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.
8 Exim version 4.89+fixes
9 -----------------------
10 Cherry-Picked from the master development branch
11 -------------------------------------------------
13 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
15 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
16 Disable the session-cache too, which might reduce our load. Since we
17 currrectly use a new context for every connection, both as server and
18 client, there is no benefit for these.
19 GnuTLS appears to not support tickets server-side by default (we don't
20 call gnutls_session_ticket_enable_server()) but client side is enabled
21 by default on recent versions (3.1.3 +) unless the PFS priority string
24 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
25 the check for any unsuccessful recipients did not notice the limit, and
26 erroneously found still-pending ones.
28 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
29 compatible one, to avoid the (poorly documented) possibility of a config
30 file in the working directory redirecting the DB files, possibly correpting
31 some existing file. CVE-2017-10140 assigned for BDB.
33 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
34 cache-hot. Previously, although the result was properly cached, the
35 initial verify call returned a defer.
37 JH/21 Bug 2151 (partial):
38 Avoid using SIZE on the MAIL for a callout verify, on any but
39 the main verify for receipient in uncached-mode.
41 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
42 the right size to place the terminating semicolon on its own folded
43 line, the header hash was calculated to an incorrect value thanks to
44 the (relaxed) space the fold became.
46 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
47 which could crash as a result. This could lead to undeliverable messages.
49 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
50 are unsuable for non-message-specific reasons - eg. connection timeout,
51 banner-time rejection.
53 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
54 matters on fast-turnover and PID-randomising systems, which were getting
55 out-of-order delivery.
57 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
58 a possibly-overlapping copy. The symptom was that "Remote host closed
59 connection in response to HELO" was logged instead of the actual 4xx
66 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
67 than -2003 did; needs libidn2 in addition to libidn.
69 JH/02 The path option on a pipe transport is now expanded before use.
71 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
72 Patch provided by "Björn", documentation fix added too.
74 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
75 missing a wire-to-host endian conversion.
77 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
78 close after a BDAT command line could be taken as a following command,
79 giving a synch failure. Fix by only checking for synch immediately
80 before acknowledging the chunk.
82 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
83 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
84 macro. Patches provided by Josh Soref.
86 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
87 Previously we did not; the RFC seems ambiguous and VRFY is not listed
88 by IANA as a service extension. However, John Klensin suggests that we
91 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
92 the dkim code may be unix-mode line endings rather than smtp wire-format
93 CRLF, so prepend a CR to any bare LF.
95 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
96 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
98 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
99 the first were themselves being wrongly included in the feed into dkim
100 processing; with most chunk sizes in use this resulted in an incorrect
101 body hash calculated value.
103 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
104 DKIM signature block, for verification. Although advised against by
105 standards it is specifically not ruled illegal.
107 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
109 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
110 missing a body hash (the bh= tag).
112 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
113 It seems that HAProxy sends the Proxy Protocol information in clear and
114 only then does a TLS startup, so do the same.
116 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
117 TCP connections (such as for Spamd) unless the daemon successfully set
118 Fast Open mode on its listening sockets. This fixes breakage seen on
119 too-old kernels or those not configured for Fast Open, at the cost of
120 requiring both directions being enabled for TFO, and TFO never being used
121 by non-daemon-related Exim processes.
123 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
124 endings, at least on the first header line. Try to canonify any that get
125 past that check, despite the cost.
127 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
128 now limited to an arbitrary five deep, while parsing addresses with the
129 strip_excess_angle_brackets option enabled.
131 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
132 instead leave the unprompted TLS handshake in socket buffer for the
133 TLS library to consume.
135 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
137 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
139 JH/16 Drop variables when they go out of scope. Memory management drops a whole
140 region in one operation, for speed, and this leaves assigned pointers
141 dangling. Add checks run only under the testsuite which checks all
142 variables at a store-reset and panics on a dangling pointer; add code
143 explicitly nulling out all the variables discovered. Fixes one known
144 bug: a transport crash, where a dangling pointer for $sending_ip_address
145 originally assigned in a verify callout, is re-used.
147 PP/06 Drop '.' from @INC in various Perl scripts.
149 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
151 PP/08 Reduce a number of compilation warnings under clang; building with
152 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
153 should be warning-free.
155 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
157 HS/01 Fix portability problems introduced by PP/08 for platforms where
158 realloc(NULL) is not equivalent to malloc() [SunOS et al].
160 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
161 chunk. This allows us to accept broken chunked messages. We need a more
162 general solution here.
164 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
165 already-broken messages in the queue.
167 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
169 JH/19 Fix reference counting bug in routing-generated-address tracking.
175 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
176 supports it and a size is available (ie. the sending peer gave us one).
178 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
179 years of being deprecated). The replacements are the ACLs
180 acl_smtp_mime and acl_not_smtp_mime.
182 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
183 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
184 or even an in-clear connection were permitted. Now, if the host lookup
185 was dnssec and dane was requested then the host is only used if the
186 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
187 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
188 if one fails this test.
189 This means that a poorly-configured remote DNS will make it incommunicado;
190 but it protects against a DNS-interception attack on it.
192 JH/04 Bug 1810: make continued-use of an open smtp transport connection
193 non-noisy when a race steals the message being considered.
195 JH/05 If main configuration option tls_certificate is unset, generate a
196 self-signed certificate for inbound TLS connections.
198 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
199 in rewrites and routers.
201 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
202 and logged a warning sing 4.83; now they are a configuration file error.
204 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
205 (lacking @domain). Apply the same qualification processing as RCPT.
207 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
209 JH/10 Support ${sha256:} applied to a string (as well as the previous
212 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
213 a cutthrough deliver is pending, as we always want to make a connection.
214 This also avoids re-routing the message when later placing the cutthrough
215 connection after a verify cache hit.
216 Do not update it with the verify result either.
218 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
219 when routing results in more than one destination address.
221 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
222 signing (which inhibits the cutthrough capability). Previously only
223 the presence of an option was tested; now an expansion evaluating as
224 empty is permissible (obviously it should depend only on data available
225 when the cutthrough connection is made).
227 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
228 the relevant preceding SMTP command did not note the pipelining mode.
230 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
231 Previously they were not counted.
233 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
234 as one having no matching records. Previously we deferred the message
235 that needed the lookup.
237 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
238 distinguished as "(=".
240 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
241 for missing MX records. Previously it only worked for missing A records.
243 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
245 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
246 after the data-go-ahead and data-ack. Patch from Jason Betts.
248 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
249 even for a "none" policy. Patch from Tony Meyer.
251 JH/22 Fix continued use of a connection for further deliveries. If a port was
252 specified by a router, it must also match for the delivery to be
255 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
256 When one of the recipients of a message was unsuitable for the connection
257 (has no matching addresses), we lost track of needing to mark it
258 deferred. As a result mail would be lost.
260 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
262 JH/25 Decoding ACL controls is now done using a binary search; the source code
263 takes up less space and should be simpler to maintain. Merge the ACL
264 condition decode tables also, with similar effect.
266 JH/26 Fix problem with one_time used on a redirect router which returned the
267 parent address unchanged. A retry would see the parent address marked as
268 delivered, so not attempt the (identical) child. As a result mail would
271 JH/27 Fix a possible security hole, wherein a process operating with the Exim
272 UID can gain a root shell. Credit to http://www.halfdog.net/ for
273 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
276 JH/28 Enable {spool,log} filesystem space and inode checks as default.
277 Main config options check_{log,spool}_{inodes,space} are now
278 100 inodes, 10MB unless set otherwise in the configuration.
280 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
281 Previously it only applied to the main-section connection policy
284 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
286 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
287 by me. Added RFC7919 DH primes as an alternative.
289 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
290 are not in the system include path.
292 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
293 GnuTLS, when a session startup failed (eg because the client disconnected)
294 Exim did stdio operations after fclose. This was exposed by a recent
295 change which nulled out the file handle after the fclose.
