuration files to cover files specified with the -C option if
1688 they are going to be used with root privileges, not just the default
1691 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1692 option (effectively making it always true).
1694 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1695 files to be used while preserving root privileges.
1697 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1698 that rogue child processes cannot use them.
1700 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1701 run-time user, instead of root.
1703 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1704 Exim run-time user without dropping privileges.
1706 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1707 result string, instead of calling string_vformat() twice with the same
1710 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1711 for other users. Others should always drop root privileges if they use
1712 -C on the command line, even for a whitelisted configure file.
1714 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1716 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1722 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1723 $data_path, and $header_path variables; fixed documentation bugs and
1726 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1727 exipick to access non-standard spools, including the "frozen" queue
1730 NM/01 Bugzilla 965: Support mysql stored procedures.
1731 Patch from Alain Williams
1733 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1735 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1736 Patch from Andreas Metzler
1738 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1739 Patch from Kirill Miazine
1741 NM/05 Bugzilla 671: Added umask to procmail example.
1743 JJ/03 installed exipick 20100323.0, fixing doc bug
1745 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1746 directory. Notification and patch from Dan Rosenberg.
1748 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1750 TK/02 Improve log output when DKIM signing operation fails.
1752 MH/01 Treat the transport option dkim_domain as a colon separated
1753 list, not as a single string, and sign the message with each element,
1754 omitting multiple occurences of the same signer.
1756 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1757 Bugzilla 985, 986. Patch by Simon Arlott
1759 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1760 Patch by Simon Arlott
1762 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1763 MBX locking. Notification from Dan Rosenberg.
1769 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1771 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1773 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1775 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1777 NM/04 Bugzilla 915: Flag broken perl installation during build.
1783 TK/01 Added patch by Johannes Berg that expands the main option
1784 "spamd_address" if it starts with a dollar sign.
1786 TK/02 Write list of recipients to X-Envelope-Sender header when building
1787 the mbox-format spool file for content scanning (suggested by Jakob
1790 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1791 (http://www.dcc-servers.net/) support via dccifd. Activated by
1792 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1794 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1795 by Mark Daniel Reidel <mr@df.eu>.
1797 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1798 When building exim an external PCRE library is now needed -
1799 PCRE is a system library on the majority of modern systems.
1800 See entry on PCRE_LIBS in EDITME file.
1802 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1803 conversation. Added nologin parameter to request.
1804 Patch contributed by Kirill Miazine.
1806 TF/01 Do not log submission mode rewrites if they do not change the address.
1808 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1810 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1811 log files in place. Contributed by Roberto Lima.
1813 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1815 TF/03 Bugzilla 615: When checking the local_parts router precondition
1816 after a local_part_suffix or local_part_prefix option, Exim now
1817 does not use the address's named list lookup cache, since this
1818 contains cached lookups for the whole local part.
1820 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1821 Robert Millan. Documentation is in experimental-spec.txt.
1823 TF/04 Bugzilla 668: Fix parallel build (make -j).
1825 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1827 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1828 Patch provided by Jan Srzednicki.
1830 TF/05 Leading white space used to be stripped from $spam_report which
1831 wrecked the formatting. Now it is preserved.
1833 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1834 that they are available at delivery time.
1836 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1838 TF/08 TLS error reporting now respects the incoming_interface and
1839 incoming_port log selectors.
1841 TF/09 Produce a more useful error message if an SMTP transport's hosts
1842 setting expands to an empty string.
1844 NM/06 Bugzilla 744: EXPN did not work under TLS.
1845 Patch provided by Phil Pennock.
1847 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1848 Patch provided by Richard Godbee.
1850 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1851 acl_smtp_notquit, added index entry.
1853 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1854 Patch provided by Eugene Bujak.
1856 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1857 accept(). Patch provided by Maxim Dounin.
1859 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1860 Patch provided by Phil Pennock.
1862 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1864 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1865 Patch provided by Brad "anomie" Jorsch.
1867 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1868 Patch provided by Dean Brooks.
1870 TK/05 Add native DKIM support (does not depend on external libraries).
1872 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1873 Patch provided by Graeme Fowler.
1875 NM/16 Bugzilla 851: Documentation example syntax fix.
1877 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1879 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1882 NM/19 Bugzilla 745: TLS version reporting.
1883 Patch provided by Phil Pennock.
1885 NM/20 Bugzilla 167: bool: condition support.
1886 Patch provided by Phil Pennock.
1888 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1889 clients. Patch provided by Phil Pennock.
1891 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1892 Patch provided by Brad "anomie" Jorsch.
1894 NM/23 Bugzilla 687: Fix misparses in eximstats.
1895 Patch provided by Heiko Schlittermann.
1897 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1898 Patch provided by Heiko Schlittermann.
1900 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1901 plus update to original patch.
1903 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1905 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1906 Patch provided by David Brownlee.
1908 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1910 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1912 NM/30 Bugzilla 888: TLS documentation bugfixes.
1914 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1916 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1917 Unlike the original bugzilla I have changed all shell scripts in src tree.
1919 NM/33 Bugzilla 898: Transport filter timeout fix.
1920 Patch by Todd Rinaldo.
1922 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1923 Patch by Serge Demonchaux.
1925 NM/35 Bugzilla 39: Base64 decode bug fixes.
1926 Patch by Jakob Hirsch.
1928 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1930 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1932 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1934 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1940 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1941 ALT-N's libdkim that I have put here:
1942 http://duncanthrax.net/exim-experimental/
1944 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1945 were called 'true' and it seems that C99 defines that as a reserved
1946 keyword to be used with 'bool' variable types. That means you could
1947 not include C99-style headers which use bools without triggering
1948 build errors in sieve.c.
1950 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1951 as mailq or other aliases. Changed the --help handling significantly
1952 to do whats expected. exim_usage() emits usage/help information.
1954 SC/01 Added the -bylocaldomain option to eximstats.
1956 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1958 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1960 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1966 PH/01 Another patch from the Sieve maintainer.
1968 PH/02 When an IPv6 address is converted to a string for single-key lookup
1969 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1970 dots are used instead of colons so that keys in lsearch files need not
1971 contain colons. This was done some time before quoting was made available
1972 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1973 (notated using the quote facility) so as to distinguish them from IPv4
1974 keys. This meant that lookups for IP addresses in host lists did not work
1975 for iplsearch lookups.
1977 This has been fixed by arranging for IPv6 addresses to be expressed with
1978 colons if the lookup type is iplsearch. This is not incompatible, because
1979 previously such lookups could never work.
1981 The situation is now rather anomalous, since one *can* have colons in
1982 ordinary lsearch keys. However, making the change in all cases is
1983 incompatible and would probably break a number of configurations.
1985 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1988 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1989 conversion specification without a maximum field width, thereby enabling
1990 a rogue spamd server to cause a buffer overflow. While nobody in their
1991 right mind would setup Exim to query an untrusted spamd server, an
1992 attacker that gains access to a server running spamd could potentially
1993 exploit this vulnerability to run arbitrary code as the Exim user.
1995 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1996 $primary_hostname instead of what libspf2 thinks the hosts name is.
1998 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1999 a directory entry by the name of the lookup key. Previously, if a
2000 symlink pointed to a non-existing file or a file in a directory that
2001 Exim lacked permissions to read, a lookup for a key matching that
2002 symlink would fail. Now it is enough that a matching directory entry
2003 exists, symlink or not. (Bugzilla 503.)
2005 PH/03 The body_linecount and body_zerocount variables are now exported in the
2008 PH/04 Added the $dnslist_matched variable.
2010 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2011 This means they are set thereafter only if the connection becomes
2014 PH/06 Added the client_condition to authenticators so that some can be skipped
2015 by clients under certain conditions.
2017 PH/07 The error message for a badly-placed control=no_multiline_responses left
2018 "_responses" off the end of the name.
2020 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2022 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2023 (without spaces) instead of just copying the configuration text.
2025 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2027 PH/11 Added $max_received_linelength.
2029 PH/12 Added +ignore_defer and +include_defer to host lists.
2031 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2032 way in which PCRE > 7.0 is built.
2034 PH/14 Implemented queue_only_load_latch.
2036 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2037 MAIL command. The effect was to mangle the value on 64-bit systems.
2039 PH/16 Another patch from the Sieve maintainer.
2041 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2043 PH/18 If a system quota error occurred while trying to create the file for
2044 a maildir delivery, the message "Mailbox is full" was not appended to the
2045 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2046 only to a quota excession during the actual writing of the file.
2048 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2049 characters?) which causes problems in log lines. The DN values are now
2050 passed through string_printing() before being added to log lines.
2052 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2053 and InterBase are left for another time.)
2055 PH/21 Added message_body_newlines option.
2057 PH/22 Guard against possible overflow in moan_check_errorcopy().
2059 PH/23 POSIX allows open() to be a macro; guard against that.
2061 PH/24 If the recipient of an error message contained an @ in the local part
2062 (suitably quoted, of course), incorrect values were put in $domain and
2063 $local_part during the evaluation of errors_copy.
2069 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2070 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2073 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2074 issue a MAIL command.
2076 PH/02 In an ACL statement such as
2078 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2080 if a client was not listed at all, or was listed with a value other than
2081 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2082 the condition was not true (as it should be), so access was not denied.
2083 The bug was that the ! inversion was incorrectly passed on to the second
2084 item. This has been fixed.
2086 PH/03 Added additional dnslists conditions == and =& which are different from
2087 = and & when the dns lookup returns more than one IP address.
2089 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2090 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2092 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2093 FSYNC, which compiles an option called disable_fsync that allows for
2094 bypassing fsync(). The documentation is heavily laced with warnings.
2096 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2098 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2099 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2100 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2101 including adding "make clean"; (3) Added -fPIC when compiling the test
2102 dynamically loaded module, to get rid of a warning.
2104 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2105 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2106 = 0s. The bug is otherwise harmless.
2108 PH/07 There was a bug in the dovecot authenticator such that the value of
2109 $auth1 could be overwritten, and so not correctly preserved, after a
2110 successful authentication. This usually meant that the value preserved by
2111 the server_setid option was incorrect.
2113 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2115 PH/09 Installed PCRE release 7.0.
2117 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2118 run for batched SMTP input. It is now run at the start of every message
2119 in the batch. While fixing this I discovered that the process information
2120 (output by running exiwhat) was not always getting set for -bs and -bS
2121 input. This is fixed, and it now also says "batched" for BSMTP.
2123 PH/11 Added control=no_pipelining.
2125 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2126 patch, slightly modified), and move the expansion of helo_data till after
2127 the connection is made in the smtp transport (so it can use these
2130 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2132 PH/14 Added log_selector = +pid.
2134 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2136 PH/16 Add ${if forany and ${if forall.
2138 PH/17 Added dsn_from option to vary the From: line in DSNs.
2140 PH/18 Flush SMTP output before performing a callout, unless control =
2141 no_callout_flush is set.
2143 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2144 was true (the default) a successful delivery failed to delete the retry
2145 item, thus causing premature timeout of the address. The bug is now
2148 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2150 PH/21 Long custom messages for fakedefer and fakereject are now split up
2151 into multiline responses in the same way that messages for "deny" and
2152 other ACL rejections are.
2154 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2155 with slight modification.
2157 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2158 draft, changing the syntax and factoring some duplicate code".
2160 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2161 for deliveries of the second and subsequent messages over the same SMTP
2164 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2165 ${reduce, with only minor "tidies".
2167 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2169 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2170 expansion side effects.
2172 PH/27 When a message times out after an over-quota error from an Exim-imposed
2173 quota, the bounce message says "mailbox is full". This message was not
2174 being given when it was a system quota that was exceeded. It now should
2177 MH/03 Made $recipients available in local_scan(). local_scan() already has
2178 better access to the recipient list through recipients_list[], but
2179 $recipients can be useful in postmaster-provided expansion strings.
2181 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2182 in the case of a MAIL command with additional options following the
2183 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2184 were accidentally chopped off.
2186 PH/29 SMTP synchronization checks are implemented when a command is read -
2187 there is a check that no more input is waiting when there shouldn't be
2188 any. However, for some commands, a delay in an ACL can mean that it is
2189 some time before the response is written. In this time, more input might
2190 arrive, invalidly. So now there are extra checks after an ACL has run for
2191 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2192 pipelining has not been advertised.
2194 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2196 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2197 returned in the SMTP response when smtp_return_error_details was false.
2198 This has been fixed.
2200 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2201 instead of the C I/O that was originally supplied, because problems were
2202 reported on Solaris.
2204 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2205 Exim which did not show up earlier: it was assuming that a call to
2206 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2207 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2208 was a macro that became an assignment, so it seemed to work. This has
2209 changed to a proper function call with a void return, hence the compile
2210 error. Exim's code has been fixed.
2212 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2215 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2217 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2219 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2220 because it turns out to be too big or there is a timeout) there is no
2221 "Completed" line in the log. When some messages of this type were
2222 selected by exigrep, they were listed as "not completed". Others were
2223 picked up by some special patterns. I have improved the selection
2224 criteria to be more general.
2226 PH/38 The host_find_failed option in the manualroute router can now be set
2227 to "ignore", to completely ignore a host whose IP address cannot be
2228 found. If all hosts are ignored, the behaviour is controlled by the new
2229 host_all_ignored option.
2231 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2232 homing or because of multiple MX records with /mx) generated more than
2233 one IP address, and the following item turned out to be the local host,
2234 all the secondary addresses of the first item were incorrectly removed
2235 from the list, along with the local host and any following hosts (which
2236 is what is supposed to happen).
2238 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2239 whoever called Exim into the -H file. In the case of the daemon it was
2240 behaving confusingly. When first started, it used values for whoever
2241 started the daemon, but after a SIGHUP it used the Exim user (because it
2242 calls itself on a restart). I have changed the code so that it now always
2245 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2246 message are rejected with the same error (e.g. no authentication or bad
2247 sender address), and a DATA command is nevertheless sent (as can happen
2248 with PIPELINING or a stupid MUA), the error message that was given to the
2249 RCPT commands is included in the rejection of the DATA command. This is
2250 intended to be helpful for MUAs that show only the final error to their
2253 PH/42 Another patch from the Sieve maintainer.
2255 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2256 Eximstats - Fixed some broken HTML links and added missing column headers
2258 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2259 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2261 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2263 PH/43 Yet another patch from the Sieve maintainer.
2265 PH/44 I found a way to check for a TCP/IP connection going away before sending
2266 the response to the final '.' that terminates a message, but only in the
2267 case where the client has not sent further data following the '.'
2268 (unfortunately, this is allowed). However, in many cases there won't be
2269 any further data because there won't be any more messages to send. A call
2270 to select() can be used: if it shows that the input is "ready", there is
2271 either input waiting, or the socket has been closed. An attempt to read
2272 the next input character can distinguish the two cases. Previously, Exim
2273 would have sent an OK response which the client would never have see.
2274 This could lead to message repetition. This fix should cure that, at
2275 least in a lot of common cases.
2277 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2278 advertised in response to EHLO.
2284 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2285 fixed by 4.65/MH/01 (is this a record?) are fixed:
2287 (i) An empty string was always treated as zero by the numeric comparison
2288 operators. This behaviour has been restored.
2290 (ii) It is documented that the numeric comparison operators always treat
2291 their arguments as decimal numbers. This was broken in that numbers
2292 starting with 0 were being interpreted as octal.
2294 While fixing these problems I realized that there was another issue that
2295 hadn't been noticed. Values of message_size_limit (both the global option
2296 and the transport option) were treated as octal if they started with 0.
2297 The documentation was vague. These values are now always treated as
2298 decimal, and I will make that clear in the documentation.
2304 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2305 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2308 MH/01 Don't check that the operands of numeric comparison operators are
2309 integers when their expansion is in "skipping" mode (fixes bug
2310 introduced by 4.64-PH/07).
2312 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2313 child addresses, Exim now panics and dies. Previously, because the count
2314 is held in a short int, deliveries were likely to be lost. As such a
2315 large number of recipients for a single message is ridiculous
2316 (performance will be very, very poor), I have chosen to impose a limit
2317 rather than extend the field.
2323 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2324 leftover -K file (the existence of which was triggered by #402).
2325 While we were at it, introduced process PID as part of the -K
2326 filename. This should rule out race conditions when creating
2329 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2330 processing considerably. Previous code took too long for large mails,
2331 triggering a timeout which in turn triggers #401.
2333 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2334 in the DK code in transports.c. sendfile() is not really portable,
2335 hence the _LINUX specificness.
2337 TF/01 In the add_headers option to the mail command in an Exim filter,
2338 there was a bug that Exim would claim a syntax error in any
2339 header after the first one which had an odd number of characters
2342 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2343 callout verification, Exim cached a "reject" for the entire domain. This
2344 is correct for most verifications, but it is not correct for a recipient
2345 verification with use_sender or use_postmaster set, because in that case
2346 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2347 case of MAIL FROM:<> rejection from other early rejections (e.g.
2348 rejection of HELO). When verifying a recipient using a non-null MAIL
2349 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2350 Whatever the result of the callout, the value of the domain cache is
2351 left unchanged (for any other kind of callout, getting as far as trying
2352 RCPT means that the domain itself is ok).
2354 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2357 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2358 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2359 session, the daemon ignores ECONNECT errors and logs others; it now
2360 ignores EPIPE as well.
2362 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2363 (quoted-printable decoding).
2365 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2366 later the small subsequent patch to fix an introduced bug.
2368 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2370 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2372 PH/08 An error is now given if message_size_limit is specified negative.
2374 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2375 to be given (somewhat) arbitrary names.
2377 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2380 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2381 miscellaneous code fixes
2383 PH/10 Added the log_reject_target ACL modifier to specify where to log
2386 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2387 hostname. This is wrong, because it relates to the incoming message (and
2388 probably the interface on which it is arriving) and not to the outgoing
2389 callout (which could be using a different interface). This has been
2390 changed to use the value of the helo_data option from the smtp transport
2391 instead - this is what is used when a message is actually being sent. If
2392 there is no remote transport (possible with a router that sets up host
2393 addresses), $smtp_active_hostname is used.
2395 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2396 tweaks were necessary in order to get it to work (see also 21 below):
2397 (a) The code assumed that strncpy() returns a negative number on buffer
2398 overflow, which isn't the case. Replaced with Exim's string_format()
2400 (b) There were several signed/unsigned issues. I just did the minimum
2401 hacking in of casts. There is scope for a larger refactoring.
2402 (c) The code used strcasecmp() which is not a standard C function.
2403 Replaced with Exim's strcmpic() function.
2404 (d) The code set only $1; it now sets $auth1 as well.
2405 (e) A simple test gave the error "authentication client didn't specify
2406 service in request". It would seem that Dovecot has changed its
2407 interface. Fortunately there's a specification; I followed it and
2408 changed what the client sends and it appears to be working now.
2410 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2413 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2414 address A is aliased to B and C, where B exists and C does not. Without
2415 -v the output is "A verified" because verification stops after a
2416 successful redirection if more than one address is generated. However,
2417 with -v the child addresses are also verified. Exim was outputting "A
2418 failed to verify" and then showing the successful verification for C,
2419 with its parentage. It now outputs "B failed to verify", showing B's
2420 parentage before showing the successful verification of C.
2422 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2423 look up a TXT record in a specific list after matching in a combined
2426 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2427 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2428 they consult the DNS. I had assumed they would set it the way they
2429 wanted; and indeed my experiments on Linux seem to show that in some
2430 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2431 To be on the safe side, however, I have now made the interface to
2432 host_find_byname() similar to host_find_bydns(), with an argument
2433 containing the DNS resolver options. The host_find_byname() function now
2434 sets these options at its start, just as host_find_bydns() does. The smtp
2435 transport options dns_qualify_single and dns_search_parents are passed to
2436 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2437 of host_find_byname() use the default settings of RES_DEFNAMES
2438 (qualify_single) but not RES_DNSRCH (search_parents).
2440 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2441 spool_read_header() do less string testing, by means of a preliminary
2442 switch on the second character of optional "-foo" lines. (This is
2443 overdue, caused by the large number of possibilities that now exist.
2444 Originally there were few.) While I was there, I also converted the
2445 str(n)cmp tests so they don't re-test the leading "-" and the first
2446 character, in the hope this might squeeze out yet more improvement.
2448 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2449 flag allowing group syntax was set by the header_syntax check but not
2450 turned off, possible causing trouble later; (2) The flag was not being
2451 set at all for the header_verify test, causing "group"-style headers to
2452 be rejected. I have now set it in this case, and also caused header_
2453 verify to ignore an empty address taken from a group. While doing this, I
2454 came across some other cases where the code for allowing group syntax
2455 while scanning a header line wasn't quite right (mostly, not resetting
2456 the flag correctly in the right place). These bugs could have caused
2457 trouble for malformed header lines. I hope it is now all correct.
2459 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2460 with the "reply" argument non-NULL. The code, however (which originally
2461 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2462 but it didn't always do it. This confused somebody who was copying the
2463 code for some other use. I have removed all the tests.
2465 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2466 feature that was used to support insecure browsers during the U.S. crypto
2467 embargo. It requires special client support, and Exim is probably the
2468 only MTA that supported it -- and would never use it because real RSA is
2469 always available. This code has been removed, because it had the bad
2470 effect of slowing Exim down by computing (never used) parameters for the
2471 RSA_EXPORT functionality.
2473 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2474 authenticator to fail if there's a tab character in the incoming data
2475 (there should never be unless someone is messing about, as it's supposed
2476 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2477 if the connection is using TLS or if the remote IP is the same as the
2478 local IP, and the "valid-client-cert option" if a client certificate has
2481 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2482 authenticators. This can be used for authorization after authentication
2483 succeeds. (In the case of plaintext, it servers for both authentication
2486 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2487 if any retry times were supplied.
2489 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2490 connection, where there is no client IP address to check. In this
2491 situation, the verify now always succeeds.
2493 PH/25 Applied John Jetmore's -Mset patch.
2495 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2497 PH/27 In a string expansion for a processed (not raw) header when multiple
2498 headers of the same name were present, leading whitespace was being
2499 removed from all of them, but trailing whitespace was being removed only
2500 from the last one. Now trailing whitespace is removed from each header
2501 before concatenation. Completely empty headers in a concatenation (as
2502 before) are ignored.
2504 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2505 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2507 PH/29 [Removed. This was a change that I later backed out, and forgot to
2508 correct the ChangeLog entry (that I had efficiently created) before
2509 committing the later change.]
2511 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2512 address errors (4xx response to RCPT) over the same connection as other
2513 messages routed to the same hosts. Such deliveries are always "forced",
2514 so retry times are not inspected. This resulted in far too many retries
2515 for the affected addresses. The effect occurred only when there were more
2516 hosts than the hosts_max_try setting in the smtp transport when it had
2517 the 4xx errors. Those hosts that it had tried were not added to the list
2518 of hosts for which the message was waiting, so if all were tried, there
2519 was no problem. Two fixes have been applied:
2521 (i) If there are any address or message errors in an SMTP delivery, none
2522 of the hosts (tried or untried) are now added to the list of hosts
2523 for which the message is waiting, so the message should not be a
2524 candidate for sending over the same connection that was used for a
2525 successful delivery of some other message. This seems entirely
2526 reasonable: after all the message is NOT "waiting for some host".
2527 This is so "obvious" that I'm not sure why it wasn't done
2528 previously. Hope I haven't missed anything, but it can't do any
2529 harm, as the worst effect is to miss an optimization.
2531 (ii) If, despite (i), such a delivery is accidentally attempted, the
2532 routing retry time is respected, so at least it doesn't keep
2533 hammering the server.
2535 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2536 in ${readsocket because some servers need this prod.
2538 PH/32 Added some extra debug output when updating a wait-xxx database.
2540 PH/33 The hint "could be header name not terminated by colon", which has been
2541 given for certain expansion errors for a long time, was not being given
2542 for the ${if def:h_colon_omitted{... case.
2544 PH/34 The spec says: "With one important exception, whenever a domain list is
2545 being scanned, $domain contains the subject domain." There was at least
2546 one case where this was not true.
2548 PH/35 The error "getsockname() failed: connection reset by peer" was being
2549 written to the panic log as well as the main log, but it isn't really
2550 panic-worthy as it just means the connection died rather early on. I have
2551 removed the panic log writing for the ECONNRESET error when getsockname()
2554 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2555 runs only) independently of the message's sender address. This meant
2556 that, if the 4xx error was in fact related to the sender, a different
2557 message to the same recipient with a different sender could confuse
2558 things. In particular, this can happen when sending to a greylisting
2559 server, but other circumstances could also provoke similar problems.
2560 I have changed the default so that the retry time for these errors is now
2561 based a combination of the sender and recipient addresses. This change
2562 can be overridden by setting address_retry_include_sender=false in the
2565 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2566 remote server are returned as part of bounce messages. This was not
2567 happening for LMTP over a pipe (the lmtp transport), but now it is the
2568 same for both kinds of LMTP.
2570 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2571 in header lines that were in fact CNAMEs. This is no longer the case.
2573 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2574 and queue runs started by the daemon processed all messages. This has
2575 been fixed so that -R and -S can now usefully be given with -q<time>.
2577 PH/40 Import PCRE release 6.7 (fixes some bugs).
2579 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2581 PH/42 Give an error if -q is specified more than once.
2583 PH/43 Renamed the variables $interface_address and $interface_port as
2584 $received_ip_address and $received_port, to make it clear that these
2585 values apply to message reception, and not to the outgoing interface when
2586 a message is delivered. (The old names remain recognized, of course.)
2588 PH/44 There was no timeout on the connect() call when using a Unix domain
2589 socket in the ${readsocket expansion. There now is.
2591 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2592 be meaningful with "accept".
2594 SC/01 Eximstats V1.43
2595 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2597 SC/02 Eximstats V1.44
2598 Use a glob alias rather than an array ref in the generated
2599 parser. This improves both readability and performance.
2601 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2602 Collect SpamAssassin and rejection statistics.
2603 Don't display local sender or destination tables unless
2604 there is data to show.
2605 Added average volumes into the top table text output.
2607 SC/04 Eximstats V1.46
2608 Collect data on the number of addresses (recipients)
2609 as well as the number of messages.
2611 SC/05 Eximstats V1.47
2612 Added 'Message too big' to the list of mail rejection
2613 reasons (thanks to Marco Gaiarin).
2615 SC/06 Eximstats V1.48
2616 Mainlog lines which have GMT offsets and are too short to
2617 have a flag are now skipped.
2619 SC/07 Eximstats V1.49 (Alain Williams)
2620 Added the -emptyok flag.
2622 SC/08 Eximstats V1.50
2623 Fixes for obtaining the IP address from reject messages.
2625 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2626 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2627 whitespace changes from 4.64-PH/27
2629 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2632 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2633 are found, allow negative numbers in numeric criteria)
2635 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2637 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2638 to match changes made in 4.64-PH/43
2640 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2642 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2643 "domain" is actually an IP address, and force a failure. This locks out
2644 those revolvers/nameservers that support "A-for-A" lookups, in
2645 contravention of the specifications.
2647 PH/48 When a host name was looked up from an IP address, and the subsequent
2648 forward lookup of the name timed out, the host name was left in
2649 $sender_host_name, contrary to the specification.
2651 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2652 restricted to single-key lookups, Exim was not diagnosing an error if
2653 * or *@ was used with a query-style lookup.
2655 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2657 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2658 long ago, but noone interested enough thought of it. Let's just say that
2659 the "1.1" means that there are some new functions that weren't there at
2660 some point in the past.
2662 PH/51 Error processing for expansion failure of helo_data from an smtp
2663 transport during callout processing was broken.
2665 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2666 tested/used via the -bh/-bhc/-bs options.
2668 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2669 bug, fixed in subsequent PCRE releases).
2671 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2672 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2674 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2680 SC/01 Use a glob alias rather than an array ref in eximstats generated
2681 parser. This improves both readability and performance.
2683 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2684 Don't display local sender or destination tables in eximstats unless
2685 there is data to show.
2686 Added average volumes into the eximstats top table text output.
2688 SC/03 Collect data on the number of addresses (recipients) as well
2689 as the number of messages in eximstats.
2691 TF/01 Correct an error in the documentation for the redirect router. Exim
2692 does (usually) call initgroups() when daemonizing.
2694 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2695 with consistent privilege compared to when running as a daemon.
2697 TF/03 Note in the spec that $authenticated_id is not set for local
2698 submissions from trusted users.
2700 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2701 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2703 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2704 by adding some example configuration directives to the default
2705 configuration file. A little bit of work is required to uncomment the
2706 directives and define how usernames and passwords are checked, but
2707 there is now a framework to start from.
2709 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2710 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2711 without this. I don't know how relevant this is to other LDAP libraries.
2713 PH/02 Add the verb name to the "unknown ACL verb" error.
2715 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2717 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2719 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2720 directory not expanded when it should be if an expanded home directory
2721 was set for the address (which is overridden by the transport).
2723 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2726 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2727 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2728 because it is too late at that time, and has no effect.
2730 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2731 security issue with \' (bugzilla #107). I could not use the
2732 PQescapeStringConn() function, because it needs a PGconn value as one of
2735 PH/08 When testing addresses using -bt, indicate those final addresses that
2736 are duplicates that would not cause an additional delivery. At least one
2737 person was confused, thinking that -bt output corresponded to deliveries.
2738 (Suppressing duplicates isn't a good idea as you lose the information
2739 about possibly different redirections that led to the duplicates.)
2741 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2742 systems where poll() doesn't work, in particular OS X.
2744 PH/10 Added more information to debugging output for retry time not reached.
2746 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2747 operations in malware.c.
2749 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2752 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2753 syslog_duplication set false, log lines that would normally be written
2754 both the the main log and to the reject log were not written to syslog at
2757 PH/14 In the default configuration, change the use of "message" in ACL warn
2758 statements to "add_header".
2760 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2761 not followed by a command (e.g. "seen endif").
2763 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2764 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2767 PH/17 Added extra conditions to the default value of delay_warning_condition
2771 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2772 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2773 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2776 The Auto-Submitted: and various List- headers are standardised, whereas I
2777 don't think Precedence: ever was.
2779 PH/18 Refactored debugging code in route_finduser() to show more information,
2780 in particular, the error code if getpwnam() issues one.
2782 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2783 This is apparently needed in addition to the PH/07 change above to avoid
2784 any possible encoding problems.
2786 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2787 but not after initializing Perl.
2789 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2790 output them only if debugging. By default they are written stderr,
2791 apparently, which is not desirable.
2793 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2796 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2799 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2801 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2802 authenticated or an ident call has been made. Suppress the default
2803 values for $authenticated_id and $authenticated_sender (but permit -oMai
2804 and -oMas) when testing with -bh.
2806 PH/24 Re-jigged the order of the tests in the default configuration so that the
2807 tests for valid domains and recipients precede the DNS black list and CSA
2808 tests, on the grounds that those ones are more expensive.
2810 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2811 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2812 This bug exists in every version of Exim that I still have, right back to
2815 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2816 However, an attempt to turn on case-sensitivity in a regex key by
2817 including (?-i) didn't work because the subject string was already
2818 lowercased, and the effects were non-intuitive. It turns out that a
2819 one-line patch can be used to allow (?-i) to work as expected.
2825 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2826 other effects) broke the use of negated acl sub-conditions.
2828 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2831 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2832 "Deny" causes Exim to reject the incoming connection with a 554 error.
2833 Unfortunately, if there is a major crisis, such as a disk failure,
2834 tcp-wrappers gives "deny", whereas what one would like would be some
2835 kind of temporary error. A kludge has been added to help with this.
2836 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2837 554 error is used if errno is still zero or contains ENOENT (which occurs
2838 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2841 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2843 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2844 errors. Otherwise a message that provokes a temporary error (when other
2845 messages do not) can cause a whole host to time out.
2847 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2848 addresses were routed directly to files or pipes from a redirect router.
2849 File deliveries just didn't batch; pipe deliveries might have suffered
2852 PH/06 A failure to get a lock for a hints database would erroneously always say
2853 "Failed to get write lock", even when it was really a read lock.
2855 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2856 of 0600. This has been changed to use the value of the lockfile_mode
2857 option (which defaults to 0600).
2859 PH/08 Applied small patch from the Sieve maintainer.
2861 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2862 folder from quota calculations, a direct delivery into this folder messed
2863 up the contents of the maildirsize file. This was because the regex was
2864 used only to exclude .Trash (or whatever) when the size of the mailbox
2865 was calculated. There was no check that a delivery was happening into an
2866 excluded directory. This bug has been fixed by ignoring all quota
2867 processing for deliveries into excluded directories.
2869 PH/10 Added the maildirfolder_create_regex option to appendfile.
2875 PH/01 The code for finding all the local interface addresses on a FreeBSD
2876 system running IPv6 was broken. This may well have applied to all BSD
2877 systems, as well as to others that have similar system calls. The broken
2878 code found IPv4 interfaces correctly, but gave incorrect values for the
2879 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2880 that it would not match correctly against @[] and not recognize the IPv6
2883 PH/02 The ipliteral router was not recognizing addresses of the form user@
2884 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2886 PH/03 Added disable_ipv6.
2888 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2889 decoded form, because it is most often used to construct To: headers
2890 lines in autoreplies, and the decoded form may well be syntactically
2891 invalid. However, $reply_address has leading white space removed, and all
2892 newlines turned into spaces so that the autoreply transport does not
2895 PH/05 If group was specified without a user on a router, and no group or user
2896 was specified on a transport, the group from the router was ignored.
2898 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2899 for visible compile-time settings that can be used to change these
2900 numbers, for those that want even more. Backwards compatibility with old
2901 spool files has been maintained. However, going back to a previous Exim
2902 release will lost any variables that are in spool files.
2904 PH/07 Two small changes when running in the test harness: increase delay when
2905 passing a TCP/IP connection to a new process, in case the original
2906 process has to generate a bounce, and remove special handling of
2907 127.0.0.2 (sic), which is no longer necessary.
2909 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2910 be the same on different OS.
2912 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2915 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2916 whether --show-vars was specified or not
2918 JJ/02 exipick: Added support for new ACL variable spool format introduced
2921 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2922 syntactically invalid From: or Reply-to: line, and a filter used this to
2923 generate an autoreply, and therefore failed to obtain an address for the
2924 autoreply, Exim could try to deliver to a non-existent relative file
2925 name, causing unrelated and misleading errors. What now happens is that
2926 it logs this as a hard delivery error, but does not attempt to create a
2929 PH/11 The exinext utility has a -C option for testing purposes, but although
2930 the given file was scanned by exinext itself; it wasn't being passed on
2931 when Exim was called.
2933 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2934 an end-of-file indication when reading a command response.
2936 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2937 compiled. In many other places in Exim, IPv6 addresses are always
2938 recognized, so I have changed this. It also means that IPv4 domain
2939 literals of the form [IPV4:n.n.n.n] are now always recognized.
2941 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2942 used if the router is not running as root, for example, when verifying at
2943 ACL time, or when using -bh. The debugging output from this situation was
2944 non-existent - all you got was a failure to exec. I have made two
2947 (a) Failures to set uid/gid, the current directory, or a process leader
2948 in a subprocess such as that created by queryprogram now generate
2949 suitable debugging output when -d is set.
2951 (b) The queryprogram router detects when it is not running as root,
2952 outputs suitable debugging information if -d is set, and then runs
2953 the subprocess without attempting to change uid/gid.
2955 PH/15 Minor change to Makefile for building test_host (undocumented testing
2958 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2959 additional section of a DNS packet that returns MX or SRV records.
2960 Instead, it always explicitly searches for A/AAAA records. This avoids
2961 major problems that occur when a DNS server includes only records of one
2962 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2963 fixed another bug: if SRV records were looked up and the corresponding
2964 address records were *not* found in the additional section, the port
2965 values from the SRV records were lost.
2967 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2968 using the correct key (the original address) when searching the retry
2969 rules in order to find which one to use for generating the retry hint.
2971 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2972 adding the default one. Similarly, if it contains a Reply-To: header, the
2973 errors_reply_to option, if set, is not used.
2975 PH/19 When calculating a retry time, Exim used to measure the "time since
2976 failure" by looking at the "first failed" field in the retry record. Now
2977 it does not use this if it is later than than the arrival time of the
2978 message. Instead it uses the arrival time. This makes for better
2979 behaviour in cases where some deliveries succeed, thus re-setting the
2980 "first failed" field. An example is a quota failure for a huge message
2981 when small messages continue to be delivered. Without this change, the
2982 "time since failure" will always be short, possible causing more frequent
2983 delivery attempts for the huge message than are intended.
2984 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2986 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2987 $1, $2, $3) because the numerical variables can be reset during some
2988 expansion items (e.g. "match"), thereby losing the authentication data.
2990 PH/21 Make -bV show the size of off_t variables so that the test suite can
2991 decide whether to run tests for quotas > 2G.
2993 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2994 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2995 filecount value is greater than 2G or if a quota value is greater than 2G
2996 on a system where the size of off_t is not greater than 4, a panic error
2999 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3000 never match. The debug and -bh output now contains an explicit error
3001 message indicating a malformed IPv4 address or mask.
3003 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3004 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3005 PH/23 above applies.
3007 PH/25 Do not write to syslog when running in the test harness. The only
3008 occasion when this arises is a failure to open the main or panic logs
3009 (for which there is an explicit test).
3011 PH/26 Added the /no_tell option to "control=freeze".
3013 PH/27 If a host name lookup failed very early in a connection, for example, if
3014 the IP address matched host_lookup and the reverse lookup yielded a name
3015 that did not have a forward lookup, an error message of the form "no IP
3016 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3017 could be logged. Now it outputs the IP address instead of "NULL".
3019 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3020 allows the sender and the authenticated sender to be set when
3021 submitting a message from within Exim. Since child_open_exim() is
3022 documented for local_scan(), the new function should be too.
3024 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3025 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3026 results in an empty string is now treated as unset.
3028 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3030 PH/31 Added sender_verify_fail logging option.
3032 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3033 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3034 tidied the source and removed it altogether.
3036 PH/33 When a queue run was abandoned because the load average was too high, a
3037 log line was always written; now it is written only if the queue_run log
3038 selector is set. In addition, the log line for abandonment now contains
3039 information about the queue run such as the pid. This is always present
3040 in "start" and "stop" lines but was omitted from the "abandon" line.
3042 PH/34 Omit spaces between a header name and the colon in the error message that
3043 is given when verify = headers_syntax fails (if there are lots of them,
3044 the message gets confusing).
3046 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3047 names, as there are now some PTR records that contain slashes. This check
3048 is only to protect against broken name servers that fall over on strange
3049 characters, so the fact that it applies to all lookups doesn't matter.
3051 PH/36 Now that the new test suite is complete, we can remove some of the
3052 special code in Exim that was needed for the old test suite. For example,
3053 sorting DNS records because real resolvers return them in an arbitrary
3054 order. The new test suite's fake resolver always returns records in the
3057 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3058 bounces) except when queue_only is set, to avoid logging races between
3059 the different processes.
3061 PH/38 Panic-die if .include specifies a non-absolute path.
3063 PH/39 A tweak to the "H" retry rule from its user.
3065 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3066 a label. They prevented compilation on older perls.
3068 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3069 a warning to be raised on newish perls.
3071 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3072 on queue. Changes to match documented behaviour of showing count of
3073 messages matching specified criteria.
3075 PH/40 Changed the default ident timeout from 30s to 5s.
3077 PH/41 Added support for the use of login_cap features, on those BSD systems
3078 that have them, for controlling the resources used by pipe deliveries.
3080 PH/42 The content-scanning code uses fopen() to create files in which to put
3081 message data. Previously it was not paying any attention to the mode of
3082 the files. Exim runs with umask(0) because the rest of the code creates
3083 files with open(), and sets the required mode explicitly. Thus, these
3084 files were ending up world-writeable. This was not a big issue, because,
3085 being within the spool directory, they were not world-accessible. I have
3086 created a function called modefopen, which takes an additional mode
3087 argument. It sets umask(777), creates the file, chmods it to the required
3088 mode, then resets the umask. All the relevant calls to fopen() in the
3089 content scanning code have been changed to use this function.
3091 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3092 to 24 hours. This avoids potential overflow problems when processing G
3093 and H retry rules. I suspect nobody ever tinkers with this value.
3095 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3097 PH/45 When the plaintext authenticator is running as a client, the server's
3098 challenges are checked to ensure they are valid base64 strings. By
3099 default, the authentication attempt is cancelled if an invalid string is
3100 received. Setting client_ignore_invalid_base64 true ignores these errors.