297 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
298 signed directly by the cert-signing cert, rather than an intermediate
299 OCSP-signing cert. This is the model used by LetsEncrypt.
301 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
303 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
304 an incoming connection.
306 HS/02 Bug 1802: Do not half-close the connection after sending a request
309 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
310 fallback to "prime256v1".
312 JH/34 SECURITY: Use proper copy of DATA command in error message.
313 Could leak key material. Remotely exploitable. CVE-2016-9963.
319 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
320 and 3.4.4 - once the server is enabled to respond to an OCSP request
321 it does even when not requested, resulting in a stapling non-aware
322 client dropping the TLS connection.
324 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
325 support variable-length bit vectors. No functional change.
327 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
328 The I= interface field on outgoing lines is now after the H= remote
329 host field, same as incoming lines. There is a separate
330 outgoing_interface log selector which allows you to disable the
333 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
334 If not running log_selector +smtp_connection the mainlog would be held
335 open indefinitely after a "too many connections" event, including to a
336 deleted file after a log rotate. Leave the per net connection logging
337 leaving it open for efficiency as that will be quickly detected by the
338 check on the next write.
340 HS/01 Bug 1671: Fix post transport crash.
341 Processing the wait-<transport> messages could crash the delivery
342 process if the message IDs didn't exist for some reason. When
343 using 'split_spool_directory=yes' the construction of the spool
344 file name failed already, exposing the same netto behaviour.
346 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
347 mime_regex ACL conditions.
349 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
350 to DSN fail messages (bounces): remote IP, remote greeting, remote response
351 to HELO, local diagnostic string.
353 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
354 log line to debug. Even when configured with a tls authenticator many
355 client connections are expected to not authenticate in this way, so
356 an authenticate fail is not an error.
358 HS/02 Add the Exim version string to the process info. This way exiwhat
359 gives some more detail about the running daemon.
361 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
362 matter for fast-change records such as DNSBLs.
364 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
365 retry record, even if constant. There may be multiple transports with
366 different interface settings and the retry behaviour needs to be kept
369 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
371 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
373 JH/10 Bug 840: fix log_defer_output option of pipe transport
375 JH/11 Bug 830: use same host for all RCPTS of a message, even under
376 hosts_randomize. This matters a lot when combined with mua_wrapper.
378 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
379 ${quote_pgsql:<string>} operator.
381 JH/13 Bug 1708: avoid misaligned access in cached lookup.
383 JH/14 Change header file name for freeradius-client. Relevant if compiling
384 with Radius support; from the Gentoo tree and checked under Fedora.
386 JH/15 Bug 1712: Introduce $prdr_requested flag variable
388 JH/16 Bug 1714: Permit an empty string as expansion result for transport
389 option transport_filter, meaning no filtering.
391 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
393 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
394 defaults to "*" (all hosts). The variable is now available when not built
395 with TLS, default unset, mainly to enable keeping the testsuite sane.
396 If a server certificate is not supplied (via tls_certificate) an error is
397 logged, and clients will find TLS connections fail on startup. Presumably
398 they will retry in-clear.
399 Packagers of Exim are strongly encouraged to create a server certificate
400 at installation time.
402 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
403 with the $config_file variable.
405 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
406 in transport context, after the attempt, and per-recipient. The latter type
407 is per host attempted. The event data is the error message, and the errno
408 information encodes the lookup type (A vs. MX) used for the (first) host,
409 and the trailing two digits of the smtp 4xx response.
411 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
412 to write to mainlog (or rejectlog, paniclog) in the window between file
413 creation and permissions/ownership being changed. Particularly affects
414 installations where exicyclog is run as root, rather than exim user;
415 result is that the running daemon panics and dies.
417 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
419 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
420 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
421 "pri" and "weight". Note that the previous implicit priority given by the
422 list order is no longer honoured.
424 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
427 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
428 by defining SUPPORT_SOCKS.
430 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
431 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
432 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
433 variables are renamed to proxy_{local,external}_{address,port}.
435 JH/27 Move Internationalisation support from Experimental to mainline, enabled
436 for a build by defining SUPPORT_I18N
438 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
439 of the query string, and make ${quote_redis:} do that quoting.
441 JH/29 Move Events support from Experimental to mainline, enabled by default
442 and removable for a build by defining DISABLE_EVENT.
444 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
446 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
447 cached by the daemon.
449 JH/32 Move Redis support from Experimental to mainline, enabled for a build
450 by defining LOOKUP_REDIS. The libhiredis library is required.
452 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
453 keys are given for lookup.
455 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
456 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
457 only supported when built with TLS support. The PolarSSL SHA routines
458 are still used when the TLS library is too old for convenient support.
460 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
461 openssl_options), for security. OpenSSL forces this from version 1.1.0
462 server-side so match that on older versions.
464 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
465 allocation for $value could be released as the expansion processing
466 concluded, but leaving the global pointer active for it.
468 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
469 and to use the domains and local_parts ACL conditions.
471 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
472 incorrectly not doubled on cutthrough transmission, hence seen as a
473 body-termination at the receiving system - resulting in truncated mails.
474 Commonly the sender saw a TCP-level error, and retransmitted the message
475 via the normal store-and-forward channel. This could result in duplicates
476 received - but deduplicating mailstores were liable to retain only the
477 initial truncated version.
479 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
481 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
483 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
484 we're in there, support oversigning also; bug 1309.
486 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
488 HS/04 Add support for keep_environment and add_environment options.
490 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
491 either intentional arithmetic overflow during PRNG, or testing config-
494 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
495 delivery resulted in actual delivery. Cancel cutthrough before DATA
498 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
499 rejects a recipient: pass the reject to the originator.
501 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
502 Many were false-positives and ignorable, but it's worth fixing the
505 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
506 for the new environment-manipulation done at startup. Move the routines
507 from being local to tls.c to being global via the os.c file.
509 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
510 an extract embedded as result-arg for a map, the first arg for extract
511 is unavailable so we cannot tell if this is a numbered or keyed
512 extraction. Accept either.
518 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
521 JH/02 The smtp transport option "multi_domain" is now expanded.
523 JH/03 The smtp transport now requests PRDR by default, if the server offers
526 JH/04 Certificate name checking on server certificates, when exim is a client,
527 is now done by default. The transport option tls_verify_cert_hostnames
528 can be used to disable this per-host. The build option
529 EXPERIMENTAL_CERTNAMES is withdrawn.
531 JH/05 The value of the tls_verify_certificates smtp transport and main options
532 default to the word "system" to access the system default CA bundle.
533 For GnuTLS, only version 3.0.20 or later.
535 JH/06 Verification of the server certificate for a TLS connection is now tried
536 (but not required) by default. The verification status is now logged by
537 default, for both outbound TLS and client-certificate supplying inbound
540 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
543 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
544 Status Notification (bounce) messages are now MIME format per RFC 3464.
545 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
546 under the control of the dsn_advertise_hosts option, and routers may
547 have a dsn_lasthop option.
549 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
550 default, modifiable by a malware= option. The list separator for
551 the options can now be changed in the usual way. Bug 68.
553 JH/10 The smtp_receive_timeout main option is now expanded before use.
555 JH/11 The incoming_interface log option now also enables logging of the
556 local interface on delivery outgoing connections.
558 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
559 if the interface and destination host and port all match.
561 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
564 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
565 Patch from Andrew Lewis.
567 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
568 now supports optional time-restrictions, weighting, and priority
569 modifiers per server. Patch originally by <rommer@active.by>.
571 JH/16 The spamd_address main option now supports a mixed list of local
572 and remote servers. Remote servers can be IPv6 addresses, and
573 specify a port-range.
575 JH/17 Bug 68: The spamd_address main option now supports an optional
576 timeout value per server.
578 JH/18 Bug 1581: Router and transport options headers_add/remove can
579 now have the list separator specified.
581 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
584 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
587 JH/21 Support for the A6 type of dns record is withdrawn.
589 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
590 rather than the verbs used.
592 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
593 from 255 to 1024 chars.
595 JH/24 Verification callouts now attempt to use TLS by default.
597 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
598 are generic router options now. The defaults didn't change.
600 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
601 Original patch from Alexander Shikoff, worked over by JH.
603 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
604 files if ZCAT_COMMAND is not executable.
606 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
608 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
610 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
611 Normally benign, it bites when the pair was led to by a CNAME;
612 modern usage is to not canonicalize the domain to a CNAME target
613 (and we were inconsistent anyway for A-only vs AAAA+A).
615 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
617 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
618 when evaluating $sender_host_dnssec.
620 JH/31 Check the HELO verification lookup for DNSSEC, adding new
621 $sender_helo_dnssec variable.
623 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
625 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
627 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
629 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
630 documented as working, but never had. Support all but $spam_report.
632 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
633 added for tls authenticator.
635 HS/03 Add perl_taintmode main config option
641 TL/01 When running the test suite, the README says that variables such as
642 no_msglog_check are global and can be placed anywhere in a specific
643 test's script, however it was observed that placement needed to be near
644 the beginning for it to behave that way. Changed the runtest perl
645 script to read through the entire script once to detect and set these
646 variables, reset to the beginning of the script, and then run through
647 the script parsing/test process like normal.
649 TL/02 The BSD's have an arc4random API. One of the functions to induce
650 adding randomness was arc4random_stir(), but it has been removed in
651 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
652 function when detected.
654 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
655 cause callback expansion.
657 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
658 syntax errors in an expansion can be treated as a string instead of
659 logging or causing an error, due to the internal use of bool_lax
660 instead of bool when processing it.
662 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
663 server certificates when making smtp deliveries.
665 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
667 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
669 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
671 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
672 Merged patch from Sebastian Wiedenroth.
674 JH/05 Fix results-pipe from transport process. Several recipients, combined
675 with certificate use, exposed issues where response data items split
676 over buffer boundaries were not parsed properly. This eventually
677 resulted in duplicates being sent. This issue only became common enough
678 to notice due to the introduction of connection certificate information,
679 the item size being so much larger. Found and fixed by Wolfgang Breyha.
681 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
682 size buffer was used, resulting in syntax errors when an expansion
685 JH/07 Add support for directories of certificates when compiled with a GnuTLS
686 version 3.3.6 or later.
688 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
689 is EXPERIMENTAL_EVENT, the main-configuration and transport options
690 both become "event_action", the variables become $event_name, $event_data
691 and $event_defer_errno. There is a new variable $verify_mode, usable in
692 routers, transports and related events. The tls:cert event is now also
693 raised for inbound connections, if the main configuration event_action
696 TL/06 In test suite, disable OCSP for old versions of openssl which contained
697 early OCSP support, but no stapling (appears to be less than 1.0.0).
699 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
700 server certificate names available under the smtp transport option
701 "tls_verify_cert_hostname" now do not permit multi-component wildcard
704 JH/10 Time-related extraction expansions from certificates now use the main
705 option "timezone" setting for output formatting, and are consistent
706 between OpenSSL and GnuTLS compilations. Bug 1541.
708 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
709 encoded parameter in the incoming message. Bug 1558.
711 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
712 include certificate info, eximon was claiming there were spoolfile
715 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
717 JH/14 Log delivery-related information more consistently, using the sequence
718 "H=<name> [<ip>]" wherever possible.
720 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
721 are problematic for Debian distribution, omit them from the release
724 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
726 JH/16 Fix string representation of time values on 64bit time_t architectures.
729 JH/17 Fix a null-indirection in certextract expansions when a nondefault
730 output list separator was used.
735 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
736 checkers that were complaining about end of non-void function with no
739 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
740 This was a regression introduced in 4.83 by another bugfix.
742 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
744 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
745 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
751 TF/01 Correctly close the server side of TLS when forking for delivery.
753 When a message was received over SMTP with TLS, Exim failed to clear up
754 the incoming connection properly after forking off the child process to
755 deliver the message. In some situations the subsequent outgoing
756 delivery connection happened to have the same fd number as the incoming
757 connection previously had. Exim would try to use TLS and fail, logging
758 a "Bad file descriptor" error.
760 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
761 utilities have not been installed.
763 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
764 temporary space as the ACL may create new global variables.
766 TL/01 LDAP support uses per connection or global context settings, depending
767 upon the detected version of the libraries at build time.
769 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
770 to extract and use the src ip:port in logging and expansions as if it
771 were a direct connection from the outside internet. PPv2 support was
772 updated based on HAProxy spec change in May 2014.
774 JH/02 Add ${listextract {number}{list}{success}{fail}}.
776 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
777 Properly escape header and check for NULL return.
779 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
782 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
784 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
785 characters in header names, implemented as a verify condition.
786 Contributed by Michael Fischer v. Mollard.
788 TL/05 Rename SPF condition results err_perm and err_temp to standardized
789 results permerror and temperror. Previous values are deprecated but
790 still accepted. In a future release, err_perm and err_temp will be
791 completely removed, which will be a backward incompatibility if the
792 ACL tests for either of these two old results. Patch contributed by
793 user bes-internal on the mailing list.
795 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
797 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
798 selectors, in both main and reject logs.
800 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
803 JH/07 Add malware type "sock" for talking to simple daemon.
805 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
807 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
808 routers/transports under cutthrough routing.
810 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
811 numbers. Touch up "bool" conditional to keep the same definition.
813 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
815 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
817 JH/12 Expand items in router/transport headers_add or headers_remove lists
818 individually rather than the list as a whole. Bug 1452.
820 Required for reasonable handling of multiple headers_ options when
821 they may be empty; requires that headers_remove items with embedded
822 colons must have them doubled (or the list-separator changed).
824 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
825 view the policy declared in the DMARC record. Currently, $dmarc_status
826 is a combined value of both the record presence and the result of the
829 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
831 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
832 dnslookup router and the smtp transport (applying to the forward
835 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
836 of ldap servers used for a specific lookup. Patch provided by Heiko
839 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
840 New variable $lookup_dnssec_authenticated for observability.
842 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
843 Patch submitted by Lars Timman.
845 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
847 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
848 Requires trusted mode and valid format message id, aborts otherwise.
849 Patch contributed by Heiko Schlichting.
851 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
852 certextract with support for various fields. Bug 1358.
854 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
855 is requested by default, modifiable by smtp transport option
858 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
859 operate on certificate variables to give certificate fingerprints
860 Also new ${sha256:cert_variable}.
862 JH/23 The PRDR feature is moved from being Experimental into the mainline.
864 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
865 Christian Aistleitner.
867 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
869 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
870 file. Patch from Wolfgang Breyha.
872 JH/25 Expand the coverage of the delivery $host and $host_address to
873 client authenticators run in verify callout. Bug 1476.
875 JH/26 Port service names are now accepted for tls_on_connect_ports, to
876 align with daemon_smtp_ports. Bug 72.
878 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
879 support and error reporting did not work properly.
881 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
882 and is readable. Patch from Andrew Colin Kissa.
884 TL/14 Enhance documentation of ${run expansion and how it parses the
885 commandline after expansion, particularly in the case when an
886 unquoted variable expansion results in an empty value.
888 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
890 PP/02 Fix internal collision of T_APL on systems which support RFC3123
891 by renaming away from it. Addresses GH issue 15, reported by
894 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
896 TL/15 SECURITY: prevent double expansion in math comparison functions
897 (can expand unsanitized data). Not remotely exploitable.