3101 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3102 they are received. Thus, the responses can be made to depend on the
3103 challenges. If an invalid string is ignored, an empty string is placed in
3106 PH/46 Messages that are created by the autoreply transport now contains a
3107 References: header, in accordance with RFCs 2822 and 3834.
3109 PH/47 Added authenticated_sender_force to the smtp transport.
3111 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3113 PH/49 Installed latest patch from the Sieve maintainer.
3115 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3116 was also set, the appendfile transport was unnecessarily scanning a
3117 directory of message files (e.g. for maildir delivery) to find the count
3118 of files (along with the size), even though it did not need this
3119 information. It now does the scan only if it needs to find either the
3120 size of the count of files.
3122 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3124 PH/52 Two bugs concerned with error handling when the smtp transport is
3127 (i) Exim was not creating retry information for temporary errors given
3128 for individual recipients after the DATA command when the smtp transport
3129 was used in LMTP mode. This meant that they could be retried too
3130 frequently, and not timed out correctly.
3132 (ii) Exim was setting the flag that allows error details to be returned
3133 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3134 recipients that were returned after the DATA command.
3136 PH/53 This is related to PH/52, but is more general: for any failing address,
3137 when detailed error information was permitted to be returned to the
3138 sender, but the error was temporary, then after the final timeout, only
3139 "retry timeout exceeded" was returned. Now it returns the full error as
3140 well as "retry timeout exceeded".
3142 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3143 do this, and (what is worse) MTAs that accept it.
3145 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3146 will now be deprecated.
3148 PH/56 New os.c-cygwin from the Cygwin maintainer.
3150 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3151 formats (previously only available in exim formats via -bpr, -bpru,
3152 and -bpra. Now also available in native and exiqgrep formats)
3154 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3155 with very large, slow to parse queues
3157 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3159 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3161 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3162 responses to authentication challenges, though it was showing the
3163 challenges; (ii) I've removed the CR characters from the debug output for
3166 PH/58 Allow for the insertion of a newline as well as a space when a string
3167 is turned into more than one encoded-word during RFC 2047 encoding. The
3168 Sieve code now uses this.
3170 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3171 data_4xx, lost_connection, tls_required.
3173 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3174 message was being sent as an SMTP response.
3176 PH/61 Add -l and -k options to exicyclog.
3178 PH/62 When verifying, if an address was redirected to one new address, so that
3179 verification continued, and the new address failed or deferred after
3180 having set something in $address_data, the value of $address_data was not
3181 passed back to the ACL. This was different to the case when no
3182 redirection occurred. The value is now passed back in both cases.
3184 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3185 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3186 use login_cap.h, so on its own it isn't the distinguishing feature. The
3187 new name refers directly to the setclassresources() function.
3189 PH/65 Added configuration files for NetBSD3.
3191 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3193 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3194 is preferred over IPv4.
3196 PH/68 The bounce_return_message and bounce_return_body options were not being
3197 honoured for bounces generated during the reception of non-SMTP messages.
3198 In particular, this applied to messages rejected by the ACL. This bug has
3199 been fixed. However, if bounce_return_message is true and bounce_return_
3200 body is false, the headers that are returned for a non-SMTP message
3201 include only those that have been read before the error was detected.
3202 (In the case of an ACL rejection, they have all been read.)
3204 PH/69 The HTML version of the specification is now built in a directory called
3205 spec_html instead of spec.html, because the latter looks like a path with
3206 a MIME-type, and this confuses some software.
3208 PH/70 Catch two compiler warnings in sieve.c.
3210 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3211 function verify_get_ident() calls ip_connect() to connect a socket, but
3212 if the "connect()" function timed out, ip_connect() used to close the
3213 socket. However, verify_get_ident() also closes the socket later, and in
3214 between Exim writes to the log, which may get opened at this point. When
3215 the socket was closed in ip_connect(), the log could get the same file
3216 descriptor number as the socket. This naturally causes chaos. The fix is
3217 not to close the socket in ip_connect(); the socket should be closed by
3218 the function that creates it. There was only one place in the code where
3219 this was missing, in the iplookup router, which I don't think anybody now
3220 uses, but I've fixed it anyway.
3222 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3223 well as to direct DNS lookups. Otherwise the handling of names in host
3224 lists is inconsistent and therefore confusing.
3230 PH/01 Two changes to the default runtime configuration:
3232 (1) Move the checks for relay_from_hosts and authenticated clients from
3233 after to before the (commented out) DNS black list checks.
3235 (2) Add control=submission to the relay_from_hosts and authenticated
3236 clients checks, on the grounds that messages accepted by these
3237 statements are most likely to be submissions.
3239 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3241 (1) Generate an error if the third argument for the ${prvs expansion is
3244 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3247 (3) Reset the variables that are obtained from the first argument of
3248 ${prvscheck and used in the second argument before leaving the code,
3249 because their memory is reclaimed, so using them afterwards may do
3252 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3253 one (it's much easier than Tom thought :-).
3255 (5) Because of (4), we can now allow for the use of $prvscheck_result
3256 inside the third argument.
3258 PH/03 For some reason, the default setting of PATH when running a command from
3259 a pipe transport was just "/usr/bin". I have changed it to
3262 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3263 anything to be listed in the output from -bV.
3265 PH/05 When a filter generated an autoreply, the entire To: header line was
3266 quoted in the delivery log line, like this:
3268 => >A.N.Other <ano@some.domain> <original@ddress> ...
3270 This has been changed so that it extracts the operative address. There
3271 may be more than one such address. If so, they are comma-separated, like
3274 => >ano@some.domain,ona@other.domain <original@ddress> ...
3276 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3277 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3278 not being looked up in the rDNS to get a host name, Exim was showing the
3279 IP address twice in Received: lines, even though the IP addresses were
3280 identical. For example:
3282 Received: from [1.2.3.4] (helo=[1.2.3.4])
3284 However, if the real host name was known, it was omitting the HELO data
3285 if it matched the actual IP address. This has been tidied up so that it
3286 doesn't show the same IP address twice.
3288 PH/07 When both +timestamp and +memory debugging was on, the value given by
3289 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3290 called by the debug printing, thereby overwriting the timestamp buffer.
3291 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3293 PH/08 When the original message was included in an autoreply transport, it
3294 always said "this is a copy of the message, including all the headers",
3295 even if body_only or headers_only was set. It now gives an appropriate
3298 PH/09 Applied a patch from the Sieve maintainer which:
3300 o fixes some comments
3301 o adds the (disabled) notify extension core
3302 o adds some debug output for the result of if/elsif tests
3303 o points to the current vacation draft in the documentation
3304 and documents the missing references header update
3308 o fixes a bug in processing the envelope test (when testing
3309 multiple envelope elements, the last element determined the
3312 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3313 Electronic Mail") by including:
3315 Auto-submitted: auto-generated
3317 in the messages that it generates (bounce messages and others, such as
3318 warnings). In the case of bounce messages for non-SMTP messages, there was
3319 also a typo: it was using "Auto_submitted" (underscore instead of
3320 hyphen). Since every message generated by Exim is necessarily in response
3321 to another message, thes have all been changed to:
3323 Auto-Submitted: auto-replied
3325 in accordance with these statements in the RFC:
3327 The auto-replied keyword:
3329 - SHOULD be used on messages sent in direct response to another
3330 message by an automatic process,
3332 - MUST NOT be used on manually-generated messages,
3334 - MAY be used on Delivery Status Notifications (DSNs) and Message
3335 Disposition Notifications (MDNs),
3337 - MUST NOT be used on messages generated by automatic or periodic
3338 processes, except for messages which are automatic responses to
3341 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3342 to the default Received: header definition.
3344 PH/12 Added log selector acl_warn_skipped (default on).
3346 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3347 variables because (a) they are in the wrong storage pool and (b) even if
3348 they were copied, it wouldn't work properly because of the caching.
3350 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3351 checking when decoding. Apparently there are clients that generate
3352 overlong encoded strings. Why am I not surprised?
3354 PH/15 If the first argument of "${if match_address" was not empty, but did not
3355 contain an "@" character, Exim crashed. Now it writes a panic log message
3356 and treats the condition as false.
3358 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3360 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3361 extension "envelope-auth". The code is finished and in agreement with
3362 other implementations, but there is no documentation so far and in fact,
3363 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3364 not changing the active code.
3366 Print executed "if" and "elsif" statements when debugging is used. This
3367 helps a great deal to understand what a filter does.
3369 Document more things not specified clearly in RFC3028. I had all this
3370 sorted out, when out of a sudden new issues came to my mind. Oops."
3372 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3375 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3376 canonical form (as documented). However, after a host name lookup from
3377 the IP address, check_host() was doing a simple string comparison with
3378 addresses acquired from the DNS when checking that the found name did
3379 have the original IP as one of its addresses. Since any found IPv6
3380 addresses are likely to be in abbreviated form, the comparison could
3381 fail. Luckily, there already exists a function for doing the comparison
3382 by converting both addresses to binary, so now that is used instead of
3383 the text comparison.
3385 PH/20 There was another similar case to PH/19, when a complete host name was
3386 given in a host list; looking up its IP address could give an abbreviated
3387 form, whereas the current host's name might or might not be abbreviated.
3388 The same fix has been applied.
3394 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3395 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3398 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3399 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3401 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3403 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3404 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3405 Also, for "Auto-Submitted", treat anything other than "no" as
3406 non-personal, in accordance with RFC 3834. (Previously it treated
3407 anything starting "auto-" as non-personal.)
3409 TF/01 The control=submission/name=... option had a problem with syntax
3410 errors if the name included a slash character. The /name= option
3411 now slurps the rest of the string, so it can include any characters
3412 but it must come last in the list of options (after /sender_retain
3415 PH/05 Some modifications to the interface to the fake nameserver for the new
3423 TK/01 Added the "success_on_redirect" address verification option. See
3424 NewStuff for rationale and an example.
3426 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3428 PH/02 Patch to exigrep to allow it to work on syslog lines.
3430 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3431 fread() to skip over the body file's header line, because in Cygwin the
3432 header line is locked and is inaccessible.
3434 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3435 co-exist for some time) to make it clear that it is the Exim ID that is
3436 referenced, not the Message-ID: header line.
3438 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3439 string_format() function, because snprintf() does not exist on all
3442 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3443 ${stat: expansion item.
3445 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3446 protocol synchronization error", to keep the pedants happy.
3448 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3449 well as for IRIX systems, when gcc is being used. See the host.c source
3452 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3454 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3457 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3458 transport and to the smtp transport in LMTP mode.
3460 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3462 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3463 run a filter in a subprocess. This could lead to confusion in subsequent
3464 lookups in the parent process. There should also be a search_tidyup() at
3465 the end of the subprocess.
3467 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3468 only if the host matched helo_try_verify_hosts, which caused the
3469 verification to occur when the EHLO/HELO command was issued. The ACL just
3470 tested the remembered result. Now, if a previous verification attempt has
3471 not happened, "verify = helo" does it there and then.
3473 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3475 TK/03 Fix log output including CR from clamd.
3477 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3478 exist provoked a memory error which could cause a segfault.
3480 PH/15 Installed PCRE 6.2
3482 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3484 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3485 of the problem. Specifically, suggested +O2 rather than +O1 for the
3488 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3490 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3491 record, the port was not correctly logged when the outgoing_port log
3492 selector was set (it logged the transort's default port).
3494 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3495 fallback_hosts, and "hosts" in the smtp transport.
3497 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3498 host errors such as "Connection refused".
3500 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3501 authentication with radiusclient 0.4.9:
3503 - Error returned from rc_read_config was caught wrongly
3504 - Username/password not passed on to radius server due to wrong length.
3506 The presumption is that some radiusclient API changes for 4.51/PH/17
3507 were not taken care of correctly. The code is still untested by me (my
3508 Linux distribution still has 0.3.2 of radiusclient), but it was
3509 contributed by a Radius user.
3511 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3512 expanding the "port" option of the smtp transport.
3514 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3515 while reading a MIME header. Thanks to Tom Hughes for a patch.
3517 PH/24 Include config.h inside local_scan.h so that configuration settings are
3520 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3521 that in an ACL for RCPT (for example), you can examine exactly what was
3524 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3525 commands, but it was not correctly comparing the address with the actual
3526 client host address. Thus, it would show the EHLO address in Received:
3527 header lines when this was not necessary.
3529 PH/27 Added the % operator to ${eval:}.
3531 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3532 it should be ignoring failures (because with -C, for example, it has lost
3533 privilege). It wasn't ignoring creation failures other than "already
3536 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3539 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3540 because an input line was too long, either on its own, or by virtue of
3541 too many continuations, the temporary file was not being removed, and the
3542 return code was incorrect.
3544 PH/31 Missing "BOOL" in function definition in filtertest.c.
3546 PH/32 Applied Sieve patches from the maintainer.
3548 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3550 PH/33 Added "verify = not_blind".
3552 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3553 Local/Makefile (with some defaults set). These are used in built scripts
3554 such as exicyclog, but they have never been used in the exim_install
3555 script (though there are many overriding facilities there). I have
3556 arranged that the exim_install script now takes note of these two
3559 PH/35 Installed configuration files for Dragonfly.
3561 PH/36 When a locally submitted message by a trusted user did not contain a
3562 From: header, and the sender address was obtained from -f or from an SMTP
3563 MAIL command, and the trusted user did not use -F to supply a sender
3564 name, $originator_name was incorrectly used when constructing a From:
3565 header. Furthermore, $originator_name was used for submission mode
3566 messages from external hosts without From: headers in a similar way,
3567 which is clearly wrong.
3569 PH/37 Added control=suppress_local_fixups.
3571 PH/38 When log_selector = +received_sender was set, and the addition of the
3572 sender made the log line's construction buffer exactly full, or one byte
3573 less than full, an overflow happened when the terminating "\n" was
3576 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3577 when the result of a list match is failure because a DNS lookup failed.
3579 PH/40 RM_COMMAND is now used in the building process.
3581 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3582 the "build-* directories that it finds.
3584 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3585 address in a domain literal was a prefix of an interface address.
3587 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3588 when verifying a sender address, unless rewrite_headers is false.
3590 PH/44 Wrote a long comment about why errors_to addresses are verified as
3591 recipients, not senders.
3593 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3594 the ratelimit ACL was added.
3596 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3598 PH/46 Added extra information about PostgreSQL errors to the error string.
3600 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3601 avoiding the need to install special zones in a real server. This is
3602 backwards compatible; if it can't find the fake resolver, it drops back.
3603 Thus, both old and new test suites can be run.
3605 TF/02 Added util/ratelimit.pl
3607 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3608 clock is set back in time.
3610 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3611 Candler <B.Candler@pobox.com>.
3613 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3614 OK for addresses that are the result of redirections.
3616 PH/48 A number of further additions for the benefit of the new test suite,
3617 including a fake gethostbyname() that interfaces to the fake DNS resolver
3620 TF/06 The fix for widen_domains has also been applied to qualify_single and
3621 search_parents which are the other dnslookup options that can cause
3624 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3627 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3629 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3630 service" if it encountered an MX record with an empty target hostname.
3631 The message is now "an MX or SRV record indicated no SMTP service".
3633 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3634 if the DNS of the sending site is misconfigured. This is quite a
3635 common situation. This change restores the behaviour of treating a
3636 helo verification defer as a failure.
3638 PH/51 If self=fail was set on a router, the bounce message did not include the
3639 actual error message.
3645 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3647 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3648 command itself ended in error, the underlying message about the transport
3649 filter timeout was being overwritten with the pipe command error. Now the
3650 underlying error message should be appended to the second error message.
3652 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3654 PH/02 Reduce the amount of output that "make" produces by default. Full output
3655 can still be requested.
3657 PH/03 The warning log line about a condition test deferring for a "warn" verb
3658 was being output only once per connection, rather than after each
3659 occurrence (because it was using the same function as for successful
3660 "warn" verbs). This seems wrong, so I have changed it.
3662 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3663 it should not have, which might have caused a crash in the right
3664 circumstances, but probably never did.
3666 PH/04 Installed a modified version of Tony Finch's patch to make submission
3667 mode fix the return path as well as the Sender: header line, and to
3668 add a /name= option so that you can make the user's friendly name appear
3671 TF/03 Added the control = fakedefer ACL modifier.
3673 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3674 Mark Lowes for thorough testing.
3676 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3678 TK/03 Merged latest SRS patch from Miles Wilton.
3680 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3681 with the definition in sysexits.h (which is #included earlier).
3682 Fortunately, Exim does not actually use EX_OK. The code used to try to
3683 preserve the sysexits.h value, by assuming that macro definitions were
3684 scanned for macro replacements. I have been disabused of this notion,
3685 so now the code just undefines EX_OK before #including unistd.h.
3687 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3688 in the smtp transport. When a block could not be written in a single
3689 write() function, the timeout was being re-applied to each part-write.
3690 This seems wrong - if the receiver was accepting one byte at a time it
3691 would take for ever. The timeout is now adjusted when this happens. It
3692 doesn't have to be particularly precise.
3694 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3695 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3697 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3698 without a domain if the check to <postmaster@domain> fails.
3700 SC/01 Eximstats: added -xls and the ability to specify output files
3701 (patch written by Frank Heydlauf).
3703 SC/02 Eximstats: use FileHandles for outputting results.
3705 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3707 SC/04 Eximstats: fixed display of large numbers with -nvr option
3709 SC/05 Eximstats: fixed merging of reports with empty tables.
3711 SC/06 Eximstats: added the -include_original_destination flag
3713 SC/07 Eximstats: removed tabs and trailing whitespace.
3715 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3717 TK/06 MBOX spool code: Add real "From " MBOX separator line
3718 so the .eml file is really in mbox format (even though
3719 most programs do not really care). Patch from Alex Miller.
3721 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3722 The latter is generated from $received_to and is only set if the
3723 message has one envelope recipient. SA can use these headers,
3724 obviously out-of-the-box. Patch from Alex Miller.
3726 PH/08 The ${def test on a variable was returning false if the variable's
3727 value was "0", contrary to what the specification has always said!
3728 The result should be true unless the variable is empty.
3730 PH/09 The syntax error of a character other than { following "${if
3731 def:variable_name" (after optional whitespace) was not being diagnosed.
3732 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3733 accidental colon was present, for example, could give incorrect results.
3735 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3736 result is used (not including appendfile, where other changes are about
3739 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3740 This involved changing a lot of size variables from int to off_t. It
3741 should work with maildirs and everything.
3743 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3744 spamd dying while we are connected to it.
3746 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3749 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3750 function for BDB 4.3.
3752 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3754 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3755 the sources for PCRE have been re-arranged and more files are now
3758 PH/15 The code I had for printing potentially long long variables in PH/11
3759 above was not the best (it lost precision). The length of off_t variables
3760 is now inspected at build time, and an appropriate printing format (%ld
3761 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3762 to be "long long int" or "long int". This is needed for the internal
3763 formatting function string_vformat().
3765 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3766 the configuration file to be ":syslog", then the script "guesses" where
3767 the logs files are, rather than using the compiled in default. In our
3768 case the guess is not the same as the compiled default, so the script
3769 suddenly stopped working when I started to use syslog. The patch checks
3770 to see if log_file_path is "". If so, it attempts to read it from exim
3771 with no configuration file to get the compiled in version, before it
3772 falls back to the previous guessing code."
3774 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3775 implementing BATV in an Exim configuration. See NewStuff for the gory
3778 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3779 Makefile that are specific to HP-UX.
3781 PH/18 If the "use_postmaster" option was set for a recipient callout together
3782 with the "random" option, the postmaster address was used as the MAIL
3783 FROM address for the random test, but not for the subsequent recipient
3784 test. It is now used for both.
3786 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3787 patch removes a few documentation additions to RFC 3028, because the
3788 latest draft now contains them. It adds the new en;ascii-case comparator
3789 and a new error check for 8bit text in MIME parts. Comparator and
3790 require names are now matched exactly. I enabled the subaddress
3791 extension, but it is not well tested yet (read: it works for me)."
3793 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3794 rework some of the code of TK/09 above to avoid the hardwired use of
3795 "%lld" and "long long". Replaced the call to snprintf() with a call to
3798 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3799 records point to non-existent hosts", "retry timeout exceeded", and
3800 "retry time not reached for any host after a long failure period".
3802 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3803 experimental DomainKeys support:
3805 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3806 (2) On an error such as an illegally used "control", the wrong name for
3807 the control was given.
3809 These problems did NOT occur unless DomainKeys support was compiled.
3811 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3813 PH/24 Added ${if match_ip condition.
3815 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3816 databases so that it will be absolutely obvious if a crash occurs in the
3817 DB library. This is a regular occurrence (often caused by mis-matched
3820 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3821 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3822 picked up on a user's system that detects such things. There doesn't seem
3823 to be a gcc warning option for this - only an attribute that has to be
3824 put on the function's prototype. It seems that in Fedora Core 4 they have
3825 set this on a number of new functions. No doubt there will be more in due
3828 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3829 specify a transport. However, if an address that was verified by such a
3830 router was the subject of a callout, Exim crashed because it tried to
3831 read the rcpt_include_affixes from the non-existent transport. Now it
3832 just assumes that the setting of that option is false. This bug was
3833 introduced by 4.51/PH/31.
3835 PH/28 Changed -d+all to exclude +memory, because that information is very
3836 rarely of interest, but it makes the output a lot bigger. People tend to
3837 do -d+all out of habit.
3839 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3840 code in os-type was giving problems when libc.so lives in lib64, like on
3843 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3844 aren't the modern standard, and it seems that some systems' include files
3845 don't always have them. Exim was already checking for some of the newer
3846 ones like T_AAAA, and defining it itself. I've added checks for all the
3847 record types that Exim uses.
3849 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3850 not automatically generating a new one, as it is supposed to. This
3851 prevented TLS from working. If the file did exist, but contained invalid
3852 data, a new version was generated, as expected. It was only the case of a
3853 non-existent file that was broken.
3855 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3856 with a change in libdomainkeys > 0.64.
3858 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3859 from DNS. If the selector record carries the flag, it now has
3860 precedence over the domain-wide flag.
3862 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3864 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3865 the use of a port name that isn't defined in /etc/services) occurred, the
3866 message was deferred as in a normal delivery, and thus remained on the
3867 spool, instead of being failed because of the mua_wrapper setting. This
3868 is now fixed, and I tidied up some of the mua_wrapper messages at the
3871 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3872 the messages in a hash of arrays rather than using individual hashes.
3873 This is a bit cleaner and results in dramatic memory savings, albeit
3874 at a slight CPU cost.
3876 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3877 as requested by Marc Sherman.
3879 SC/10 Eximstats: added histograms for user specified patterns as requested
3882 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3884 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3885 fopen() in the content-scanning modules that did not already have it.
3891 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3892 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3894 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3896 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3898 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3899 to test Sieve filters that use "vacation".
3901 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3902 that changes the way the GnuTLS parameters are stored in the cache file.
3903 The new format can be generated externally. For backward compatibility,
3904 if the data in the cache doesn't make sense, Exim assumes it has read an
3905 old-format file, and it generates new data and writes a new file. This
3906 means that you can't go back to an older release without removing the
3909 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3910 work if there are any delivery delays because "one_time" forces the
3911 parent to be marked "delivered", so its unseen clone is never tried
3912 again. For this reason, Exim now forbids the simultaneous setting of
3915 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3916 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3917 entry if you want to know the details. The fix, however, neglected to
3918 consider the case when local delivery batching is involved. The test for
3919 "previously delivered" was not happening when checking to see if an
3920 address could be batched with a previous (undelivered) one; under
3921 certain circumstances this could lead to multiple deliveries to the same
3924 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3925 in its include files, and this causes problems building Exim.
3927 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3928 header_syntax) but Exim was just ignoring anything given after a slash.
3929 In particular, this caused confusion with an attempt to use "verify =
3930 reverse_host_lookup/defer_ok". An error is now given when options are
3931 supplied for verify items that do not have them. (Maybe reverse_host_
3932 lookup should have a defer_ok option, but that's a different point.)
3934 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3935 defined by RFC 821) to 2048, because there were problems with some AUTH
3936 commands, and RFC 1869 says the size should be increased for extended
3937 SMTP commands that take arguments.
3939 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3942 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3943 "unknown" error; now it says that the functionality isn't in the binary.
3945 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3946 an address' error message when a string expansion fails (syntax or
3947 whatever). Otherwise the password may appear in the log. Following change
3948 PH/42 below, there is no longer a chance of it appearing in a bounce
3951 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3953 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3954 of its addresses was ever tried. (Bugzilla bug #2.)
3956 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3957 the result incorrectly in the debug output. (It correctly added a newline
3958 to what was transported.)
3960 TF/01 Added $received_time.
3962 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3963 commented out examples of how to interface to a virus scanner and to
3964 SpamAssassin. Also added commented examples of av_scanner and
3965 spamd_address settings.
3967 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3968 and controls are allowed in which ACLs. There were a couple of minor
3969 errors. Some of the entries in the conditions table (which is a table of
3970 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3971 negation of where the condition IS allowed.
3973 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3975 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3976 header file does not have a version number, so I've had to invent a new
3977 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3978 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3979 radiusclient), but it was contributed by a Radius user.
3981 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3982 files or directories, for OpenSSL.
3984 PH/19 When an Exim process that is running as root has to create an Exim log
3985 file, it does so in a subprocess that runs as exim:exim so as to get the
3986 ownership right at creation (otherwise, other Exim processes might see
3987 the file with the wrong ownership). There was no test for failure of this
3988 fork() call, which would lead to the process getting stuck as it waited
3989 for a non-existent subprocess. Forks do occasionally fail when resources
3990 run out. I reviewed all the other calls to fork(); they all seem to check
3993 PH/20 When checking for unexpected SMTP input at connect time (before writing
3994 the banner), Exim was not dealing correctly with a non-positive return
3995 from the read() function. If the client had disconnected by this time,
3996 the result was a log entry for a synchronization error with an empty
3997 string after "input=" when read() returned zero. If read() returned -1
3998 (an event I could not check), uninitialized data bytes were printed.
3999 There were reports of junk text (parts of files, etc) appearing after
4002 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4004 PH/22 Added support for macro redefinition, and (re)definition in between
4005 driver and ACL definitions.
4007 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4008 forgetting to use the resulting value; it was using the unexpanded value.
4010 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4011 hadn't been configured. The fix is from Juergen Kreileder, who
4012 understands it better than I do:
4014 "Here's what I see happening with three configured cyrus_sasl
4015 authenticators configured (plain, login, cram-md5):
4017 On startup auth_cyrus_sasl_init() gets called for each of these.
4018 This means three calls to sasl_listmech() without a specified mech_list.
4019 => SASL tests which mechs of all available mechs actually work
4020 => three warnings about OTP not working
4021 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4023 With the patch, sasl_listmech() also gets called three times. But now
4024 SASL's mech_list option is set to the server_mech specified in the the
4025 authenticator. Or in other words, the answer from sasl_listmech()
4026 gets limited to just the mech you're testing for (which is different
4028 => the return list contains just 'plain' or 'login', 'cram-md5' or
4029 nothing depending on the value of ob->server_mech.
4031 I've just tested the patch: Authentication still works fine,
4032 unavailable mechs specified in the exim configuration are still
4033 caught, and the auth.log warnings about OTP are gone."
4035 PH/25 When debugging is enabled, the contents of the command line are added
4036 to the debugging output, even when log_selector=+arguments is not
4039 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4040 answer is "GNU", and only if the return is "GNU/something" is the answer
4043 PH/27 $acl_verify_message is now set immediately after the failure of a
4044 verification in an ACL, and so is available in subsequent modifiers. In
4045 particular, the message can be preserved by coding like this:
4047 warn !verify = sender
4048 set acl_m0 = $acl_verify_message
4050 Previously, $acl_verify_message was set only while expanding "message"
4051 and "log_message" when a very denied access.
4053 PH/28 Modified OS/os.c-Linux with
4055 -#ifndef OS_LOAD_AVERAGE
4056 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4058 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4059 nomenclature these days.)
4061 PH/29 Installed patch from the Sieve maintainer that adds the options
4062 sieve_useraddress and sieve_subaddress to the redirect router.
4064 PH/30 In these circumstances:
4065 . Two addresses routed to the same list of hosts;
4066 . First host does not offer TLS;
4067 . First host accepts first address;
4068 . First host gives temporary error to second address;
4069 . Second host offers TLS and a TLS session is established;
4070 . Second host accepts second address.
4071 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4072 and peerdn, if requested) that were in fact used only for the second
4075 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4076 attention to any local part prefix or suffix that was matched by the
4077 router that accepted the address. It now behaves in the same way as it
4078 does for delivery: the affixes are removed from the local part unless
4079 rcpt_include_affixes is set on the transport.
4081 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4082 timeout during the DATA phase of an incoming message.
4084 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4085 applied a patch sanctioned by the Sieve maintainer.
4087 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4088 the uid or gid is negative. A case of a negative gid caused this to be
4089 noticed. The fix allows for either to be negative.
4091 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4092 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4095 PH/36 The change PH/12 above was broken. Fixed it.
4097 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4098 the grounds that routing the same address twice would always produce the
4099 same answer. This might have been true once, but it is certainly no
4100 longer true now. Routing a child address may depend on the previous
4101 routing that produced that child. Some complicated redirection strategies
4102 went wrong when messages had multiple recipients, and made Exim's
4103 behaviour dependent on the order in which the addresses were given.
4105 I have moved the duplicate checking until after the routing is complete.
4106 Exim scans the addresses that are assigned to local and remote
4107 transports, and removes any duplicates. This means that more work will be
4108 done, as duplicates will always all be routed, but duplicates are
4109 presumably rare, so I don't expect this is of any significance.
4111 For deliveries to pipes, files, and autoreplies, the duplicate checking
4112 still happens during the routing process, since they are not going to be
4115 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4116 It corrects a timeout issue with spamd. This is Ian's comment: "The
4117 background is that sometimes spamd either never reads data from a
4118 connection it has accepted, or it never writes response data. The exiscan
4119 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4120 blindly assumes that writes won't block so it may never time out."
4122 PH/39 Allow G after quota size as well as K and M.
4124 PH/40 The value set for $authenticated_id in an authenticator may not contain
4125 binary zeroes or newlines because the value is written to log lines and
4126 to spool files. There was no check on this. Now the value is run through
4127 the string_printing() function so that such characters are converted to
4128 printable escape sequences.
4130 PH/41 $message_linecount is a new variable that contains the total number of
4131 lines in the message. Compare $body_linecount, which is the count for the
4134 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4135 bounce and delay warning messages, except in certain special cases, which
4138 (a) An SMTP error message from a remote host;
4139 (b) A message specified in a :fail: redirection;
4140 (c) A message specified in a "fail" command in a system filter;
4141 (d) A message specified in a FAIL return from the queryprogram router;
4142 (e) A message specified by the cannot_route_message router option.
4144 In these cases only, Exim does include the error details in bounce and
4145 warning messages. There are also a few cases where bland messages such
4146 as "unrouteable address" or "local delivery error" are given.
4148 PH/43 $value is now also set for the "else" part of a ${run expansion.
4150 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4151 being worked on, but at least Exim now implements the latest version to
4154 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4155 process to complete was treated as a delivery failure, a timeout while
4156 writing the message to the pipe was logged, but erroneously treated as a
4157 successful delivery. Such timeouts include transport filter timeouts. For
4158 consistency with the overall process timeout, these timeouts are now
4159 treated as errors, giving rise to delivery failures by default. However,
4160 there is now a new Boolean option for the pipe transport called
4161 timeout_defer, which, if set TRUE, converts the failures into defers for
4162 both kinds of timeout. A transport filter timeout is now identified in
4165 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4166 systems where "make" and "gmake" are different, calling "gmake" at top
4167 level broke things. I've arranged for the value of $(MAKE) to be passed
4168 from the Makefile to this script so that it can call the same version of
4172 A note about Exim versions 4.44 and 4.50
4173 ----------------------------------------
4175 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4176 changes of various kinds. As a consequence, a big documentation update was
4177 needed. This delayed the release for rather longer than seemed good, especially
4178 in the light of a couple of (minor) security issues. Therefore, the changes
4179 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4180 release. So 4.44 and 4.50 are in effect two different branches that both start
4183 I have left the 4.50 change log unchanged; it contains all the changes since
4184 4.43. The change log for 4.44 is below; many of its items are identical to
4185 those for 4.50. This seems to be the most sensible way to preserve the
4186 historical information.
4192 1. Minor wording change to the doc/README.SIEVE file.
4194 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4195 computation of the current number of files was incorrect.
4197 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4198 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4201 4. Give more explanation in the error message when the command for a transport
4202 filter fails to execute.
4204 5. There are several places where Exim runs a non-Exim command in a
4205 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4206 was being done only for the command run by the queryprogram router. It is
4207 now done for all such subprocesses. The other cases are: ${run, transport
4208 filters, and the commands run by the lmtp and pipe transports.
4210 6. Added CONFIGURE_GROUP build-time option.
4212 7. Some older OS have a limit of 256 on the maximum number of file
4213 descriptors. Exim was using setrlimit() to set 1000 as a large value
4214 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4215 systems. I've change it so that if it can't get 1000, it tries for 256.
4217 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4218 was an oversight, and furthermore, ever since the addition of extra
4219 controls (e.g. 4.43/32), the checks on when to allow different forms of
4220 "control" were broken. There should now be diagnostics for all cases when a
4221 control that does not make sense is encountered.
4223 9. Added the /retain_sender option to "control=submission".
4225 10. $recipients is now available in the predata ACL (oversight).
4227 11. Tidy the search cache before the fork to do a delivery from a message
4228 received from the command line. Otherwise the child will trigger a lookup
4229 failure and thereby defer the delivery if it tries to use (for example) a
4230 cached ldap connection that the parent has called unbind on.
4232 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4233 of $address_data from the recipient verification was clobbered by the
4234 sender verification.
4236 13. The value of address_data from a sender verification is now available in
4237 $sender_address_data in subsequent conditions in the ACL statement.
4239 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4241 15. Added a new option "connect=<time>" to callout options, to set a different
4244 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4245 was its contents. (It was OK if the option was not defined at all.)
4247 17. A "Completed" log line is now written for messages that are removed from
4248 the spool by the -Mrm option.
4250 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4251 information about exactly what failed.
4253 19. Added -dd to debug only the daemon process.
4255 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4256 handles timeouts, both on the server side and network timeouts. Renamed the
4257 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4259 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4260 It is now set to "smtps".
4262 22. $host_address is now set to the target address during the checking of
4263 ignore_target_hosts.
4265 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4266 being passed; this would have caused $sender_host_name to have been used if
4267 matching the list had actually called for a host name (not very likely,
4268 since this list is usually IP addresses). A host name is now passed as
4271 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4272 code that specifies a non-restarting handler (typically sigaction() in
4273 modern systems) in an attempt to fix a rare and obscure crash bug.
4275 25. Narrowed the window for a race in the daemon that could cause it to ignore
4276 SIGCHLD signals. This is not a major problem, because they are used only to
4277 wake it up if nothing else does.
4279 26. A malformed maildirsize file could cause Exim to calculate negative values
4280 for the mailbox size or file count. Odd effects could occur as a result.
4281 The maildirsize information is now recalculated if the size or filecount
4284 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4285 support for a long time. Removed HAVE_SYS_VFS_H.
4287 28. Installed the latest version of exipick from John Jetmore.
4289 29. In an address list, if the pattern was not a regular expression, an empty
4290 subject address (from a bounce message) matched only if the pattern was an
4291 empty string. Non-empty patterns were not even tested. This was the wrong
4292 because it is perfectly reasonable to use an empty address as part of a
4293 database query. An empty address is now tested by patterns that are
4294 lookups. However, all the other forms of pattern expect the subject to
4295 contain a local part and a domain, and therefore, for them, an empty
4296 address still always fails if the pattern is not itself empty.
4298 30. Exim went into a mad DNS loop when attempting to do a callout where the
4299 host was specified on an smtp transport, and looking it up yielded more
4300 than one IP address.
4302 31. Re-factored the code for checking spool and log partition space into a
4303 function that finds that data and another that does the check. The former
4304 is then used to implement four new variables: $spool_space, $log_space,
4305 $spool_inodes, and $log_inodes.
4307 32. The RFC2047 encoding function was originally intended for short strings
4308 such as real names; it was not keeping to the 75-character limit for
4309 encoded words that the RFC imposes. It now respects the limit, and
4310 generates multiple encoded words if necessary. To be on the safe side, I
4311 have increased the buffer size for the ${rfc2047: expansion operator from
4314 33. It is now permitted to omit both strings after an "if" condition; if the
4315 condition is true, the result is "true". As before, when the second string
4316 is omitted, a false condition yields an empty string. This makes it less
4317 cumbersome to write custom ACL and router conditions.
4319 34. Failure to deliver a bounce message always caused it to be frozen, even if
4320 there was an errors_to setting on the router. The errors_to setting is now
4323 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4324 canonical form (fully expanded) before being placed in
4325 $sender_host_address.
4327 36. The table in the code that translates DNS record types into text (T_A to
4328 "A" for instance) was missing entries for NS and CNAME. It is just possible
4329 that this could have caused confusion if both these types were looked up
4330 for the same domain, because the text type is used as part of Exim's
4331 per-process caching. But the chance of anyone hitting this buglet seems
4334 37. The dnsdb lookup has been extended in a number of ways.
4336 (1) There is a new type, "zns", which walks up the domain tree until it
4337 finds some nameserver records. It should be used with care.
4339 (2) There is a new type, "mxh", which is like "mx" except that it returns
4340 just the host names, not the priorities.
4342 (3) It is now possible to give a list of domains (or IP addresses) to be
4343 looked up. The behaviour when one of the lookups defers can be
4344 controlled by a keyword.
4346 (4) It is now possible to specify the separator character for use when
4347 multiple records are returned.
4349 38. The dnslists ACL condition has been extended: it is now possible to supply
4350 a list of IP addresses and/or domains to be looked up in a particular DNS
4353 39. Added log_selector=+queue_time_overall.
4355 40. When running the queue in the test harness, wait just a tad after forking a
4356 delivery process, to get repeatability of debugging output.
4358 41. Include certificate and key file names in error message when GnuTLS fails
4359 to set them up, because the GnuTLS error message doesn't include the name
4360 of the failing file when there is a problem reading it.
4362 42. Allow both -bf and -bF in the same test run.
4364 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4366 44. The "Exiscan patch" is now merged into the mainline Exim source.
4368 45. Sometimes the final signoff response after QUIT could fail to get
4369 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4370 < 0 before doing a fflush(). This bug looks as though it goes back to the
4371 introduction of TLS in release 3.20, but "sometimes" must have been rare
4372 because the tests only now provoked it.
4374 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4375 (this can affect the format of dates).
4377 47. exim_tidydb, when checking for the continued existence of a message for
4378 which it has found a message-specific retry record, was not finding
4379 messages that were in split spool directories. Consequently, it was
4380 deleting retry records that should have stayed in existence.
4382 48. Steve fixed some bugs in eximstats.
4384 49. The SPA authentication driver was not abandoning authentication and moving
4385 on to the next authenticator when an expansion was forced to fail,
4386 contradicting the general specification for all authenticators. Instead it
4387 was generating a temporary error. It now behaves as specified.
4389 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4390 (the order specifies the preference for clients). The order is now AES256,
4391 AES128, 3DES, ARCFOUR128.
4393 51. Small patch to Sieve code - explicitly set From: when generating an
4396 52. Exim crashed if a remote delivery caused a very long error message to be
4397 recorded - for instance if somebody sent an entire SpamAssassin report back
4398 as a large number of 550 error lines. This bug was coincidentally fixed by
4399 increasing the size of one of Exim's internal buffers (big_buffer) that
4400 happened as part of the Exiscan merge. However, to be on the safe side, I
4401 have made the code more robust (and fixed the comments that describe what
4404 53. Now that there can be additional text after "Completed" in log lines (if
4405 the queue_time_overall log selector is set), a one-byte patch to exigrep
4406 was needed to allow it to recognize "Completed" as not the last thing in
4409 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4410 patch that reportedly fixes this has been added. I am not expert enough to
4411 create a test for it. This is what the patch creator wrote:
4413 "I found a little strange behaviour of ldap code when working with
4414 Windows 2003 AD Domain, where users was placed in more than one
4415 Organization Units. When I tried to give exim partial DN, the exit code
4416 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4417 But simultaneously result of request was absolutely normal ldap result,
4418 so I produce this patch..."
4420 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4421 REFERENCE, so I have modified the code to exclude the patch when that macro
4424 55. Some experimental protocols are using DNS PTR records for new purposes. The
4425 keys for these records are domain names, not reversed IP addresses. The
4426 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4427 leaves it alone. Component reversal etc. now happens only for IP addresses.