904 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
906 PP/02 Make -n do something, by making it not do something.
907 When combined with -bP, the name of an option is not output.
909 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
912 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
913 $sender_host_name and config options to manage this, and basic check
916 PP/05 DSCP support for outbound connections and control modifier for inbound.
918 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
919 (Only plugin which currently uses this is kerberos4, which nobody should
920 be using, but we should make it available and other future plugins might
921 conceivably use it, even though it would break NAT; stuff *should* be
922 using channel bindings instead).
924 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
925 name; added for Sendmail compatibility; requires admin caller.
926 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
927 just ignore it); requires trusted caller.
928 Also parse but ignore: -Ac -Am -X<logfile>
931 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
933 TL/02 Add +smtp_confirmation as a default logging option.
935 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
936 Patch by Magnus Holmgren from 2007-02-20.
938 TL/04 Bugzilla 1281 - Spec typo.
939 Bugzilla 1283 - Spec typo.
940 Bugzilla 1290 - Spec grammar fixes.
942 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
944 TL/06 Add Experimental DMARC support using libopendmarc libraries.
946 TL/07 Fix an out of order global option causing a segfault. Reported to dev
947 mailing list by by Dmitry Isaikin.
949 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
951 JH/02 Support "G" suffix to numbers in ${if comparisons.
953 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
955 NM/01 Bugzilla 1197 - Spec typo
956 Bugzilla 1196 - Spec examples corrections
958 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
960 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
961 gnutls_enable_pkcs11, but renamed to more accurately indicate its
964 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
965 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
967 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
968 "acl {{name}{arg}...}", and optional args on acl condition
971 JH/05 Permit multiple router/transport headers_add/remove lines.
973 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
975 JH/07 Avoid using a waiting database for a single-message-only transport.
976 Performance patch from Paul Fisher. Bugzilla 1262.
978 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
981 JH/09 Add $headers_added variable, with content from use of ACL modifier
982 add_header (but not yet added to the message). Bugzilla 199.
984 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
985 Pulled from Bugzilla 817 by Wolfgang Breyha.
987 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
989 (nb: this is the same fix as in Exim 4.80.1)
991 JH/11 Add A= logging on delivery lines, and a client_set_id option on
994 JH/12 Add optional authenticated_sender logging to A= and a log_selector
997 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
999 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1000 advertise SMTP AUTH mechanism to us, instead of a generic
1001 protocol violation error. Also, make Exim more robust to bad
1002 data from the Dovecot auth socket.
1004 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1006 When a queue runner is handling a message, Exim first routes the
1007 recipient addresses, during which it prunes them based on the retry
1008 hints database. After that it attempts to deliver the message to
1009 any remaining recipients. It then updates the hints database using
1012 So if a recipient address works intermittently, it can get repeatedly
1013 deferred at routing time. The retry hints record remains fresh so the
1014 address never reaches the final cutoff time.
1016 This is a fairly common occurrence when a user is bumping up against
1017 their storage quota. Exim had some logic in its local delivery code
1018 to deal with this. However it did not apply to per-recipient defers
1019 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1021 This change adds a proper retry rule check during routing so that the
1022 final cutoff time is checked against the message's age. We only do
1023 this check if there is an address retry record and there is not a
1024 domain retry record; this implies that previous attempts to handle
1025 the address had the retry_use_local_parts option turned on. We use
1026 this as an approximation for the destination being like a local
1027 delivery, as in LMTP.
1029 I suspect this new check makes the old local delivery cutoff check
1030 redundant, but I have not verified this so I left the code in place.
1032 TF/02 Correct gecos expansion when From: is a prefix of the username.
1034 Test 0254 submits a message to Exim with the header
1038 When I ran the test suite under the user fanf2, Exim expanded
1039 the header to contain my full name, whereas it should have added
1040 a Resent-Sender: header. It erroneously treats any prefix of the
1041 username as equal to the username.
1043 This change corrects that bug.
1045 GF/01 DCC debug and logging tidyup
1046 Error conditions log to paniclog rather than rejectlog.
1047 Debug lines prefixed by "DCC: " to remove any ambiguity.
1049 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1051 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1052 Bug spotted by Jeremy Harris; was flawed since initial commit.
1053 Would have resulted in OCSP responses post-SNI triggering an Exim
1054 NULL dereference and crash.
1056 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1058 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1059 Bug detection, analysis and fix by Samuel Thibault.
1060 Bugzilla 1331, Debian bug #698092.
1062 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1064 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1065 Server implementation by Todd Lyons, client by JH.
1066 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1067 config variable "prdr_enable" controls whether the server
1068 advertises the facility. If the client requests PRDR a new
1069 acl_data_smtp_prdr ACL is called once for each recipient, after
1070 the body content is received and before the acl_smtp_data ACL.
1071 The client is controlled by both of: a hosts_try_prdr option
1072 on the smtp transport, and the server advertisement.
1073 Default client logging of deliveries and rejections involving
1074 PRDR are flagged with the string "PRDR".
1076 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1077 fclose(). Diagnosis by Todd Lyons.
1079 PP/17 Update configure.default to handle IPv6 localhost better.
1080 Patch by Alain Williams (plus minor tweaks).
1083 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1084 This is now consistent with GnuTLS, and is now documented: the
1085 previous undocumented portable approach to treating the option as
1086 unset was to force an expansion failure. That still works, and
1087 an empty string is now equivalent.
1089 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1090 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1091 not performing validation itself.
1093 PP/20 Added force_command boolean option to pipe transport.
1094 Patch from Nick Koston, of cPanel Inc.
1096 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1099 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1101 PP/21 Fix eximon continuous updating with timestamped log-files.
1102 Broken in a format-string cleanup in 4.80, missed when I repaired the
1103 other false fix of the same issue.
1104 Report and fix from Heiko Schlichting.
1107 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1108 Report from Prashanth Katuri.
1110 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1111 It's SecureTransport, so affects any MacOS clients which use the
1112 system-integrated TLS libraries, including email clients.
1114 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1115 using a MIME ACL for non-SMTP local injection.
1116 Report and assistance in diagnosis by Warren Baker.
1118 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1120 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1122 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1123 last id that failed so it may be referenced in subsequent ACL's.
1125 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1128 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1129 ldap library initialization, allowing self-signed CA's to be
1130 used. Also properly sets require_cert option later in code by
1131 using NULL (global ldap config) instead of ldap handle (per
1132 session). Bug diagnosis and testing by alxgomz.
1134 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1135 the src/util/ subdirectory.
1137 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1138 renamed to Transport Post Delivery Action by Jeremy Harris, as
1141 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1142 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1143 redis_servers = needs to be configured which will be used by the redis
1144 lookup. Patch from Warren Baker, of The Packet Hub.
1146 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1148 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1149 hostname or reverse DNS when processing a host list. Used suggestions
1150 from multiple comments on this bug.
1152 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1154 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1155 Missed a few lines, added it to make the runtest require no keyboard
1158 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1159 contains upper case chars. Make router use caseful_local_part.
1161 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1162 support when GnuTLS has been built with p11-kit.
1168 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1170 This, or similar/improved, will also be change PP/11 of 4.82.
1176 PP/01 Handle short writes when writing local log-files.
1177 In practice, only affects FreeBSD (8 onwards).
1178 Bugzilla 1053, with thanks to Dmitry Isaikin.
1180 NM/01 Bugzilla 949 - Documentation tweak
1182 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1185 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1187 PP/02 Implemented gsasl authenticator.
1189 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1191 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1192 `pkg-config foo` for cflags/libs.
1194 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1195 with rest of GSASL and with heimdal_gssapi.
1197 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1198 `pkg-config foo` for cflags/libs for the TLS implementation.
1200 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1201 properties get this fed in as external SSF. A number of robustness
1202 and debugging improvements to the cyrus_sasl authenticator.
1204 PP/08 cyrus_sasl server now expands the server_realm option.
1206 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1207 Patch by Jeremy Harris.
1209 PP/10 Added dbmjz lookup type.
1211 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1213 PP/12 MAIL args handles TAB as well as SP, for better interop with
1214 non-compliant senders.
1215 Analysis and variant patch by Todd Lyons.
1217 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1218 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1219 Patch from Dirk Mueller <dmueller@suse.com>
1221 PP/13 tls_peerdn now print-escaped for spool files.
1222 Observed some $tls_peerdn in wild which contained \n, which resulted
1223 in spool file corruption.
1225 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1226 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1227 or write after TLS renegotiation, which otherwise led to messages
1230 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1231 as a tracking header (ie: a signed header comes before the signature).
1232 Patch from Wolfgang Breyha.
1234 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1235 comma-sep list; embedded commas doubled.
1237 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1239 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1241 Report and patch from Dmitry Banschikov.
1243 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1244 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1245 failures. We appear to now support TLS1.1+ with Exim.
1247 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1248 lets Exim select keys and certificates based upon TLS SNI from client.
1249 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1250 before an outbound SMTP session. New log_selector, +tls_sni.
1252 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1253 NULL dereference. Report and patch from Alun Jones.
1255 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1256 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1257 Not seeing resolver debug output on NetBSD, but suspect this is a
1258 resolver implementation change.
1260 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1261 Left warnings. Added "eximon gdb" invocation mode.
1263 PP/21 Defaulting "accept_8bitmime" to true, not false.
1265 PP/22 Added -bw for inetd wait mode support.
1267 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1268 locate the relevant includes and libraries. Made this the default.
1270 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1271 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1273 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1274 This may cause build issues on older platforms.
1276 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1277 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1278 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1279 Added SNI support via GnuTLS too.
1280 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1282 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1284 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1285 Applied second patch from Janne, implementing suggestion to default
1286 multiple-strings-in-record handling to match SPF spec.
1288 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1290 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1291 read-only, out of scope).
1292 Patch from Wolfgang Breyha, report from Stuart Northfield.
1294 PP/29 Fix three issues highlighted by clang analyser static analysis.
1295 Only crash-plausible issue would require the Cambridge-specific
1296 iplookup router and a misconfiguration.
1297 Report from Marcin Mirosław.
1299 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1301 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1302 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1303 As part of this, removing so much warning spew let me fix some minor
1304 real issues in debug logging.
1306 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1307 assignment on my part. Fixed.
1309 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1310 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1311 Janne Snabb (who went above and beyond: thank you).
1313 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1314 string otherwise requires a connection and a bunch more work and it's
1315 relatively easy to get wrong. Should also expose TLS library linkage
1318 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1319 64-bit ${eval} (JH/03).
1321 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1322 GNU libc to support some of the 64-bit stuff, should not lead to
1323 conflicts. Defined before os.h is pulled in, so if a given platform
1324 needs to override this, it can.
1326 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1327 protection layer was required, which is not implemented.
1328 Bugzilla 1254, patch from Wolfgang Breyha.
1330 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1331 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1332 tls_dhparam take prime identifiers. Also unbreak combination of
1333 OpenSSL+DH_params+TLSSNI.
1335 PP/39 Disable SSLv2 by default in OpenSSL support.
1341 PP/01 Solaris build fix for Oracle's LDAP libraries.
1342 Bugzilla 1109, patch from Stephen Usher.
1344 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1346 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1349 TF/02 Fix a couple more cases where we did not log the error message
1350 when unlink() failed. See also change 4.74-TF/03.
1352 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1353 lock up or crash if it happened to be inside a call to libc when it
1354 got a SIGUSR1 from exiwhat.
1356 The SIGUSR1 handler appends the current process status to the process
1357 log which is later printed by exiwhat. It used to use the general
1358 purpose logging code to do this, but several functions it calls are
1359 not safe for signals.
1361 The new output code in the SIGUSR1 handler is specific to the process
1362 log, and simple enough that it's easy to inspect for signal safety.
1363 Removing some special cases also simplifies the general logging code.
1364 Removing the spurious timestamps from the process log simplifies
1367 TF/04 Improved ratelimit ACL condition.
1369 The /noupdate option has been deprecated in favour of /readonly which
1370 has clearer semantics. The /leaky, /strict, and /readonly update modes
1371 are mutually exclusive. The update mode is no longer included in the
1372 database key; it just determines when the database is updated. (This
1373 means that when you upgrade Exim will forget old rate measurements.)
1375 Exim now checks that the per_* options are used with an update mode that
1376 makes sense for the current ACL. For example, when Exim is processing a
1377 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1378 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1379 must specify per_mail/readonly. If you omit the update mode it defaults to
1380 /leaky where that makes sense (as before) or /readonly where required.
1382 The /noupdate option is now undocumented but still supported for
1383 backwards compatibility. It is equivalent to /readonly except that in
1384 ACLs where /readonly is required you may specify /leaky/noupdate or
1385 /strict/noupdate which are treated the same as /readonly.
1387 A useful new feature is the /count= option. This is a generalization
1388 of the per_byte option, so that you can measure the throughput of other
1389 aggregate values. For example, the per_byte option is now equivalent
1390 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1392 The per_rcpt option has been generalized using the /count= mechanism
1393 (though it's more complicated than the per_byte equivalence). When it is
1394 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1395 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1396 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1397 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1398 Note that using per_rcpt with a non-readonly update mode in more than
1399 one ACL will cause the recipients to be double-counted. (The per_mail
1400 and per_byte options don't have this problem.)
1402 The handling of very low rates has changed slightly. If the computed rate
1403 is less than the event's count (usually one) then this event is the first
1404 after a long gap. In this case the rate is set to the same as this event's
1405 count, so that the first message of a spam run is counted properly.
1407 The major new feature is a mechanism for counting the rate of unique
1408 events. The new per_addr option counts the number of different
1409 recipients that someone has sent messages to in the last time period. It
1410 behaves like per_rcpt if all the recipient addresses are different, but
1411 duplicate recipient addresses do not increase the measured rate. Like
1412 the /count= option this is a general mechanism, so the per_addr option
1413 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1414 example, measure the rate that a client uses different sender addresses
1415 with the options per_mail/unique=$sender_address. There are further
1416 details in the main documentation.
1418 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1420 TF/06 Removed a few PCRE remnants.
1422 TF/07 Automatically extract Exim's version number from tags in the git
1423 repository when doing development or release builds.
1425 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1426 Bugzilla 879. Patch from Paul Fisher.
1428 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1429 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1432 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1434 PP/05 Variable $av_failed, true if the AV scanner deferred.
1435 Bugzilla 1078. Patch from John Horne.
1437 PP/06 Stop make process more reliably on build failure.
1438 Bugzilla 1087. Patch from Heiko Schlittermann.
1440 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1441 Bugzilla 1089. Patch from Heiko Schlittermann.
1443 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1444 Bugzilla 1131. Patch from Holger Weiß.
1446 PP/09 Handle IPv6 addresses with SPF.
1447 Bugzilla 860. Patch from Wolfgang Breyha.
1449 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1451 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1454 PP/11 match_* no longer expand right-hand-side by default.
1455 New compile-time build option, EXPAND_LISTMATCH_RHS.
1456 New expansion conditions, "inlist", "inlisti".
1458 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1460 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1462 PP/14 fix log_write() format string regression from TF/03.
1463 Bugzilla 1152. Patch from Dmitry Isaikin.
1469 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1471 PP/02 Harmonised TLS library version reporting; only show if debugging.
1472 Layout now matches that introduced for other libraries in 4.74 PP/03.
1474 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1476 PP/04 New "dns_use_edns0" global option.
1478 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1481 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1482 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1484 TK/01 Updated PolarSSL code to 0.14.2.
1485 Bugzilla 1097. Patch from Andreas Metzler.
1487 PP/07 Catch divide-by-zero in ${eval:...}.
1488 Fixes bugzilla 1102.
1490 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1493 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1494 format-string attack -- SECURITY: remote arbitrary code execution.