4430 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4432 57. Double the size of the debug message buffer (to 2048) so that more of very
4433 long debug lines gets shown.
4435 58. The exicyclog utility now does better if the number of log files to keep
4436 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4438 59. Two changes related to the smtp_active_hostname option:
4440 (1) $smtp_active_hostname is now available as a variable.
4441 (2) The default for smtp_banner uses $smtp_active_hostname instead
4442 of $primary_hostname.
4444 60. The host_aton() function is supposed to be passed a string that is known
4445 to be a valid IP address. However, in the case of IPv6 addresses, it was
4446 not checking this. This is a hostage to fortune. Exim now panics and dies
4447 if the condition is not met. A case was found where this could be provoked
4448 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4449 components; fortuitously, this particular loophole had already been fixed
4450 by change 4.50/55 above.
4452 If there are any other similar loopholes, the new check in host_aton()
4453 itself should stop them being exploited. The report I received stated that
4454 data on the command line could provoke the exploit when Exim was running as
4455 exim, but did not say which command line option was involved. All I could
4456 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4457 running as the user.
4460 61. There was a buffer overflow vulnerability in the SPA authentication code
4461 (which came originally from the Samba project). I have added a test to the
4462 spa_base64_to_bits() function which I hope fixes it.
4465 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4466 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4468 63. The daemon start-up calls getloadavg() while still root for those OS that
4469 need the first call to be done as root, but it missed one case: when
4470 deliver_queue_load_max is set with deliver_drop_privilege. This is
4471 necessary for the benefit of the queue runner, because there is no re-exec
4472 when deliver_drop_privilege is set.
4474 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4475 This has been fixed.
4477 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4478 was in use, was not putting the data itself into the right store pool;
4479 consequently, it could be overwritten for a subsequent message in the same
4480 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4483 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4485 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4486 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4487 the function were treating the return as a boolean value, which happened to
4488 work because 0=false and not-0=true, but is not correct code.
4490 68. The host_aton() function was not handling scoped IPv6 addresses (those
4491 with, for example, "%eth0" on the end) correctly.
4493 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4494 negated items (that is, ~something) in unsigned ints. Some compilers
4495 apparently mutter when there is no cast.
4497 70. If an address verification called from an ACL failed, and did not produce a
4498 user-specific message (i.e. there was only a "system" message), nothing was
4499 put in $acl_verify_message. In this situation, it now puts the system
4502 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4503 session; change 4.31/43 added the unwanted input to the log line - except
4504 that it did not do this in the start of session case. It now does.
4506 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4507 This is wrong and can cause the other end to generate a synchronization
4508 error if it is another Exim or anything else that does the synchronization
4509 check. A QUIT command is no longer sent after a timeout.
4511 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4512 during host lookups.
4514 74. The defer_ok option of callout verification was not working if it was used
4515 when verifying addresses in header lines, that is, for this case:
4517 verify = header_sender/callout=defer_ok
4519 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4520 those file descriptors could be used for SMTP connections. If anything
4521 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4522 could be sent to the SMTP client, causing chaos. The daemon now opens
4523 stdin, stdout, and stderr to /dev/null when it puts itself into the
4526 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4527 log by default. The user can override this with suitable Perl magic.
4529 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4530 the log message when discard triggers, was not working for the DATA ACL or
4531 for the non-SMTP ACL.
4533 78. Error message wording change in sieve.c.
4535 79. If smtp_accept_max_per_host was set, the number of connections could be
4536 restricted to fewer than expected, because the daemon was trying to set up
4537 a new connection before checking whether the processes handling previous
4538 connections had finished. The check for completed processes is now done
4539 earlier. On busy systems, this bug wouldn't be noticed because something
4540 else would have woken the daemon, and it would have reaped the completed
4543 80. If a message was submitted locally by a user whose login name contained one
4544 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4545 It caused a spool format error. I have fixed the spool reading code. A
4546 related problem was that the "from" clause in the Received: line became
4547 illegal because of the space(s). It is now covered by ${quote_local_part.
4549 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4552 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4554 83. Patch from Sieve maintainer to fix unterminated string problem in
4555 "vacation" handling.
4557 84. Some minor changes to the Linux configuration files to help with other
4558 OS variants using glibc.
4560 85. One more patch for Sieve to update vacation handling to latest spec.
4563 ----------------------------------------------------
4564 See the note above about the 4.44 and 4.50 releases.
4565 ----------------------------------------------------
4571 1. Change 4.43/35 introduced a bug that caused file counts to be
4572 incorrectly computed when quota_filecount was set in an appendfile
4575 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4576 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4579 3. Give more explanation in the error message when the command for a transport
4580 filter fails to execute.
4582 4. There are several places where Exim runs a non-Exim command in a
4583 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4584 was being done only for the command run by the queryprogram router. It is
4585 now done for all such subprocesses. The other cases are: ${run, transport
4586 filters, and the commands run by the lmtp and pipe transports.
4588 5. Some older OS have a limit of 256 on the maximum number of file
4589 descriptors. Exim was using setrlimit() to set 1000 as a large value
4590 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4591 systems. I've change it so that if it can't get 1000, it tries for 256.
4593 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4594 was an oversight, and furthermore, ever since the addition of extra
4595 controls (e.g. 4.43/32), the checks on when to allow different forms of
4596 "control" were broken. There should now be diagnostics for all cases when a
4597 control that does not make sense is encountered.
4599 7. $recipients is now available in the predata ACL (oversight).
4601 8. Tidy the search cache before the fork to do a delivery from a message
4602 received from the command line. Otherwise the child will trigger a lookup
4603 failure and thereby defer the delivery if it tries to use (for example) a
4604 cached ldap connection that the parent has called unbind on.
4606 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4607 of $address_data from the recipient verification was clobbered by the
4608 sender verification.
4610 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4611 was its contents. (It was OK if the option was not defined at all.)
4613 11. A "Completed" log line is now written for messages that are removed from
4614 the spool by the -Mrm option.
4616 12. $host_address is now set to the target address during the checking of
4617 ignore_target_hosts.
4619 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4620 being passed; this would have caused $sender_host_name to have been used if
4621 matching the list had actually called for a host name (not very likely,
4622 since this list is usually IP addresses). A host name is now passed as
4625 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4626 code that specifies a non-restarting handler (typically sigaction() in
4627 modern systems) in an attempt to fix a rare and obscure crash bug.
4629 15. Narrowed the window for a race in the daemon that could cause it to ignore
4630 SIGCHLD signals. This is not a major problem, because they are used only to
4631 wake it up if nothing else does.
4633 16. A malformed maildirsize file could cause Exim to calculate negative values
4634 for the mailbox size or file count. Odd effects could occur as a result.
4635 The maildirsize information is now recalculated if the size or filecount
4638 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4639 support for a long time. Removed HAVE_SYS_VFS_H.
4641 18. Updated exipick to current release from John Jetmore.
4643 19. Allow an empty sender to be matched against a lookup in an address list.
4644 Previously the only cases considered were a regular expression, or an
4647 20. Exim went into a mad DNS lookup loop when doing a callout where the
4648 host was specified on the transport, if the DNS lookup yielded more than
4651 21. The RFC2047 encoding function was originally intended for short strings
4652 such as real names; it was not keeping to the 75-character limit for
4653 encoded words that the RFC imposes. It now respects the limit, and
4654 generates multiple encoded words if necessary. To be on the safe side, I
4655 have increased the buffer size for the ${rfc2047: expansion operator from
4658 22. Failure to deliver a bounce message always caused it to be frozen, even if
4659 there was an errors_to setting on the router. The errors_to setting is now
4662 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4663 canonical form (fully expanded) before being placed in
4664 $sender_host_address.
4666 24. Updated eximstats to version 1.33
4668 25. Include certificate and key file names in error message when GnuTLS fails
4669 to set them up, because the GnuTLS error message doesn't include the name
4670 of the failing file when there is a problem reading it.
4672 26. Expand error message when OpenSSL has problems setting up cert/key files.
4675 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4676 (this can affect the format of dates).
4678 28. exim_tidydb, when checking for the continued existence of a message for
4679 which it has found a message-specific retry record, was not finding
4680 messages that were in split spool directories. Consequently, it was
4681 deleting retry records that should have stayed in existence.
4683 29. eximstats updated to version 1.35
4684 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4685 1.35 - bugfix such that pie charts by volume are generated correctly
4687 30. The SPA authentication driver was not abandoning authentication and moving
4688 on to the next authenticator when an expansion was forced to fail,
4689 contradicting the general specification for all authenticators. Instead it
4690 was generating a temporary error. It now behaves as specified.
4692 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4693 (the order specifies the preference for clients). The order is now AES256,
4694 AES128, 3DES, ARCFOUR128.
4696 31. Small patch to Sieve code - explicitly set From: when generating an
4699 32. Exim crashed if a remote delivery caused a very long error message to be
4700 recorded - for instance if somebody sent an entire SpamAssassin report back
4701 as a large number of 550 error lines. This bug was coincidentally fixed by
4702 increasing the size of one of Exim's internal buffers (big_buffer) that
4703 happened as part of the Exiscan merge. However, to be on the safe side, I
4704 have made the code more robust (and fixed the comments that describe what
4707 33. Some experimental protocols are using DNS PTR records for new purposes. The
4708 keys for these records are domain names, not reversed IP addresses. The
4709 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4710 leaves it alone. Component reversal etc. now happens only for IP addresses.
4713 34. The host_aton() function is supposed to be passed a string that is known
4714 to be a valid IP address. However, in the case of IPv6 addresses, it was
4715 not checking this. This is a hostage to fortune. Exim now panics and dies
4716 if the condition is not met. A case was found where this could be provoked
4717 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4718 components; fortuitously, this particular loophole had already been fixed
4719 by change 4.50/55 or 4.44/33 above.
4721 If there are any other similar loopholes, the new check in host_aton()
4722 itself should stop them being exploited. The report I received stated that
4723 data on the command line could provoke the exploit when Exim was running as
4724 exim, but did not say which command line option was involved. All I could
4725 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4726 running as the user.
4729 35. There was a buffer overflow vulnerability in the SPA authentication code
4730 (which came originally from the Samba project). I have added a test to the
4731 spa_base64_to_bits() function which I hope fixes it.
4734 36. The daemon start-up calls getloadavg() while still root for those OS that
4735 need the first call to be done as root, but it missed one case: when
4736 deliver_queue_load_max is set with deliver_drop_privilege. This is
4737 necessary for the benefit of the queue runner, because there is no re-exec
4738 when deliver_drop_privilege is set.
4740 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4741 was in use, was not putting the data itself into the right store pool;
4742 consequently, it could be overwritten for a subsequent message in the same
4743 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4746 38. Sometimes the final signoff response after QUIT could fail to get
4747 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4748 < 0 before doing a fflush(). This bug looks as though it goes back to the
4749 introduction of TLS in release 3.20, but "sometimes" must have been rare
4750 because the tests only now provoked it.
4756 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4757 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4758 Now it returns an integer. A number of calls were still expecting a T/F
4759 return. Fortuitously, in all cases, the tests worked in OK situations,
4760 which is the norm. However, things would have gone wrong on any write
4761 failures on the smtp file descriptor. This function is used when sending
4762 messages over SMTP and also when doing verify callouts.
4764 2. When Exim is called to do synchronous delivery of a locally submitted
4765 message (the -odf or -odi options), it no longer closes stderr before doing
4768 3. Implemented the mua_wrapper option.
4770 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4772 5. Implemented the functions header_remove(), header_testname(),
4773 header_add_at_position(), and receive_remove_recipient(), and exported them
4776 6. If an ACL "warn" statement specified the addition of headers, Exim already
4777 inserted X-ACL-Warn: at the start if there was no header name. However, it
4778 was not making this test for the second and subsequent header lines if
4779 there were newlines in the string. This meant that an invalid header could
4780 be inserted if Exim was badly configured.
4782 7. Allow an ACL "warn" statement to add header lines at the start or after all
4783 the Received: headers, as well as at the end.
4785 8. Added the rcpt_4xx retry error code.
4787 9. Added postmaster_mailfrom=xxx to callout verification option.
4789 10. Added mailfrom=xxxx to the callout verification option, for verify=
4792 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4793 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4795 12. Inserted some casts to stop certain compilers warning when using pointer
4796 differences as field lengths or precisions in printf-type calls (mostly
4797 affecting debugging statements).
4799 13. Added optional readline() support for -be (dynamically loaded).
4801 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4802 same clock tick as a message's arrival, so that its received time was the
4803 same as the "first fail" time on the retry record, and that message
4804 remained on the queue past the ultimate address timeout, every queue runner
4805 would try a delivery (because it was past the ultimate address timeout) but
4806 after another failure, the ultimate address timeout, which should have then
4807 bounced the address, did not kick in. This was a "< instead of <=" error;
4808 in most cases the first failure would have been in the next clock tick
4809 after the received time, and all would be well.
4811 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4812 being recognized when the domain list was tested by the match_domain
4813 condition in an expansion string.
4815 16. Added the ${str2b64: operator.
4817 17. Exim was always calling setrlimit() to set a large limit for the number of
4818 processes, without checking whether the existing limit was already
4819 adequate. (It did check for the limit on file descriptors.) Furthermore,
4820 errors from getrlimit() and setrlimit() were being ignored. Now they are
4821 logged to the main and panic logs, but Exim does carry on, to try to do its
4822 job under whatever limits there are.
4824 18. Imported PCRE 5.0.
4826 19. Trivial typo in log message " temporarily refused connection" (the leading
4829 20. If the log selector return_path_on_delivery was set and an address was
4830 redirected to /dev/null, the delivery process crashed because it assumed
4831 that a return path would always be set for a "successful" delivery. In this
4832 case, the whole delivery is bypassed as an optimization, and therefore no
4835 21. Internal re-arrangement: the function for sending a challenge and reading
4836 a response while authentication was assuming a zero-terminated challenge
4837 string. It's now changed to take a pointer and a length, to allow for
4838 binary data in such strings.
4840 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4842 23. Exim was not respecting finduser_retries when seeking the login of the
4843 uid under which it was called; it was always trying 10 times. (The default
4844 setting of finduser_retries is zero.) Also, it was sleeping after the final
4845 failure, which is pointless.
4847 24. Implemented tls_on_connect_ports.
4849 25. Implemented acl_smtp_predata.
4851 26. If the domain in control=submission is set empty, Exim assumes that the
4852 authenticated id is a complete email address when it generates From: or
4853 Sender: header lines.
4855 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4856 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4857 chown and chgrp in /bin and hostname in /usr/bin.
4859 28. Exim was keeping the "process log" file open after each use, just as it
4860 does for the main log. This opens the possibility of it remaining open for
4861 long periods when the USR1 signal hits a daemon. Occasional processlog
4862 errors were reported, that could have been caused by this. Anyway, it seems
4863 much more sensible not to leave this file open at all, so that is what now
4866 29. The long-running daemon process does not normally write to the log once it
4867 has entered its main loop, and it closes the log before doing so. This is
4868 so that log files can straightforwardly be renamed and moved. However,
4869 there are a couple of unusual error situations where the daemon does write
4870 log entries, and I had neglected to close the log afterwards.
4872 30. The text of an SMTP error response that was received during a remote
4873 delivery was being truncated at 512 bytes. This is too short for some of
4874 the long messages that one sometimes sees. I've increased the limit to
4877 31. It is now possible to make retry rules that apply only when a message has a
4878 specific sender, in particular, an empty sender.
4880 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4881 it possible to be selective about when SMTP synchronization is enforced.
4883 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4885 32. Implemented hosts_connection_nolog.
4887 33. Added an ACL for QUIT.
4889 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4892 35. Added mailbox_size and mailbox_filecount to appendfile.
4894 36. Added control = no_multiline_responses to ACLs.
4896 37. There was a bug in the logic of the code that waits for the clock to tick
4897 in the case where the clock went backwards by a substantial amount such
4898 that the microsecond fraction of "now" was more than the microsecond
4899 fraction of "then" (but the whole seconds number was less).
4901 38. Added support for the libradius Radius client library this is found on
4902 FreeBSD (previously only the radiusclient library was supported).
4908 1. When certain lookups returned multiple values in the form name=value, the
4909 quoting of the values was not always being done properly. Specifically:
4910 (a) If the value started with a double quote, but contained no whitespace,
4912 (b) If the value contained whitespace other than a space character (i.e.
4913 tabs or newlines or carriage returns) it was not quoted.
4914 This fix has been applied to the mysql and pgsql lookups by writing a
4915 separate quoting function and calling it from the lookup code. The fix
4916 should probably also be applied to nisplus, ibase and oracle lookups, but
4917 since I cannot test any of those, I have not disturbed their existing code.
4919 2. A hit in the callout cache for a specific address caused a log line with no
4920 reason for rejecting RCPT. Now it says "Previous (cached) callout
4921 verification failure".
4923 3. There was an off-by-one bug in the queryprogram router. An over-long
4924 return line was truncated at 256 instead of 255 characters, thereby
4925 overflowing its buffer with the terminating zero. As well as fixing this, I
4926 have increased the buffer size to 1024 (and made a note to document this).
4928 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4929 when Exim is waiting for an SMTP response from a remote server, Exim
4930 restarts its select() call on the socket, thereby resetting its timeout.
4931 This is not a problem when such interrupts are rare. Somebody set up a cron
4932 job to run exiwhat every 2 minutes, which is less than the normal select()
4933 timeout (5 or 10 minutes). This meant that the select() timeout never
4934 kicked in because it was always reset. I have fixed this by comparing the
4935 time when an interrupt arrives with the time at the start of the first call
4936 to select(). If more time than the timeout has elapsed, the interrupt is
4937 treated as a timeout.
4939 5. Some internal re-factoring in preparation for the addition of Sieve
4940 extensions (by MH). In particular, the "personal" test is moved to a
4941 separate function, and given an option for scanning Cc: and Bcc: (which is
4942 not set for Exim filters).
4944 6. When Exim created an email address using the login of the caller as the
4945 local part (e.g. when creating a From: or Sender: header line), it was not
4946 quoting the local part when it contained special characters such as @.
4948 7. Installed new OpenBSD configuration files.
4950 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4951 try to make them clearer.
4953 9. Callout options, other than the timeout value, were being ignored when
4954 verifying sender addresses in header lines. For example, when using
4956 verify = header_sender/callout=no_cache
4958 the cache was (incorrectly) being used.
4960 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4961 only the Cygwin environment.
4963 11. When return_path_on_delivery was set as a log selector, if different remote
4964 addresses in the same message used different return paths and parallel
4965 remote delivery occurred, the wrong values would sometimes be logged.
4966 (Whenever a remote delivery process finished, the return path value from
4967 the most recently started remote delivery process was logged.)
4969 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4970 lines when AUTH and/or TLS are in use. This is the "received protocol"
4971 field. Exim used to use "asmtp" for authenticated SMTP, without any
4972 indication (in the protocol name) for TLS use. Now it follows the RFC and
4973 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4974 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4975 names appear in log lines as well as in Received: header lines.
4977 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4978 extensions, and comparison tests, and to fix some bugs.
4980 14. Changes to the "personal" filter test:
4982 (1) The test was buggy in that it was just doing the equivalent of
4983 "contains" tests on header lines. For example, if a user's address was
4984 anne@some.where, the "personal" test would incorrectly be true for
4986 To: susanne@some.where
4988 This test is now done by extracting each address from the header in turn,
4989 and checking the entire address. Other tests that are part of "personal"
4990 are now done using regular expressions (for example, to check local parts
4991 of addresses in From: header lines).
4993 (2) The list of non-personal local parts in From: addresses has been
4994 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4995 taken from the Sieve specification recommendations.
4997 (3) If the message contains any header line starting with "List-" it is
4998 treated as non-personal.
5000 (4) The test for "circular" in the Subject: header line has been removed
5001 because it now seems ill-conceived.
5003 15. Minor typos in src/EDITME comments corrected.
5005 16. Installed latest exipick from John Jetmore.
5007 17. If headers_add on a router specified a text string that was too long for
5008 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5009 of string_sprintf() is now avoided.
5011 18. $message_body_size was not set (it was always zero) when running the DATA
5012 ACL and the local_scan() function.
5014 19. For the "mail" command in an Exim filter, no default was being set for
5015 the once_repeat time, causing a random time value to be used if "once" was
5016 specified. (If the value happened to be <= 0, no repeat happened.) The
5017 default is now 0s, meaning "never repeat". The "vacation" command was OK
5018 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5019 (I found it when inspecting the code).