1496 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1497 time unintentionally subject to list matching rules, letting the header
1498 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1499 arbitrary string expansion). This allowed for information disclosure.
1501 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1502 INT_MIN/-1 -- value coerced to INT_MAX.
1508 NM/01 Workaround for PCRE version dependency in version reporting
1511 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1512 This fixes portability to compilers other than gcc, notably
1513 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1515 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1516 makefiles for portability to HP-UX and POSIX correctness.
1518 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1519 Also via indented variable definition in the Makefile.
1520 (Debugging by Oliver Heesakkers).
1522 PP/02 Restore caching of spamd results with expanded spamd_address.
1523 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1525 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1526 Improves build reliability. Fix from: Frank Elsner
1528 NM/02 Fix wide character breakage in the rfc2047 coding
1529 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1531 NM/03 Allow underscore in dnslist lookups
1532 Fixes bug 1026. Patch from Graeme Fowler
1534 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1535 Code patches from Adam Ciarcinski of NetBSD.
1537 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1540 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1541 is logged, to avoid truncation. Patch from John Horne.
1543 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1544 Patch from Jakob Hirsch.
1546 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1547 SQL string expansion failure details.
1548 Patch from Andrey Oktyabrski.
1550 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1551 Patch from Simon Arlott.
1553 PP/09 New lookups functionality failed to compile on old gcc which rejects
1554 extern declarations in function scope.
1555 Patch from Oliver Fleischmann
1557 PP/10 Use sig_atomic_t for flags set from signal handlers.
1558 Check getgroups() return and improve debugging.
1559 Fixed developed for diagnosis in bug 927 (which turned out to be
1562 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1563 Patch from Mark Zealey.
1565 PP/12 Bugzilla 1056: Improved spamd server selection.
1566 Patch from Mark Zealey.
1568 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1569 Based on patch from Heiko Schlittermann.
1571 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1572 Patch from Uwe Doering, sign-off by Michael Haardt.
1574 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1575 variable declaration deep within a block. Bug and patch from
1578 PP/15 lookups-Makefile IRIX compatibility coercion.
1580 PP/16 Make DISABLE_DKIM build knob functional.
1582 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1583 Patch by Simon Arlott
1585 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1586 variable argument macros. Our copy now differs from upstream.
1592 TF/01 Failure to get a lock on a hints database can have serious
1593 consequences so log it to the panic log.
1595 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1596 controlled using the smtp_confirmation log selector.
1598 TF/03 Include the error message when we fail to unlink a spool file.
1600 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1601 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1602 for maintaining out-of-tree patches for some time.
1604 PP/01 Bugzilla 139: Documentation and portability issues.
1605 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1606 Handle per-OS dynamic-module compilation flags.
1608 PP/02 Let /dev/null have normal permissions.
1609 The 4.73 fixes were a little too stringent and complained about the
1610 permissions on /dev/null. Exempt it from some checks.
1611 Reported by Andreas M. Kirchwitz.
1613 PP/03 Report version information for many libraries, including
1614 Exim version information for dynamically loaded libraries. Created
1615 version.h, now support a version extension string for distributors
1616 who patch heavily. Dynamic module ABI change.
1618 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1619 privilege escalation vulnerability whereby the Exim run-time user
1620 can cause root to append content of the attacker's choosing to
1623 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1626 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1627 If dropping privileges for untrusted macros, we disabled normal logging
1628 on the basis that it would fail; for the Exim run-time user, this is not
1629 the case, and it resulted in successful deliveries going unlogged.
1630 Fixed. Reported by Andreas Metzler.
1636 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1637 only prepend for the Resent-* case. Fixes regression introduced in
1638 Exim 4.70 by NM/22 for Bugzilla 607.
1640 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1641 increasing numbers of administrators be bitten by this.
1643 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1645 PP/03 Bugzilla 994: added openssl_options main configuration option.
1647 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1649 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1651 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1653 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1654 without a peer certificate, leading to a segfault because of an
1655 assumption that peers always have certificates. Be a little more
1656 paranoid. Problem reported by Martin Tscholak.
1658 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1659 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1660 NB: ClamAV planning to remove STREAM in "middle of 2010".
1661 CL also introduces -bmalware, various -d+acl logging additions and
1662 more caution in buffer sizes.
1664 PP/09 Implemented reverse_ip expansion operator.
1666 PP/10 Bugzilla 937: provide a "debug" ACL control.
1668 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1670 PP/12 Bugzilla 973: Implement --version.
1672 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1674 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1676 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1678 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1679 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1680 ignore trailing whitespace.
1682 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1684 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1687 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1688 Notification from Dr Andrew Aitchison.
1690 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1691 ExtendedDetectionInfo response format.
1692 Notification from John Horne.
1694 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1697 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1698 XSL and documented dependency on system catalogs, with examples of how
1701 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1704 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1705 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1706 configuration file which is writeable by the Exim user or group.
1708 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1709 of configuration files to cover files specified with the -C option if
1710 they are going to be used with root privileges, not just the default
1713 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1714 option (effectively making it always true).
1716 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1717 files to be used while preserving root privileges.
1719 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1720 that rogue child processes cannot use them.
1722 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1723 run-time user, instead of root.
1725 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1726 Exim run-time user without dropping privileges.
1728 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1729 result string, instead of calling string_vformat() twice with the same
1732 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1733 for other users. Others should always drop root privileges if they use
1734 -C on the command line, even for a whitelisted configure file.
1736 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1738 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1744 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1745 $data_path, and $header_path variables; fixed documentation bugs and
1748 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1749 exipick to access non-standard spools, including the "frozen" queue
1752 NM/01 Bugzilla 965: Support mysql stored procedures.
1753 Patch from Alain Williams
1755 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1757 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1758 Patch from Andreas Metzler
1760 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1761 Patch from Kirill Miazine
1763 NM/05 Bugzilla 671: Added umask to procmail example.
1765 JJ/03 installed exipick 20100323.0, fixing doc bug
1767 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1768 directory. Notification and patch from Dan Rosenberg.
1770 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1772 TK/02 Improve log output when DKIM signing operation fails.
1774 MH/01 Treat the transport option dkim_domain as a colon separated
1775 list, not as a single string, and sign the message with each element,
1776 omitting multiple occurences of the same signer.
1778 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1779 Bugzilla 985, 986. Patch by Simon Arlott
1781 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1782 Patch by Simon Arlott
1784 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1785 MBX locking. Notification from Dan Rosenberg.
1791 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1793 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1795 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1797 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1799 NM/04 Bugzilla 915: Flag broken perl installation during build.
1805 TK/01 Added patch by Johannes Berg that expands the main option
1806 "spamd_address" if it starts with a dollar sign.
1808 TK/02 Write list of recipients to X-Envelope-Sender header when building
1809 the mbox-format spool file for content scanning (suggested by Jakob
1812 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1813 (http://www.dcc-servers.net/) support via dccifd. Activated by
1814 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1816 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1817 by Mark Daniel Reidel <mr@df.eu>.
1819 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1820 When building exim an external PCRE library is now needed -
1821 PCRE is a system library on the majority of modern systems.
1822 See entry on PCRE_LIBS in EDITME file.
1824 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1825 conversation. Added nologin parameter to request.
1826 Patch contributed by Kirill Miazine.
1828 TF/01 Do not log submission mode rewrites if they do not change the address.
1830 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1832 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1833 log files in place. Contributed by Roberto Lima.
1835 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1837 TF/03 Bugzilla 615: When checking the local_parts router precondition
1838 after a local_part_suffix or local_part_prefix option, Exim now
1839 does not use the address's named list lookup cache, since this
1840 contains cached lookups for the whole local part.
1842 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1843 Robert Millan. Documentation is in experimental-spec.txt.