5021 20. There is now an overall timeout for performing a callout verification. It
5022 defaults to 4 times the callout timeout, which applies to individual SMTP
5023 commands during the callout. The overall timeout applies when there is more
5024 than one host that can be tried. The timeout is checked before trying the
5025 next host. This prevents very long delays if there are a large number of
5026 hosts and all are timing out (e.g. when the network connections are timing
5027 out). The value of the overall timeout can be changed by specifying an
5028 additional sub-option for "callout", called "maxwait". For example:
5030 verify = sender/callout=5s,maxwait=20s
5032 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5033 to the end before writing, but this should make it even safer).
5035 22. Exim was forgetting that it had advertised PIPELINING for the second and
5036 subsequent messages on an SMTP connection. It was also not resetting its
5037 memory on STARTTLS and an internal HELO.
5039 23. When Exim logs an SMTP synchronization error within a session, it now
5040 records whether PIPELINING has been advertised or not.
5042 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5043 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5044 rather than long int.
5046 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5048 26. Added the never_mail option to autoreply.
5054 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5055 crash if the getsockname() call failed; this can happen if a connection is
5056 closed very soon after it is established. The problem was simply in the
5057 order in which certain operations were done, causing Exim to try to write
5058 to the SMTP stream before it had set up the file descriptor. The bug has
5059 been fixed by making things happen in the correct order.
5065 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5066 before the connection was closed, thus losing the rejection response.
5068 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5069 some early Solaris releases, but causes trouble in current releases where
5070 socklen_t is defined.
5072 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5075 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5078 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5079 that releasing the top of it at the end releases what was used for sub-
5080 expansions (unless the block got too big). However, discard this block if
5081 the first thing is a variable or header, so that we can use its block when
5082 it is dynamic (useful for very large $message_headers, for example).
5084 6. Lookups now cache *every* query, not just the most recent. A new, separate
5085 store pool is used for this. It can be recovered when all lookup caches are
5086 flushed. Lookups now release memory at the end of their result strings.
5087 This has involved some general refactoring of the lookup sources.
5089 7. Some code has been added to the store_xxx() functions to reduce the amount
5090 of flapping under certain conditions.
5092 8. log_incoming_interface used to affect only the <= reception log lines. Now
5093 it causes the local interface and port to be added to several more SMTP log
5094 lines, for example "SMTP connection from", and rejection lines.
5096 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5098 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5100 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5101 could be overwritten at the end of the current message (or the start of a
5102 new message if it was set in a HELO ACL). The value is now preserved for
5103 the duration of the SMTP connection.
5105 12. If a transport had a headers_rewrite setting, and a matching header line
5106 contained an unqualified address, that address was qualified, even if it
5107 did not match any rewriting rules. The underlying bug was that the values
5108 of the flags that permit the existence of unqualified sender and recipient
5109 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5110 non-local messages, and by -bnq for local messages) were not being
5111 preserved with the message after it was received.
5113 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5114 "next input=" as part of the text comprising the host identity instead of
5115 the correct text. The code was using the same buffer for two different
5116 strings. However, depending on which order the printing function evaluated
5117 its arguments, the bug did not always show up. Under Linux, for example, my
5118 test suite worked just fine.
5120 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5121 doesn't work with some older versions of Perl. It has been changed to "my",
5122 which in any case is probably the better facility to use.
5124 15. A really picky compiler found some instances of statements for creating
5125 error messages that either had too many or two few arguments for the format
5128 16. The size of the buffer for calls to the DNS resolver has been increased
5129 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5130 for addresses that have a lot of PTR records. This alleviates a problem; it
5131 does not fully solve it.
5133 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5134 buffer now truncates the list and logs the incident, which is the same
5135 action as happens when Exim is looking up a host name and its aliases.
5136 Previously in this situation something unpredictable would happen;
5137 sometimes it was "internal error: store_reset failed".
5139 18. If a server dropped the connection unexpectedly when an Exim client was
5140 using GnuTLS and trying to read a response, the client delivery process
5141 crashed while trying to generate an error log message.
5143 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5144 string, for example:
5146 warn message = H1: something\nH2: something
5148 the text was added as a single header line from Exim's point of view
5149 though it ended up OK in the delivered message. However, searching for the
5150 second and subsequent header lines using $h_h2: did not work. This has been
5151 fixed. Similarly, if a system filter added multiple headers in this way,
5152 the routers could not see them.
5154 20. Expanded the error message when iplsearch is called with an invalid key to
5155 suggest using net-iplsearch in a host list.
5157 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5158 ACLs are no longer actually imposed (and a message to that effect is
5161 22. If a "gecos" field in a passwd entry contained escaped characters, in
5162 particular, if it contained a \" sequence, Exim got it wrong when building
5163 a From: or a Sender: header from that name. A second bug also caused
5164 incorrect handling when an unquoted " was present following a character
5165 that needed quoting.
5167 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5168 was not being matched caselessly.
5170 24. Arranged for all hyphens in the exim.8 source to be escaped with
5173 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5174 properly. Recipient callout cache records were still being keyed to include
5175 the sender, even when use_sender was set false. This led to far more
5176 callouts that were necessary. The sender is no longer included in the key
5177 when use_sender is false.
5179 26. Added "control = submission" modifier to ACLs.
5181 27. Added the ${base62d: operator to decode base 62 numbers.
5183 28. dnsdb lookups can now access SRV records.
5185 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5186 the configuration file.
5188 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5189 (-v) mode. This makes the output for a verbose queue run more intelligible.
5191 31. Added a use_postmaster feature to recipient callouts.
5193 32. Added the $body_zerocount variable, containing the number of binary zero
5194 bytes in the message body.
5196 33. The time of last modification of the "new" subdirectory is now used as the
5197 "mailbox time last read" when there is a quota error for a maildir
5200 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5202 35. Added +ignore_unknown as a special item in host lists.
5204 36. Code for decoding IPv6 addresses in host lists is now included, even if
5205 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5206 address was recognized as an IP address, but was then not correctly decoded
5207 into binary, causing unexpected and incorrect effects when compared with
5214 1. Very minor rewording of debugging text in manualroute to say "list of
5215 hosts" instead of "hostlist".
5217 2. If verify=header_syntax was set, and a header line with an unqualified
5218 address (no domain) and a large number of spaces between the end of the
5219 name and the colon was received, the reception process suffered a buffer
5220 overflow, and (when I tested it) crashed. This was caused by some obsolete
5221 code that should have been removed. The fix is to remove it!
5223 3. When running in the test harness, delay a bit after writing a bounce
5224 message to get a bit more predictability in the log output.
5226 4. Added a call to search_tidyup() just before forking a reception process. In
5227 theory, someone could use a lookup in the expansion of smtp_accept_max_
5228 per_host which, without the tidyup, could leave open a database connection.
5230 5. Added the variables $recipient_data and $sender_data which get set from a
5231 lookup success in an ACL "recipients" or "senders" condition, or a router
5232 "senders" option, similar to $domain_data and $local_part_data.
5234 6. Moved the writing of debug_print from before to after the "senders" test
5237 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5238 problems for message scanning, either using a data ACL, or using
5239 local_scan() because the Received: header was not generated till after they
5240 were called (in order to set the time as the time of reception completion).
5241 I have revised the way this works. The header is now generated after the
5242 body is received, but before the ACL or local_scan() are called. After they
5243 are run, the timestamp in the header is updated.
5249 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5250 before starting a queue runner without re-exec. This happened only when
5251 deliver_drop_privilege was set or when the Exim user was set to root. The
5252 effect of the bug was that timeouts during subsequent deliveries caused
5253 crashes instead of being properly handled. The handler is now left at its
5254 default (and expected) setting.
5256 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5257 message, again when deliver_drop_privilege is set or Exim is run as root.
5258 The bug described in (1) was not present in this case, but the tidying up
5259 of the other signals was missing. I have made the two cases consistent.
5261 3. The ignore_target_hosts setting on a manualroute router was being ignored
5262 for hosts that were looked up using the /MX notation.
5264 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5267 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5268 operated on the sender address. After changing the $sender_address to <>
5269 for the sender address verify, Exim was re-instated it as the original
5270 (before rewriting) address, but remembering that it had rewritten it, so it
5271 wasn't rewriting it again. This bug also had the effect of breaking the
5272 sender address verification caching when the sender address was rewritten.
5274 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5275 This has been changed so that if the ip literal address matches
5276 ignore_target_hosts, the router declines.
5278 7. Added expansion conditions match_domain, match_address, and match_local_
5279 part (NOT match_host).
5281 8. The placeholder for the Received: header didn't have a length field set.
5283 9. Added code to Exim itself and to exim_lock to test for a specific race
5284 condition that could lead to file corruption when using MBX delivery. The
5285 issue is with the lockfile that is created in /tmp. If this file is removed
5286 after a process has opened it but before that process has acquired a lock,
5287 there is the potential for a second process to recreate the file and also
5288 acquire a lock. This could lead to two Exim processes writing to the file
5289 at the same time. The added code performs the same test as UW imapd; it
5290 checks after acquiring the lock that its file descriptor still refers to
5291 the same named file.
5293 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5294 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5297 11. Added the smtp_active_hostname option. If used, this will typically be made
5298 to depend on the incoming interface address. Because $interface_address is
5299 not set up until the daemon has forked a reception process, error responses
5300 that can happen earlier (such as "too many connections") no longer contain
5303 12. If an expansion in a condition on a "warn" statement fails because a lookup
5304 defers, the "warn" statement is abandoned, and the next ACL statement is
5305 processed. Previously this caused the whole ACL to be aborted.
5307 13. Added the iplsearch lookup type.
5309 14. Added ident_timeout as a log selector.
5311 15. Added tls_certificate_verified as a log selector.
5313 16. Added a global option tls_require_ciphers (compare the smtp transport
5314 option of the same name). This controls incoming TLS connections.
5316 17. I finally figured out how to make tls_require_ciphers do a similar thing
5317 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5318 before starting the TLS session.
5320 18. Tabs are now shown as \t in -bP output.
5322 19. If the log selector return_path_on_delivery was set, Exim crashed when
5323 bouncing a message because it had too many Received: header lines.
5325 20. If two routers both had headers_remove settings, and the first one included
5326 a superfluous trailing colon, the final name in the first list and the
5327 first name in the second list were incorrectly joined into one item (with a
5328 colon in the middle).
5334 1. Added -C and -D options to the exinext utility, mainly to make it easier
5335 to include in the automated testing, but these could be helpful when
5336 multiple configurations are in use.
5338 2. The exinext utility was not formatting the output nicely when there was
5339 an alternate port involved in the retry record key, nor when there was a
5340 message id as well (for retries that were specific to a specific message
5341 and a specific host). It was also confused by IPv6 addresses, because of
5342 the additional colons they contain. I have fixed the IPv4 problem, and
5343 patched it up to do a reasonable job for IPv6.
5345 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5346 delivery, the log line now contains "pipelined" if PIPELINING was used.
5348 4. An SMTP transport process used to panic and die if the bind() call to set
5349 an explicit outgoing interface failed. This has been changed; it is now
5350 treated in the same way as a connect() failure.
5352 5. A reference to $sender_host_name in the part of a conditional expansion
5353 that was being skipped was still causing a DNS lookup. This no longer
5356 6. The def: expansion condition was not recognizing references to header lines
5357 that used bh_ and bheader_.
5359 7. Added the _cache feature to named lists.
5361 8. The code for checking quota_filecount in the appendfile transport was
5362 allowing one more file than it should have been.
5364 9. For compatibility with Sendmail, the command line option
5372 and sets the incoming protocol and host name (for trusted callers). The
5373 host name and its colon can be omitted when only the protocol is to be set.
5374 Note the Exim already has two private options, -pd and -ps, that refer to
5375 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5376 "s", but I don't think that's a major issue.
5378 10. A number of refactoring changes to the code, none of which should affect
5381 (a) The number of logging options was getting close to filling up the
5382 32-bit word that was used as a bit map. I have split them into two classes:
5383 those that are passed in the argument to log_write(), and those that are
5384 only ever tested independently outside of that function. These are now in
5385 separate 32-bit words, so there is plenty of room for expansion again.
5386 There is no change in the user interface or the logging behaviour.
5388 (b) When building, for example, log lines, the code previously used a
5389 macro that called string_cat() twice, in order to add two strings. This is
5390 not really sufficiently general. Furthermore, there was one instance where
5391 it was actually wrong because one of the argument was used twice, and in
5392 one call a function was used. (As it happened, calling the function twice
5393 did not affect the overall behaviour.) The macro has been replaced by a
5394 function that can join an arbitrary number of extra strings onto a growing
5397 (c) The code for expansion conditions now uses a table and a binary chop
5398 instead of a serial search (which was left over from when there were very
5399 few conditions). Also, it now recognizes conditions like "pam" even when
5400 the relevant support is not compiled in: a suitably worded error message is
5401 given if an attempt is made to use such a condition.
5403 11. Added ${time_interval:xxxxx}.
5405 12. A bug was causing one of the ddress fields not to be passed back correctly
5406 from remote delivery subprocesses. The field in question was not being
5407 subsequently used, so this caused to problems in practice.
5409 13. Added new log selectors queue_time and deliver_time.
5411 14. Might have fixed a bug in maildirsizefile handling that threw up
5412 "unexpected character" debug warnings, and recalculated the data
5413 unnecessarily. In any case, I expanded the warning message to give more
5416 15. Added the message "Restricted characters in address" to the statements in
5417 the default ACL that block characters like @ and % in local parts.
5419 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5420 Three changes have been made:
5422 (a) There was a serious bug; a negative response to MAIL caused the whole
5423 recipient domain to be cached as invalid, thereby blocking all messages
5424 to all local parts at the same domain, from all senders. This bug has
5425 been fixed. The domain is no longer cached after a negative response to
5426 MAIL if the sender used is not empty.
5428 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5431 (c) A new callout option, "use_sender" has been added for people who want
5432 the modified behaviour.
5438 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5441 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5442 indeed breaks things for older releases.
5444 3. Added additional logging to the case where there is a problem reading data
5445 from a filter that is running in a subprocess using a pipe, in order to
5446 try to track down a specific problem.
5448 4. Testing facility fudge: when running in the test harness and attempting
5449 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5450 getting "No route to host". Convert this to a timeout.
5452 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5455 6. Some OS don't have socklen_t but use size_t instead. This affects the
5456 fifth argument of getsockopt() amongst other things. This is now
5457 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5458 can be set for individual OS. I have set it for SunOS5, OSF1, and
5459 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5460 some earlier ones do not.
5462 7. Change 4.30/15 was not doing the test caselessly.
5464 8. The standard form for an IPv6 address literal was being rejected by address
5465 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5466 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5467 this, as well as the form without the "IPv6" on the front (but only when
5468 address literals are enabled, of course).
5470 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5472 10. Exim crashed if a message with an empty sender address specified by -f
5473 encountered a router with an errors_to setting. This could be provoked only
5474 by a command such as
5478 where an empty string was supplied; "<>" did not hit this bug.
5480 11. Installed PCRE release 4.5.
5482 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5483 remained set. It is now erased.
5485 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5486 times from message ids (which are base 36 rather than the normal 62).
5488 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5489 were being counted as actual protocol errors, and logged if the log
5490 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5491 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5492 and DATA following a set of rejected RCPTs do not count as protocol errors.
5493 In other words, Exim assumes they were pipelined, though this may not
5494 actually be the case. Of course, in all cases the client gets an
5495 appropriate error code.
5497 15. If a lookup fails in an ACL condition, a message about the failure may
5498 be available; it is used if testing the ACL cannot continue, because most
5499 such messages specify what the cause of the deferral is. However, some
5500 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5501 that caused an old message to be retained and used if a later statement
5502 caused a defer, replacing the real cause of the deferral.
5504 16. If an IP address had so many PTR records that the DNS lookup buffer
5505 was not large enough to hold them, Exim could crash while trying to process
5506 the truncated data. It now detects and logs this case.
5508 17. Further to 4.21/58, another change has been made: if (and only if) the
5509 first line of a message (the first header line) ends with CRLF, a bare LF
5510 in a subsequent header line has a space inserted after it, so as not to
5511 terminate the header.
5513 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5514 unnecessarily, used atoi() instead of strtol(), and didn't check the
5515 termination when getting file sizes from file names by regex.
5517 19. Completely re-implemented the support for maildirsize files, in the light
5518 of a number of problems with the previous contributed implementation
5519 (4.30/29). In particular:
5521 . If the quota is zero, the maildirsize file is maintained, but no quota is
5524 . If the maildir directory does not exist, it is created before any attempt
5525 to write a maildirsize file.
5527 . The quota value in the file is just a cache; if the quota is changed in
5528 the transport, the new value overrides.
5530 . A regular expression is available for excluding directories from the
5533 20. The autoreply transport checks the characters in options that define the
5534 message's headers; it allows continued headers, but it was checking with
5535 isspace() after an embedded newline instead of explicitly looking for a
5538 21. If all the "regular" hosts to which an address was routed had passed their
5539 expiry times, and had not reached their retry times, the address was
5540 bounced, even if fallback hosts were defined. Now Exim should go on to try
5543 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5544 equivalent code in the SMTP transport. Some hosts send humungous responses
5545 to HELO/EHLO, more than 1024 it seems.
5547 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5548 gives compiler warnings in some environments. I've now done it "properly",
5551 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5552 (because of problems with the built-in one) was declared to return uschar *
5553 instead of char *, causing compiler failure.
5555 25. Fixed a file descriptor leak when processing alias/forward files.
5557 26. Fixed a minor format string issue in dbfn.c.
5559 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5561 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5562 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5565 29. When a sender address is verified, it is cached, to save repeating the test
5566 when there is more than one recipient in a message. However, when the
5567 verification involves a callout, it is possible for different callout
5568 options to be set for different recipients. It is too complicated to keep
5569 track of this in the cache, so now Exim always runs a verification when a
5570 callout is required, relying on the callout cache for the optimization.
5571 The overhead is duplication of the address routing, but this should not be
5574 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5575 to be verified with callout=postmaster, and the main callout worked but the
5576 postmaster check failed, the verification correctly failed. However, if a
5577 subsequent RCPT command asked for sender verification *without* the
5578 postmaster check, incorrect caching caused this verification also to fail,
5581 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5582 it was not caching the DNS options (qualify_single, search_parents) that
5583 were used when the lookup failed. A subsequent lookup with different
5584 options therefore always gave the same answer, though there were cases
5585 where it should not have. (Example: a "domains = !$mx_any" option on a
5586 dnslookup router: the "domains" option is always processed without any
5587 widening, but the router might have qualify_single set.) Now Exim uses the
5588 cached value only when the same options are set.
5590 32. Added John Jetmore's "exipick" utility to the distribution.
5592 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5593 than a timeout (e.g. a certificate is required, and is not provided), an
5594 Exim server now closes the connection immediately. Previously it waited for
5595 the client to close - but if the client is SSL, it seems that they each
5596 wait for each other, leading to a delay before one of them times out.
5598 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5599 maintained 0.8.x compatibility because I don't think many are using it, and
5600 it is clearly obsolete.
5602 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5605 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5606 client certificate was expired. A simple patch fixes this, though I don't
5607 understand the full logic of why the verify callback is called multiple
5610 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5611 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5612 which causes problems with some clients (such as the Certicom SSL Plus
5613 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5614 disables the coutermeasure allowing Eudora to connect."
5616 38. Exim was not checking that a write() to a log file succeeded. This could
5617 lead to Bad Things if a log got too big, in particular if it hit a file
5618 size limit. Exim now panics and dies if it cannot write to a log file, just
5619 as it does if it cannot open a log file.
5621 39. Modified OS/Makefile-Linux so that it now contains
5623 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5625 The two -D definitions ensure that Exim is compiled with large file
5626 support, which makes it possible to handle log files that are bigger than
5629 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5630 instance) a domain was checked against a named list that involved a lookup,
5631 causing $domain_data to be set, then another domain was checked against the
5632 same list, then the first domain was re-checked, the value of $domain_data
5633 after the final check could be wrong. In particular, if the second check
5634 failed, it could be set empty. This bug probably also applied to
5637 41. The strip_trailing_dot option was not being applied to the address given
5638 with the -f command-line option.