1845 TF/04 Bugzilla 668: Fix parallel build (make -j).
1847 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1849 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1850 Patch provided by Jan Srzednicki.
1852 TF/05 Leading white space used to be stripped from $spam_report which
1853 wrecked the formatting. Now it is preserved.
1855 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1856 that they are available at delivery time.
1858 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1860 TF/08 TLS error reporting now respects the incoming_interface and
1861 incoming_port log selectors.
1863 TF/09 Produce a more useful error message if an SMTP transport's hosts
1864 setting expands to an empty string.
1866 NM/06 Bugzilla 744: EXPN did not work under TLS.
1867 Patch provided by Phil Pennock.
1869 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1870 Patch provided by Richard Godbee.
1872 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1873 acl_smtp_notquit, added index entry.
1875 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1876 Patch provided by Eugene Bujak.
1878 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1879 accept(). Patch provided by Maxim Dounin.
1881 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1882 Patch provided by Phil Pennock.
1884 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1886 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1887 Patch provided by Brad "anomie" Jorsch.
1889 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1890 Patch provided by Dean Brooks.
1892 TK/05 Add native DKIM support (does not depend on external libraries).
1894 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1895 Patch provided by Graeme Fowler.
1897 NM/16 Bugzilla 851: Documentation example syntax fix.
1899 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1901 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1904 NM/19 Bugzilla 745: TLS version reporting.
1905 Patch provided by Phil Pennock.
1907 NM/20 Bugzilla 167: bool: condition support.
1908 Patch provided by Phil Pennock.
1910 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1911 clients. Patch provided by Phil Pennock.
1913 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1914 Patch provided by Brad "anomie" Jorsch.
1916 NM/23 Bugzilla 687: Fix misparses in eximstats.
1917 Patch provided by Heiko Schlittermann.
1919 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1920 Patch provided by Heiko Schlittermann.
1922 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1923 plus update to original patch.
1925 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1927 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1928 Patch provided by David Brownlee.
1930 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1932 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1934 NM/30 Bugzilla 888: TLS documentation bugfixes.
1936 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1938 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1939 Unlike the original bugzilla I have changed all shell scripts in src tree.
1941 NM/33 Bugzilla 898: Transport filter timeout fix.
1942 Patch by Todd Rinaldo.
1944 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1945 Patch by Serge Demonchaux.
1947 NM/35 Bugzilla 39: Base64 decode bug fixes.
1948 Patch by Jakob Hirsch.
1950 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1952 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1954 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1956 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1962 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1963 ALT-N's libdkim that I have put here:
1964 http://duncanthrax.net/exim-experimental/
1966 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1967 were called 'true' and it seems that C99 defines that as a reserved
1968 keyword to be used with 'bool' variable types. That means you could
1969 not include C99-style headers which use bools without triggering
1970 build errors in sieve.c.
1972 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1973 as mailq or other aliases. Changed the --help handling significantly
1974 to do whats expected. exim_usage() emits usage/help information.
1976 SC/01 Added the -bylocaldomain option to eximstats.
1978 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1980 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1982 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1988 PH/01 Another patch from the Sieve maintainer.
1990 PH/02 When an IPv6 address is converted to a string for single-key lookup
1991 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1992 dots are used instead of colons so that keys in lsearch files need not
1993 contain colons. This was done some time before quoting was made available
1994 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1995 (notated using the quote facility) so as to distinguish them from IPv4
1996 keys. This meant that lookups for IP addresses in host lists did not work
1997 for iplsearch lookups.
1999 This has been fixed by arranging for IPv6 addresses to be expressed with
2000 colons if the lookup type is iplsearch. This is not incompatible, because
2001 previously such lookups could never work.
2003 The situation is now rather anomalous, since one *can* have colons in
2004 ordinary lsearch keys. However, making the change in all cases is
2005 incompatible and would probably break a number of configurations.
2007 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2010 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2011 conversion specification without a maximum field width, thereby enabling
2012 a rogue spamd server to cause a buffer overflow. While nobody in their
2013 right mind would setup Exim to query an untrusted spamd server, an
2014 attacker that gains access to a server running spamd could potentially
2015 exploit this vulnerability to run arbitrary code as the Exim user.
2017 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2018 $primary_hostname instead of what libspf2 thinks the hosts name is.
2020 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2021 a directory entry by the name of the lookup key. Previously, if a
2022 symlink pointed to a non-existing file or a file in a directory that
2023 Exim lacked permissions to read, a lookup for a key matching that
2024 symlink would fail. Now it is enough that a matching directory entry
2025 exists, symlink or not. (Bugzilla 503.)
2027 PH/03 The body_linecount and body_zerocount variables are now exported in the
2030 PH/04 Added the $dnslist_matched variable.
2032 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2033 This means they are set thereafter only if the connection becomes
2036 PH/06 Added the client_condition to authenticators so that some can be skipped
2037 by clients under certain conditions.
2039 PH/07 The error message for a badly-placed control=no_multiline_responses left
2040 "_responses" off the end of the name.
2042 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2044 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2045 (without spaces) instead of just copying the configuration text.
2047 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2049 PH/11 Added $max_received_linelength.
2051 PH/12 Added +ignore_defer and +include_defer to host lists.
2053 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2054 way in which PCRE > 7.0 is built.
2056 PH/14 Implemented queue_only_load_latch.
2058 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2059 MAIL command. The effect was to mangle the value on 64-bit systems.
2061 PH/16 Another patch from the Sieve maintainer.
2063 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2065 PH/18 If a system quota error occurred while trying to create the file for
2066 a maildir delivery, the message "Mailbox is full" was not appended to the
2067 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2068 only to a quota excession during the actual writing of the file.
2070 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2071 characters?) which causes problems in log lines. The DN values are now
2072 passed through string_printing() before being added to log lines.
2074 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2075 and InterBase are left for another time.)
2077 PH/21 Added message_body_newlines option.
2079 PH/22 Guard against possible overflow in moan_check_errorcopy().
2081 PH/23 POSIX allows open() to be a macro; guard against that.
2083 PH/24 If the recipient of an error message contained an @ in the local part
2084 (suitably quoted, of course), incorrect values were put in $domain and
2085 $local_part during the evaluation of errors_copy.
2091 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2092 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2095 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2096 issue a MAIL command.
2098 PH/02 In an ACL statement such as
2100 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2102 if a client was not listed at all, or was listed with a value other than
2103 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2104 the condition was not true (as it should be), so access was not denied.
2105 The bug was that the ! inversion was incorrectly passed on to the second
2106 item. This has been fixed.
2108 PH/03 Added additional dnslists conditions == and =& which are different from
2109 = and & when the dns lookup returns more than one IP address.
2111 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2112 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2114 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2115 FSYNC, which compiles an option called disable_fsync that allows for
2116 bypassing fsync(). The documentation is heavily laced with warnings.
2118 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2120 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2121 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2122 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2123 including adding "make clean"; (3) Added -fPIC when compiling the test
2124 dynamically loaded module, to get rid of a warning.
2126 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2127 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2128 = 0s. The bug is otherwise harmless.
2130 PH/07 There was a bug in the dovecot authenticator such that the value of
2131 $auth1 could be overwritten, and so not correctly preserved, after a
2132 successful authentication. This usually meant that the value preserved by
2133 the server_setid option was incorrect.
2135 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2137 PH/09 Installed PCRE release 7.0.
2139 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2140 run for batched SMTP input. It is now run at the start of every message
2141 in the batch. While fixing this I discovered that the process information
2142 (output by running exiwhat) was not always getting set for -bs and -bS
2143 input. This is fixed, and it now also says "batched" for BSMTP.
2145 PH/11 Added control=no_pipelining.
2147 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2148 patch, slightly modified), and move the expansion of helo_data till after
2149 the connection is made in the smtp transport (so it can use these
2152 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2154 PH/14 Added log_selector = +pid.