5640 42. The code for reading a message's header from the spool was incrementing
5641 $received_count, but never initializing it. This meant that the value was
5642 incorrect (doubled) while delivering a message in the same process in which
5643 it was received. In the most common configuration of Exim, this never
5644 happens - a fresh exec is done - but it can happen when
5645 deliver_drop_privilege is set.
5647 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5648 it now includes up to 150 characters of the unexpected data in the log
5651 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5652 and building data strings. The size of both of these buffers was 10 000
5653 bytes - far larger than anybody would *ever* want, thought I. Needless to
5654 say, somebody hit the limit. I have increased the maximum line length to
5655 20 000 and the maximum data length of concatenated lines to 100 000. I have
5656 also fixed two bugs, because there was no checking on these buffers. Tsk,
5657 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5658 buffer is too small.
5660 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5661 lsearch lookups. Now it does.
5663 46. When parsing a route_list item in a manualroute router, a fixed-length
5664 buffer was used for the list of hosts. I made this 1024 bytes long,
5665 thinking that nobody would ever have a list of hosts that long. Wrong.
5666 Somebody had a whole pile of complicated expansion conditions, and the
5667 string was silently truncated, leading to an expansion error. It turns out
5668 that it is easier to change to an unlimited length (owing to other changes
5669 that have happened since this code was originally written) than to build
5670 structure for giving a limitation error. The length of the item that
5671 expands into the list of hosts is now unlimited.
5673 47. The lsearch lookup could not handle data where the length of text line was
5674 more than 4095 characters. Such lines were truncated, leading to shortened
5675 data being returned. It should now handle lines of any length.
5677 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5678 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5681 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5682 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5683 no longer happens in comment lines. A list of replacements is now placed
5684 at the head of all of the source files, except those whose only change is
5685 to replace PERL_COMMAND in the very first #! line.
5687 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5688 messages on the queue, with a bottom-up merge sort, using code contributed
5689 by Michael Haardt. This should make operations like -bp somewhat faster on
5690 large queues. It won't affect queue runners, except when queue_run_in_order
5693 51. Installed eximstats 1.31 in the distribution.
5695 52. Added support for SRV lookups to the dnslookup router.
5697 53. If an ACL referred to $message_body or $message_body_end, the value was not
5698 reset for any messages that followed in the same SMTP session.
5700 54. The store-handling optimization for building very long strings was not
5701 differentiating between the different store pools. I don't think this
5702 actually made any difference in practice, but I've tidied it.
5704 55. While running the routers to verify a sender address, $sender_address
5705 was still set to the sender address. This is wrong, because when routing to
5706 send a bounce to the sender, it would be empty. Therefore, I have changed
5707 it so that, while verifying a sender address, $sender_address is set to <>.
5708 (There is no change to what happens when verifying a recipient address.)
5710 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5711 target A or AAAA records (if not already returned) without resetting the
5712 qualify_single or search_parents options of the DNS resolver. These are
5713 inappropriate in this case because the targets of MX and SRV records must
5714 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5715 target that, when qualified, matched something in the local domain. These
5716 two options are now turned off when doing these lookups.
5718 57. It seems that at least some releases of Reiserfs (which does not have the
5719 concept of a fixed number of inodes) returns zero and not -1 for the
5720 number of available inodes. This interacted badly with check_spool_inodes,
5721 which assumed that -1 was the "no such thing" setting. What I have done is
5722 to check that the total number of inodes is greater than zero before doing
5723 the test of how many are available.
5725 58. When a "warn" ACL statement has a log_message modifier, the message is
5726 remembered, and not repeated. This is to avoid a lot of repetition when a
5727 message has many recipients that cause the same warning to be written.
5728 However, Exim was preserving the list of already written lines for an
5729 entire SMTP session, which doesn't seem right. The memory is now reset if a
5730 new message is started.
5732 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5733 debugging output unless log_rewrite was also set.
5735 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5736 of (int)(handle) when we know that handle contains (void *)(-1).
5738 61. The Exim daemon panic-logs an error return when it closes the incoming
5739 connection. However "connection reset by peer" seems to be common, and
5740 isn't really an error worthy of noting specially, so that particular error
5743 62. When Exim is trying to find all the local interfaces, it used to panic and
5744 die if the ioctl to get the interface flags failed. However, it seems that
5745 on at least one OS (Solaris 9) it is possible to have an interface that is
5746 included in the list of interfaces, but for which you get a failure error
5747 for this call. This happens when the interface is not "plumbed" into a
5748 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5749 failure of the "get flags" call assumes that the interface is down.
5751 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5752 makes life easier for people who are doing arithmetic on fields extracted
5753 from dates, where you often get leading zeros that should not be
5754 interpreted as octal.
5756 64. Added qualify_domain to the redirect router, to override the global
5759 65. If a pathologically long header line contained very many addresses (the
5760 report of this problem mentioned 10 000) and each of them was rewritten,
5761 Exim could use up a very large amount of memory. (It kept on making new
5762 copies of the header line as it rewrote, and never released the old ones.)
5763 At the expense of a bit more processing, the header rewriting function has
5764 been changed so that it no longer eats memory in this way.
5766 66. The generation of the Received: header has been moved from the time that a
5767 message starts to be received, to the time that it finishes. The timestamp
5768 in the Received: header should now be very close to that of the <= log
5769 line. There are two side-effects of this change:
5771 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5772 logged header lines no longer include the local Received: line, because
5773 it has not yet been created. The same applies to a copy of the message
5774 that is returned to a non-SMTP sender when a message is rejected.
5776 (b) When a filter file is tested using -bf, no additional Received: header
5777 is added to the test message. After some thought, I decided that this
5780 This change does not affect the value of $received_for. It is still set
5781 after address rewriting, but before local_scan() is called.
5783 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5785 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5786 gave an unhelpful panic error message, and a defer error. I have managed to
5787 change this behaviour so that it now rejects any supplied certificate,
5788 which seems right, as the list of acceptable certificates is empty.
5790 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5791 gave an unhelpful defer error. I have not managed to make this reject any
5792 supplied certificates, but the error message it gives is "no certificate
5793 supplied", which is not helpful.
5795 70. exigrep's output now also includes lines that are not associated with any
5796 message, but which match the given pattern. Implemented by a patch from
5797 Martin Sluka, which also tidied up the Perl a bit.
5799 71. Recipient callout verification, like sender verification, was using <> in
5800 the MAIL FROM command. This isn't really the right thing, since the actual
5801 sender may affect whether the remote host accepts the recipient or not. I
5802 have changed it to use the actual sender in the callout; this means that
5803 the cache record is now keyed on a recipient/sender pair, not just the
5804 recipient address. There doesn't seem to be a real danger of callout loops,
5805 since a callout by the remote host to check the sender would use <>.
5806 [SEE ABOVE: changed after hitting problems.]
5808 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5809 temporary errors. However, in the case of such a code being given after
5810 the end of a data transmission (i.e. after ".") Exim was failing to write
5811 a retry record for the message. (Yes, there was some broken host that was
5812 actually sending 8xx at this point.)
5814 73. An unknown lookup type in a host list could cause Exim to panic-die when
5815 the list was checked. (An example that provoked this was putting <; in the
5816 middle of a list instead of at the start.) If this happened during a DATA
5817 ACL check, a -D file could be left lying around. This kind of configuration
5818 error no longer causes Exim to die; instead it causes a defer error. The
5819 incident is still logged to the main and panic logs.
5821 74. Buglet left over from Exim 3 conversion. The message "too many messages
5822 in one connection" was written to the rejectlog but not the mainlog, except
5823 when address rewriting (yes!) was being logged.
5825 75. Added write_rejectlog option.
5827 76. When a system filter was run not as root (that is, when system_filter_user
5828 was set), the values of the $n variables were not being returned to the
5829 main process; thus, they were not subsequently available in the $sn
5832 77. Added +return_path_on_delivery log selector.
5834 78. A connection timeout was being treated differently from recipients deferred
5835 when testing hosts_max_try with a message that was older than the host's
5836 retry timeout. (The host should not be counted, thus allowing all hosts to
5837 be tried at least once before bouncing.) This may have been the cause of an
5838 occasionally reported bug whereby a message would remain on the queue
5839 longer than the retry timeout, but would be bounced if a delivery was
5840 forced. I say "may" because I never totally pinned down the problem;
5841 setting up timeout/retry tests is difficult. See also the next item.
5843 79. The ultimate address timeout was not being applied to errors that involved
5844 a combination of host plus message (for example, a timeout on a MAIL
5845 command). When an address resolved to a number of possible hosts, and they
5846 were not all tried for each delivery (e.g. because of hosts_max_try), a
5847 message could remain on the queue longer than the retry timeout.
5849 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5852 81. Fixed an obscure SMTP outgoing bug which required at least the following
5853 conditions: (a) there was another message waiting for the same server;
5854 (b) the server returned 5xx to all RCPT commands in the first message so
5855 that the message was not completed; (c) the server dropped the connection
5856 or gave a negative response to the RSET that Exim sends to abort the
5857 transaction. The observed case was a dropped connection after DATA that had
5858 been sent in pipelining mode. That is, the server had advertised PIPELINING
5859 but was not implementing it correctly. The effect of the bug was incorrect
5860 behaviour, such as trying another host, and this could lead to a crash.
5866 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5867 and daemon.c were passed as pointers to ints; they should have been
5868 pointers to socklen_t variables (which are typically unsigned ints).
5870 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5873 3. Fixed a really odd bug that affected only the testing scheme; patching a
5874 certain fixed string in the binary changed the value of another string that
5875 happened to be identical to the end of the original first string.
5877 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5878 name", it returns that address as the IP address. On some operating
5879 systems (e.g. Solaris), it also passes back the IP address string as the
5880 "host name". However, on others (e.g. Linux), it passes back an empty
5881 string. Exim wasn't checking for this, and was changing the host name to an
5882 empty string, assuming it had been canonicalized.
5884 5. Although rare, it is permitted to have more than one PTR record for a given
5885 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5886 all the names associated with an address, because they do in Solaris.
5887 However, it seems that they do not in Linux for data that comes from the
5888 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5889 I found this out when I moved to a new Linux workstation and tried to run
5890 the Exim test suite.
5892 To get round this problem I have changed the code so that it now does its
5893 own call to the DNS to look up PTR records when searching for a host name.
5894 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5895 addresses that are only in /etc/hosts are still found.
5897 This behaviour is, however, controlled by an option called host_lookup_
5898 order, which defaults to "bydns:byaddr". If people want to use the other
5899 order, or indeed, just use one or the other means of lookup, they can
5900 specify it in this variable.
5902 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5903 some operating systems, this comes back from gethostbyaddr() as an empty
5904 string, and this is what Exim used to test for. However, it seems that in
5905 other systems, "." is yielded. Exim now tests for this case too.
5907 7. The values of check_spool_space and check_log_space are now held internally
5908 as a number of kilobytes instead of an absolute number of bytes. If a
5909 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5910 kilobyte. This means that much larger values can be stored.
5912 8. Exim monitor: an attempt to get the action menu when not actually pointing
5913 at a message produces an empty menu entitled "No message selected". This
5914 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5915 no entries in it ("Shell widget menu has zero width and/or height"). So I
5916 have added a single, blank menu entry in this case.
5918 9. Added ${quote_local_part.
5920 10. MIME decoding is now applied to the contents of Subject: header lines when
5923 11. Now that a reference to $sender_host_address automatically causes a reverse
5924 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5925 host lookup before query-style lookups in lists that might use this
5926 variable. This has therefore been abolished, and the "net-" prefix is no
5927 longer necessary for query-style lookups.
5929 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5930 have been a typo for LFLAGS, so it has been changed.
5932 13. The install script calls Exim with "-C /dev/null" in order to find the
5933 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5934 to be output. However, since Exim outputs its version number before the
5935 error, it didn't break the script. It just looked ugly. I fixed this by
5936 always allowing "-C /dev/null" if the caller is root.
5938 14. Ignore overlarge ACL variable number when reading spool file - insurance
5939 against a later release with more variables having written the file.
5941 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5942 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5943 this, as well as the form without the "IPv6" on the front.
5945 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5946 OS/Makefile-Darwin file.
5948 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5949 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5951 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5952 variable. However, some OS use ints for inodes. I've added cast to long int
5953 to get rid of the compiler warning.
5955 19. I had forgotten to lock out "/../" in configuration file names when
5956 ALT_CONFIG_PREFIX was set.
5958 20. Routers used for verification do not need to specify transports. However,
5959 if such a router generated a host list, and callout was configured, Exim
5960 crashed, because it could not find a port number from the (non-existent)
5961 transport. It now assumes port 25 in this circumstance.
5963 21. Added the -t option to exigrep.
5965 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5966 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5967 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5968 them is set, LOOKUP_LSEARCH is forced.
5970 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5972 24. Added sender and host information to the "rejected by local_scan()" log
5973 line; previously there was no indication of these.
5975 25. Added .include_if_exists.
5977 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5978 Linux. It turns out that not all file systems support this. Apparently some
5979 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5980 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5981 not supported on the file descriptor, is now ignored when Exim is trying to
5982 sync a directory. This applies only to Linux.
5984 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5986 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5987 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5990 29. Added support for maildirsize files from supplied patch (modified a bit).
5992 30. The use of :fail: followed by an empty string could lead Exim to respond to
5993 sender verification failures with (e.g.):
5995 550 Verification failed for <xxx>
5996 550 Sender verify failed
5998 where the first response line was missing the '-' that indicates it is not
5999 the final line of the response.
6001 31. The loop for finding the name of the user that called Exim had a hardwired
6002 limit of 10; it now uses the value of finduser_retries, which is used for
6003 all other user lookups.
6005 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6008 33. Exim was neglecting to zero errno before one call of strtol() when
6009 expanding a string and expecting an integer value. On some systems this
6010 resulted in spurious "integer overflow" errors. Also, it was casting the
6011 result into an int without checking.
6013 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6014 did not work. The code looks as if it has *never* worked, though it appears
6015 to have been documented since at least release 1.62. I have made it work.
6017 35. The "timeout_DNS" error in retry rules, also documented since at least
6018 1.62, also never worked. As it isn't clear exactly what this means, and
6019 clearly it isn't a major issue, I have abolished the feature by treating it
6020 as "timeout", and writing a warning to the main and panic logs.
6022 36. The display of retry rules for -brt wasn't always showing the error code
6025 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6026 timeout_connect_A, timeout_connect_MX.
6028 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6029 to the empty sender.
6031 39. The daemon was not analysing the content of -oX till after it had closed
6032 stderr and disconnected from the controlling terminal. This meant that any
6033 syntax errors were only noted on the panic log, and the return code from
6034 the command was 0. By re-arranging the code a little, I've made the
6035 decoding happen first, so such errors now appear on stderr, and the return
6036 code is 1. However, the actual setting up of the sockets still happens in
6037 the disconnected process, so errors there are still only recorded on the
6040 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6041 connections (as happens on some IP stacks) was logged at start up time as
6042 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6043 it from "IPv6 and IPv4", which means that two separate sockets are being
6046 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6047 says whether AF_INET or AF_INET6 was passed as an argument.
6049 42. Exiwhat output was messed up when time zones were included in log
6052 43. Exiwhat now gives more information about the daemon's listening ports,
6053 and whether -tls-on-connect was used.
6055 44. The "port" option of the smtp transport is now expanded.
6057 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6058 silently ignored. Now an error message is written to the main and panic
6061 46. There's a new ACL modifier called "logwrite" which writes to a log file
6062 as soon as it is encountered.
6064 47. Added $local_user_uid and $local_user_gid at routing time.
6066 48. Exim crashed when trying to verify a sender address that was being
6069 49. Exim was recognizing only a space character after ".include". It now also
6070 recognizes a tab character.
6072 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6073 extracting the relevant information from the specification. The man page no
6074 longer contains scrambled data for the -d option, and I've added a section
6075 at the front about calling Exim under different names.
6077 51. Added "extra_headers" argument to the "mail" command in filter files.
6079 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6082 53. Installed eximstats 1.29.
6084 54. Added transport_filter_timeout as a generic transport option.
6086 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6087 Cc: header lines. This was required by RFC 822, but it not required by RFC
6090 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6091 incoming messages that did not have them. Now it does so only if the
6092 message originates locally, that is, if there is no associated remote host
6093 address. When Resent- header lines are present, this applies to the Resent-
6094 lines rather than the non-Resent- lines.
6096 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6097 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6099 58. Messages for configuration errors now include the name of the main
6100 configuration file - useful now that there may be more than one file in a
6101 list (.included file names were always shown).
6103 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6104 for those rare installations that do not start the daemon as root or run it
6105 setuid root. I've cut out the call to initgroups() if the daemon is not
6108 60. The Exim user and group can now be bound into the binary as text strings
6109 that are looked up at the start of Exim's processing.
6111 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6113 62. Added $mailstore_basename variable.
6115 63. Installed patch to sieve.c from Michael Haardt.
6117 64. When Exim failed to open the panic log after failing to open the main log,
6118 the original message it was trying to log was written to stderr and debug
6119 output, but if they were not available (the usual case in production), it
6120 was lost. Now it is written to syslog before the two lines that record the
6121 failures to open the logs.
6123 65. Users' Exim filters run in subprocesses under the user's uid. It is
6124 possible for a "deliver" command or an alias in a "personal" command to
6125 provoke an address rewrite. If logging of address rewriting is configured,
6126 this fails because the process is not running as root or exim. There may be
6127 a better way of dealing with this, but for the moment (because 4.30 needs
6128 to be released), I have disabled address rewrite logging when running a
6129 filter in a non-root, non-exim process.
6135 1. The buildconfig auxiliary program wasn't quoting the value set for
6136 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6137 not defined. This bug was masked in 4.22 by the effect that was fixed in
6140 2. Some messages that were rejected after a message id was allocated were
6141 shown as "incomplete" by exigrep. It no longer does this for messages that
6142 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6144 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6145 have allow_domain_literals set, the ID did not get logged in the <= line.
6146 Domain literals are now always recognized in Message-ID: header lines.
6148 4. The first argument for a ${extract expansion item is the key name or field
6149 number. Leading and trailing spaces in this item were not being ignored,
6150 causing some misleading effects.
6152 5. When deliver_drop_privilege was set, single queue runner processes started
6153 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6154 same command in the process it spins off) were not dropping privilege.
6156 6. When the daemon running as "exim" started a queue runner, it always
6157 re-executed Exim in the spun-off process. This is a waste of effort when
6158 deliver_drop_privilege is set. The new process now just calls the
6159 queue-runner function directly.
6165 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6168 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6169 ignored. Though the use of -oP was forcing the writing of a pid file, it
6170 was always written to the default place.
6172 3. If the message "no IP address found for host xxxx" is generated during
6173 incoming verification, it is now followed by identification of the incoming
6174 connection (so you can more easily find what provoked it).
6176 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6178 5. Added some features to "harden" Exim a bit more against certain attacks:
6180 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6181 be put in Local/Makefile. This is like the never_users runtime option,
6182 but it cannot be overridden. The default setting is "root".
6184 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6185 prefix string with which any file named in a -C command line option
6188 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6189 is retained for -C and -D only if the caller of Exim is root. Without
6190 it, the exim user may also use -C and -D and retain privilege.
6192 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6193 command line option is disabled.
6195 6. Macro names set by the -D option must start with an upper case letter, just
6196 like macro names defined in the configuration file.
6198 7. Added "dereference=" facility to LDAP.
6200 8. Two instances of the typo "uknown" in the source files are fixed.
6202 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6203 the Configure-Makefile script screwed up while processing it.
6205 10. Incorporated PCRE 4.4.
6207 11. The SMTP synchronization check was not operating right at the start of an
6208 SMTP session. For example, it could not catch a HELO sent before the client
6209 waited for the greeting. There is now a check for outstanding input at the
6210 point when the greeting is written. Because of the duplex, asynchronous
6211 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6212 way, but not yet received, when the check is performed.
6214 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6215 on TCP/IP sockets, because this apparently causes some broken clients to
6218 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6219 unchanged) from the Cygwin maintainer.
6221 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6222 was supported instead of testing for mbx. Effectively a typo.
6224 15. The spa authenticator server code was not checking that the input it
6225 received was valid base64.
6227 16. The debug output line for the "set" modifier in ACLs was not showing the
6228 name of the variable that was being set.
6230 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6232 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6233 lookup on its own. Something else was needed to trigger the lookup. For
6234 example, a match in host_lookup or the need for a host name in a host list.
6235 Now, if $sender_host_name is referenced and the host name has not yet been
6236 looked up, a lookup is performed. If the lookup fails, the variable remains
6237 empty, and $host_lookup_failed is set to "1".
6239 19. Added "eqi" as a case-independent comparison operator.
6241 20. The saslauthd authentication condition could segfault if neither service
6242 nor realm was specified.
6244 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6245 error message that was logged was misleading, and incoming SMTP
6246 connections were dropped. The message is now more accurate, and temporary
6247 errors are given to SMTP connections.
6249 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6250 (except RSET) with a 421 error, until QUIT is received. However, it was
6251 failing to send a response to QUIT.
6253 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6254 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6255 ACL is now run after the helo_try_verify_hosts code.
6257 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6258 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6259 case-independent, so other case variants are also recognized). Apparently
6260 some systems use these upper case variants.
6262 25. If more than two messages were waiting for the same host, and a transport
6263 filter was specified for the transport, Exim sent two messages over the
6264 same TCP/IP connection, and then failed with "socket operation on non-
6265 socket" when it tried to send the third.
6267 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6269 27. The extern definition of crypt16() in expand.c was not being excluded when
6270 the OS had its own crypt16() function.
6272 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6273 as a preferred synonym for return_size_limit, both as an option and as an
6276 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6278 30. Changed the default configuration ACL to relax the local part checking rule
6279 for addresses that are not in any local domains. For these addresses,
6280 slashes and pipe symbols are allowed within local parts, but the sequence
6281 /../ is explicitly forbidden.
6283 31. SPA server authentication was not clearing the challenge buffer before
6286 32. log_message in a "warn" ACL statement was writing to the reject log as
6287 well as to the main log, which contradicts the documentation and doesn't
6288 seem right (because no rejection is happening). So I have stopped it.
6290 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6291 However, I am unable to do any testing of this.
6293 34. Fixed an infelicity in the appendfile transport. When checking directories
6294 for a mailbox, to see if any needed to be created, it was accidentally
6295 using path names with one or more superfluous leading slashes; tracing
6296 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6298 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6299 contents are added to the log line that is written for every discarded
6300 recipient. (Previously a log_message setting was ignored.)
6302 36. The ${quote: operator now quotes the string if it is empty.
6304 37. The install script runs exim in order to find its version number. If for
6305 some reason other than non-existence or emptiness, which it checks, it
6306 could not run './exim', it was installing it with an empty version number,
6307 i.e. as "exim-". This error state is now caught, and the installation is
6310 38. An argument was missing from the function that creates an error message
6311 when Exim fails to connect to the socket for saslauthd authentication.
6312 This could cause Exim to crash, or give a corrupted message.
6314 39. Added isip, isip4, and isip6 to ${if conditions.
6316 40. The ACL variables $acl_xx are now saved with the message, and can be
6317 accessed later in routers, transports, and filters.
6319 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6320 strings in the file are not string-expanded.
6322 42. If a MAIL command specified a SIZE value that was too large to fit into an
6323 int variable, the check against message_size_limit failed. Such values are
6324 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6325 day this will have to be increased, but I don't think I want to be around
6326 when emails are that large.
6333 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6334 iconv() is not standard in FreeBSD.
6336 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6337 IPv6 enabled. The observed symptom was a segmentation fault on return from
6338 the function os_common_find_running_interfaces() in src/os.c.
6340 3. In the check_special_case() function in daemon.c I had used "errno" as an
6341 argument name, which causes warnings on some systems. This was basically a
6342 typo, since it was named "eno" in the comments!
6344 4. The code that waits for the clock to tick (at a resolution of some fraction
6345 of a second) so as to ensure message-id uniqueness was always waiting for
6346 at least one whole tick, when it could have waited for less. [This is
6347 almost certainly not relevant at current processor speeds, where it is
6348 unlikely to ever wait at all. But we try to future-proof.]
6350 5. The function that sleeps for a time interval that includes fractions of a
6351 second contained a race. It did not block SIGALRM between setting the
6352 timer, and suspending (a couple of lines later). If the interval was short
6353 and the sigsuspend() was delayed until after it had expired, the suspension
6354 never ended. On busy systems this could lead to processes getting stuck for
6357 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6358 process, before it forks any delivery processes. The open lookup caching
6359 mechanism meant that the open file or database connection was passed into
6360 the delivery process. The problem was that delivery processes always tidy
6361 up cached lookup data. This could cause a problem for the next delivery
6362 process started by the queue runner, because the external queue runner
6363 process does not know about the closure. So the next delivery process
6364 still has data in the lookup cache. In the case of a file lookup, there was
6365 no problem because closing a file descriptor in a subprocess doesn't affect
6366 the parent. However, if the lookup was caching a connection to a database,
6367 the connection was closed, and the second delivery process was likely to
6368 see errors such as "PGSQL: query failed: server closed the connection
6369 unexpectedly". The problem has been fixed by closing all cached lookups
6370 in a queue runner before running a delivery process.
6372 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6373 seem to have the "const" qualifier which it has on other OS. I've
6376 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6377 *of the same type* (client or server) with the same public name that an
6378 error should be diagnosed.
6380 9. When Exim looked up a host name for an IP address, but failed to find the
6381 original IP address when looking up the host name (a safety check), it
6382 output the message "<ip address> does not match any IP for NULL", which was
6383 confusing, to say the least. The bug was that the host name should have
6384 appeared instead of "NULL".
6386 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6387 user that is built into the binary, and the -C or -D options is used, root
6388 privilege is dropped before the configuration file is read. In addition,
6389 logging is switched to stderr instead of the normal log files. If the
6390 configuration then re-defines the Exim user, the unprivileged environment
6391 is probably not what is expected, so Exim logs a panic warning message (but
6394 However, if deliver_drop_privilege is set, the unprivileged state may well
6395 be exactly what is intended, so the warning has been cut out in that case,
6396 and Exim is allowed to try to write to its normal log files.
6402 1. smtp_return_error_details was not giving details for temporary sender
6403 or receiver verification errors.
6405 2. Diagnose a configuration error if two authenticators have the same public
6408 3. Exim used not to create the message log file for a message until the first
6409 delivery attempt. This could be confusing when incoming messages were held
6410 for policy or load reasons. The message log file is now created at the time
6411 the message is received, and an initial "Received" line is written to it.
6413 4. The automatically generated man page for command line options had a minor
6414 bug that caused no ill effects; however, a more serious problem was that
6415 the procedure for building the man page automatically didn't always
6416 operate. Consequently, release 4.20 contains an out-of-date version. This
6417 shouldn't happen again.
6419 5. When building Exim with embedded Perl support, the script that builds the
6420 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6421 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6423 6. The freeze_tell option was not being used for messages that were frozen on
6424 arrival, either by an ACL or by local_scan().
6426 7. Added the smtp_incomplete_transaction log selector.
6428 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6429 was accepting AUTH without a new EHLO.
6431 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6432 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6435 10. Logging of TCP/IP connections (when configured) now happens in the main
6436 daemon process instead of the child process, so that the TCP/IP connection
6437 count is more accurate (but it can never be perfect).
6439 11. The use of "drop" in a nested ACL was not being handled correctly in the
6440 outer ACL. Now, if condition failure induced by the nested "drop" causes
6441 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6442 or "require"), the connection is dropped.
6444 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6445 that yield "discard" can now be used with an "accept" or a "discard" verb,
6446 but an error is generated for any others (because I can't see a useful way
6447 to define what should happen).
6449 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6450 are now processed in the same way as lines in the Exim configuration file.
6451 In particular, continuation lines are supported.
6453 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6455 15. Added -ti meaning -t -i.
6457 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6458 domains, and warn by logging if others are found.
6460 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6461 returned from GIFCONF when Exim is trying to find the list of interfaces on
6462 a host. The code in os.c has been modified to copy each ifreq to an aligned
6463 structure in all cases.
6465 Also, in some cases, the returned ifreq's were being copied to a 'struct
6466 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6467 means the last couple of bytes of an IPv6 address could be chopped if the
6468 ifreq contained only a normal sockaddr (14 bytes storage).
6470 18. Named domain lists were not supported in the hosts_treat_as_local option.
6471 An entry such as +xxxx was not recognized, and was treated as a literal
6474 19. Ensure that header lines added by a DATA ACL are included in the reject log
6475 if the ACL subsequently rejects the message.
6477 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6478 MD5 (which is deprecated).
6480 21. When testing a filter file using -bf, Exim was writing a message when it
6481 took the sender from a "From " line in the message, but it was not doing so
6482 when it took $return_path from a Return-Path: header line. It now does.
6484 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6485 with a valid header line field name (a series of printing characters
6486 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6488 23. Changed "disc" in the source to "disk" to conform to the documentation and
6489 the book and for uniformity.
6491 24. Ignore Sendmail's -Ooption=value command line item.
6493 25. When execve() failed while trying to run a command in a pipe transport,
6494 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6495 could be confused with a return value of 69 from the command itself. This
6496 has been changed to 127, the value the shell returns if it is asked to run
6497 a non-existent command. The wording for the related log line suggests a
6498 non-existent command as the problem.
6500 26. If received_header_text expands to an empty string, do not add a Received:
6501 header line to the message. (Well, it adds a token one on the spool, but
6502 marks it "old" so that it doesn't get used or transmitted.)
6504 27. Installed eximstats 1.28 (addition of -nt option).
6506 28. There was no check for failure on the call to getsockname() in the daemon
6507 code. This can fail if there is a shortage of resources on the system, with
6508 ENOMEM, for example. A temporary error is now given on failure.
6510 29. Contrary to the C standard, it seems that in some environments, the
6511 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6512 program. Exim now does this explicitly; it affects the formatting of
6513 timestamps using strftime().
6515 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6516 complaints. I've now initialized all the variables, to avoid this.
6518 32. Header lines added by a system filter were not being "seen" during
6519 transport-time rewrites.
6521 33. The info_callback() function passed to OpenSSL is set up with type void
6522 (*)(SSL *, int, int), as described somewhere. However, when calling the
6523 function (actually a macro) that sets it up, the type void(*)() is
6524 expected. I've put in a cast to prevent warnings from picky compilers.
6526 34. If a DNS black list lookup found a CNAME record, but there were no A
6527 records associated with the domain it pointed at, Exim crashed.
6529 35. If a DNS black list lookup returned more than one A record, Exim ignored
6530 all but the first. It now scans all returned addresses if a particular IP
6531 value is being sought. In this situation, the contents of the
6532 $dnslist_value variable are a list of all the addresses, separated by a
6535 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6536 to accept a host name and all its aliases if the forward lookup for any of
6537 them yielded the IP address of the incoming connection. Now it accepts only
6538 those names whose forward lookup yields the correct IP address. Any other
6539 names are discarded. This closes a loophole whereby a rogue DNS
6540 administrator could create reverse DNS records to break through a
6541 wildcarded host restriction in an ACL.
6543 37. If a user filter or a system filter that ran in a subprocess used any of
6544 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6545 the wrong values were passed to the pipe command ($thisaddress had the
6546 value of $0, $0 had the value of $1, etc). This bug was introduced by
6547 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6549 38. Improved the line breaking for long SMTP error messages from ACLs.
6550 Previously, if there was no break point between 40 and 75 characters, Exim
6551 left the rest of the message alone. Two changes have been made: (a) I've
6552 reduced the minimum length to 35 characters; (b) if it can't find a break
6553 point between 35 and 75 characters, it looks ahead and uses the first one
6554 that it finds. This may give the occasional overlong line, but at least the
6555 remaining text gets split now.
6557 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6558 file descriptors might be low, and that setting 1000 would always raise it.
6559 It turns out that in some environments, the limit is already over 1000 and
6560 that lowering it causes trouble. So now Exim takes care not to decrease it.
6562 40. When delivering a message, the value of $return_path is set to $sender_
6563 address at the start of routing (routers may change the value). By an
6564 oversight, this default was not being set up when an address was tested by
6565 -bt or -bv, which affected the outcome if any router or filter referred to
6568 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6569 log only when "message" or "log_message" are set. However, if one of the
6570 conditions was an address verification, or a call to a nested ACL, the
6571 messages generated by the underlying test were being passed through. This
6572 no longer happens. The underlying message is available in $acl_verify_
6573 message for both "message" and "log_message" expansions, so it can be
6574 passed through if needed.
6576 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6577 new expansion $bh_ to give the encoded byte string without charset
6578 translation. Translation happens only if iconv() is available; HAVE_ICONV
6579 indicates this at build time. HEADERS_CHARSET gives the charset to
6580 translate to; headers_charset can change it in the configuration, and
6581 "headers charset" can change it in an individual filter file.
6583 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6584 that creates RFC 2047 encoded "words" labels them as that charset instead
6585 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6586 expansion operator; (ii) when Exim creates a From: line for a local
6587 message; (iii) when a header line is rewritten to include a "phrase" part.
6589 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6590 buggy, causing it to skip the first lines of messages whose message ID
6591 ended in 'D'. This would not have bitten before Exim release 4.14, because
6592 message IDs were unlikely to end in 'D' before then. The effect was to have
6593 incorrect size information for certain domains.
6595 45. #include "config.h" was missing at the start of the crypt16.c module. This
6596 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6599 46. If there was a timeout during a "random" callout check, Exim treated it as
6600 a failure of the random address, and carried on sending RSET and the real
6601 address. If the delay was just some slowness somewhere, the response to the
6602 original RCPT would be taken as a response to RSET and so on, causing
6603 mayhem of various kinds.
6605 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6606 when I implemented it. It didn't allow for the fact that some option values
6607 may legitimately be negative (e.g. size_addition), and it didn't even do
6608 the right test for positive values.
6610 48. Domain names in DNS records are case-independent. Exim always looks them up
6611 in lower case. Some resolvers return domain names in exactly the case they
6612 appear in the zone file, that is, they may contain uppercase letters. Not
6613 all resolvers do this - some return always lower case. Exim was treating a
6614 change of case by a resolver as a change of domain, similar to a widening
6615 of a domain abbreviation. This triggered its re-routing code and so it was
6616 trying to route what was effectively the same domain again. This normally
6617 caused routing to fail (because the router wouldn't handle the domain
6618 twice). Now Exim checks for this case specially, and just changes the
6619 casing of the domain that it ultimately uses when it transmits the message
6622 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6625 50. If a filter generated a file delivery with a non-absolute name (possible if
6626 no home directory exists for the router), the forbid_file option was not
6629 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6630 the existing equality matching.
6632 52. Exim was using ints instead of ino_t variables in some places where it was
6633 dealing with inode numbers.
6635 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6636 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6637 called TMPDIR, and if it finds it is different, it changes its value.
6639 54. The smtp_printf() function is now made available to local_scan() so
6640 additional output lines can be written before returning. There is also an
6641 smtp_fflush() function to enable the detection of a dropped connection.
6642 The variables smtp_input and smtp_batched_input are exported to
6645 55. Changed the default runtime configuration: the message "Unknown user"
6646 has been removed from the ACL, and instead placed on the localuser router,
6647 using the cannot_route_message feature. This means that any verification
6648 failures that generate their own messages won't get overridden. Similarly,
6649 the "Unrouteable address" message that was in the ACL for unverifiable
6650 relay addresses has also been removed.
6652 56. Added hosts_avoid_esmtp to the smtp transport.
6654 57. The exicyclog script was not checking for the esoteric option
6655 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6656 will work only if exicyclog is run under the appropriate euid.
6658 58. Following a discussion on the list, the rules by which Exim recognises line
6659 endings on incoming messages have been changed. The -dropcr and drop_cr
6660 options are now no-ops, retained only for backwards compatibility. The
6661 following line terminators are recognized: LF CRLF CR. However, special
6662 processing applies to CR:
6664 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6665 nor a local message in the state where . is a terminator.
6667 (ii) If a bare CR is encountered in a header line, an extra space is added
6668 after the line terminator so as not to end the header. The reasoning
6669 behind this is that bare CRs in header lines are most likely either
6670 to be mistakes, or people trying to play silly games.
6672 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6673 was being incorrectly given as 18 bytes larger than it should have been.
6674 This is a VOB (very old bug).
6676 60. This may never have affected anything current, but just in case it has:
6677 When the local host is found other than at the start of a list of hosts,
6678 the local host, those with the same MX, and any that follow, are discarded.
6679 When the list in question was part of a longer list of hosts, the following
6680 hosts (not currently being processed) were also being discarded. This no
6681 longer happens. I'm not sure if this situation could ever has previously
6684 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6687 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6688 header. This is something that is recommended in a new Internet Draft, and
6689 is something that is documented as being done by Sendmail. There are two
6690 possible values. For messages generated by the autoreply transport, Exim
6693 Auto-Submitted: auto-replied
6695 whereas for all other generated messages (e.g. bounces) it adds
6697 Auto-Submitted: auto-generated
6699 63. The "personal" condition in filters now includes a test for the
6700 Auto-Submitted: header. If it contains the string "auto-" the message it
6701 not considered personal.
6703 64. Added rcpt_include_affixes as a generic transport option.
6705 65. Added queue_only_override (default true).
6707 66. Added the syslog_duplication option.
6709 67. If what should have been the first header line of a message consisted of
6710 a space followed by a colon, Exim was mis-interpreting it as a header line.
6711 It isn't of course - it is syntactically invalid and should therefore be
6712 treated as the start of the message body. The misbehaviour could have
6713 caused a number of strange effects, including loss of data in subsequent
6714 header lines, and spool format errors.
6716 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6717 client host had authenticated. This control can now be exercised by an ACL
6718 for more flexibility.
6720 69. By default, callouts do not happen when testing with -bh. There is now a
6721 variant, -bhc, which does actually run the callout code, including
6722 consulting and updating the callout cache.
6724 70. Added support for saslauthd authentication, courtesy of Alexander
6727 71. If statvfs() failed on the spool or log directories while checking their
6728 size for availability, Exim confusingly gave the error "space shortage".
6729 Furthermore, in debugging mode it crashed with a floating point exception.
6730 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6731 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6732 serious problem, Exim now writes to the main and panic logs when this
6733 happens, with details of the failure. It then refuses to accept the
6734 incoming message, giving the message "spool directory problem" or "log
6735 directory problem" with a 421 code for SMTP messages.
6737 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6738 0, 1, and 2 exist, because some OS complain for execs without them (see
6739 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6740 descriptors. However, the code omitted to check that the open succeeded,
6741 causing mysterious errors if for some reason the permissions on /dev/null
6742 got screwed. Now Exim writes a message to the main and panic logs, and
6743 bombs out if it can't open /dev/null.
6745 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6746 interact so that it is all more flexible. It is supposed to remain
6747 backwards compatible. Also added extra_local_interfaces.
6749 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6750 to bomb out with an assertion failure - to the client this appears as a
6751 connection drop. This problem occurs in the part of the code that was taken
6752 from the Samba project. Fortunately, the assertion is in a very simple
6753 function, so I have fixed this by reproducing the function inline in the
6754 one place where it is called, and arranging for authentication to fail
6755 instead of killing the process with assert().
6757 75. The SPA client code was not working when the server requested OEM rather
6758 than Unicode encoding.
6760 76. Added code to make require_files with a specific uid setting more usable in
6761 the case where statting the file as root fails - usually a non-root-mounted
6762 NFS file system. When this happens and the failure is EACCES, Exim now
6763 forks a subprocess and does the per-uid checking as the relevant uid.
6765 77. Added process_log_path.
6767 78. If log_file_path was not explicitly set, a setting of check_log_space or
6768 check_log_inodes was ignored.
6770 79. If a space check for the spool or log partitions fails, the incident is now
6771 logged. Of course, in the latter case the data may get lost...
6773 80. Added the %p formatting code to string_format() so that it can be used to
6774 print addresses in debug_print(). Adjusted all the address printing in the
6775 debugging in store.c to use %p rather than %d.
6777 81. There was a concern that a line of code in smtp_in.c could overflow a
6778 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6779 initially expressed, the concern was not well-founded, because trailing
6780 spaces are removed early. However, if the trailing spaces were followed by
6781 a NULL, they did not get removed, so the overflow was possible. Two fixes
6784 (a) I re-wrote the offending code in a cleaner fashion.
6785 (b) If an incoming SMTP command contains a NULL character, it is rejected
6788 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6789 runs initgroups(), so that if the Exim user is in any additional groups,
6790 they will be used during message reception.
6796 The change log for 4.20 and earlier releases has been archived.