2156 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2158 PH/16 Add ${if forany and ${if forall.
2160 PH/17 Added dsn_from option to vary the From: line in DSNs.
2162 PH/18 Flush SMTP output before performing a callout, unless control =
2163 no_callout_flush is set.
2165 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2166 was true (the default) a successful delivery failed to delete the retry
2167 item, thus causing premature timeout of the address. The bug is now
2170 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2172 PH/21 Long custom messages for fakedefer and fakereject are now split up
2173 into multiline responses in the same way that messages for "deny" and
2174 other ACL rejections are.
2176 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2177 with slight modification.
2179 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2180 draft, changing the syntax and factoring some duplicate code".
2182 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2183 for deliveries of the second and subsequent messages over the same SMTP
2186 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2187 ${reduce, with only minor "tidies".
2189 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2191 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2192 expansion side effects.
2194 PH/27 When a message times out after an over-quota error from an Exim-imposed
2195 quota, the bounce message says "mailbox is full". This message was not
2196 being given when it was a system quota that was exceeded. It now should
2199 MH/03 Made $recipients available in local_scan(). local_scan() already has
2200 better access to the recipient list through recipients_list[], but
2201 $recipients can be useful in postmaster-provided expansion strings.
2203 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2204 in the case of a MAIL command with additional options following the
2205 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2206 were accidentally chopped off.
2208 PH/29 SMTP synchronization checks are implemented when a command is read -
2209 there is a check that no more input is waiting when there shouldn't be
2210 any. However, for some commands, a delay in an ACL can mean that it is
2211 some time before the response is written. In this time, more input might
2212 arrive, invalidly. So now there are extra checks after an ACL has run for
2213 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2214 pipelining has not been advertised.
2216 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2218 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2219 returned in the SMTP response when smtp_return_error_details was false.
2220 This has been fixed.
2222 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2223 instead of the C I/O that was originally supplied, because problems were
2224 reported on Solaris.
2226 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2227 Exim which did not show up earlier: it was assuming that a call to
2228 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2229 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2230 was a macro that became an assignment, so it seemed to work. This has
2231 changed to a proper function call with a void return, hence the compile
2232 error. Exim's code has been fixed.
2234 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2237 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2239 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2241 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2242 because it turns out to be too big or there is a timeout) there is no
2243 "Completed" line in the log. When some messages of this type were
2244 selected by exigrep, they were listed as "not completed". Others were
2245 picked up by some special patterns. I have improved the selection
2246 criteria to be more general.
2248 PH/38 The host_find_failed option in the manualroute router can now be set
2249 to "ignore", to completely ignore a host whose IP address cannot be
2250 found. If all hosts are ignored, the behaviour is controlled by the new
2251 host_all_ignored option.
2253 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2254 homing or because of multiple MX records with /mx) generated more than
2255 one IP address, and the following item turned out to be the local host,
2256 all the secondary addresses of the first item were incorrectly removed
2257 from the list, along with the local host and any following hosts (which
2258 is what is supposed to happen).
2260 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2261 whoever called Exim into the -H file. In the case of the daemon it was
2262 behaving confusingly. When first started, it used values for whoever
2263 started the daemon, but after a SIGHUP it used the Exim user (because it
2264 calls itself on a restart). I have changed the code so that it now always
2267 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2268 message are rejected with the same error (e.g. no authentication or bad
2269 sender address), and a DATA command is nevertheless sent (as can happen
2270 with PIPELINING or a stupid MUA), the error message that was given to the
2271 RCPT commands is included in the rejection of the DATA command. This is
2272 intended to be helpful for MUAs that show only the final error to their
2275 PH/42 Another patch from the Sieve maintainer.
2277 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2278 Eximstats - Fixed some broken HTML links and added missing column headers
2280 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2281 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2283 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2285 PH/43 Yet another patch from the Sieve maintainer.
2287 PH/44 I found a way to check for a TCP/IP connection going away before sending
2288 the response to the final '.' that terminates a message, but only in the
2289 case where the client has not sent further data following the '.'
2290 (unfortunately, this is allowed). However, in many cases there won't be
2291 any further data because there won't be any more messages to send. A call
2292 to select() can be used: if it shows that the input is "ready", there is
2293 either input waiting, or the socket has been closed. An attempt to read
2294 the next input character can distinguish the two cases. Previously, Exim
2295 would have sent an OK response which the client would never have see.
2296 This could lead to message repetition. This fix should cure that, at
2297 least in a lot of common cases.
2299 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2300 advertised in response to EHLO.
2306 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2307 fixed by 4.65/MH/01 (is this a record?) are fixed:
2309 (i) An empty string was always treated as zero by the numeric comparison
2310 operators. This behaviour has been restored.
2312 (ii) It is documented that the numeric comparison operators always treat
2313 their arguments as decimal numbers. This was broken in that numbers
2314 starting with 0 were being interpreted as octal.
2316 While fixing these problems I realized that there was another issue that
2317 hadn't been noticed. Values of message_size_limit (both the global option
2318 and the transport option) were treated as octal if they started with 0.
2319 The documentation was vague. These values are now always treated as
2320 decimal, and I will make that clear in the documentation.
2326 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2327 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2330 MH/01 Don't check that the operands of numeric comparison operators are
2331 integers when their expansion is in "skipping" mode (fixes bug
2332 introduced by 4.64-PH/07).
2334 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2335 child addresses, Exim now panics and dies. Previously, because the count
2336 is held in a short int, deliveries were likely to be lost. As such a
2337 large number of recipients for a single message is ridiculous
2338 (performance will be very, very poor), I have chosen to impose a limit
2339 rather than extend the field.
2345 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2346 leftover -K file (the existence of which was triggered by #402).
2347 While we were at it, introduced process PID as part of the -K
2348 filename. This should rule out race conditions when creating
2351 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2352 processing considerably. Previous code took too long for large mails,
2353 triggering a timeout which in turn triggers #401.
2355 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2356 in the DK code in transports.c. sendfile() is not really portable,
2357 hence the _LINUX specificness.
2359 TF/01 In the add_headers option to the mail command in an Exim filter,
2360 there was a bug that Exim would claim a syntax error in any
2361 header after the first one which had an odd number of characters
2364 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2365 callout verification, Exim cached a "reject" for the entire domain. This
2366 is correct for most verifications, but it is not correct for a recipient
2367 verification with use_sender or use_postmaster set, because in that case
2368 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2369 case of MAIL FROM:<> rejection from other early rejections (e.g.
2370 rejection of HELO). When verifying a recipient using a non-null MAIL
2371 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2372 Whatever the result of the callout, the value of the domain cache is
2373 left unchanged (for any other kind of callout, getting as far as trying
2374 RCPT means that the domain itself is ok).
2376 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2379 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2380 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2381 session, the daemon ignores ECONNECT errors and logs others; it now
2382 ignores EPIPE as well.
2384 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2385 (quoted-printable decoding).
2387 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2388 later the small subsequent patch to fix an introduced bug.
2390 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2392 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2394 PH/08 An error is now given if message_size_limit is specified negative.
2396 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2397 to be given (somewhat) arbitrary names.
2399 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2402 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2403 miscellaneous code fixes
2405 PH/10 Added the log_reject_target ACL modifier to specify where to log
2408 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2409 hostname. This is wrong, because it relates to the incoming message (and
2410 probably the interface on which it is arriving) and not to the outgoing
2411 callout (which could be using a different interface). This has been
2412 changed to use the value of the helo_data option from the smtp transport
2413 instead - this is what is used when a message is actually being sent. If
2414 there is no remote transport (possible with a router that sets up host
2415 addresses), $smtp_active_hostname is used.
2417 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2418 tweaks were necessary in order to get it to work (see also 21 below):
2419 (a) The code assumed that strncpy() returns a negative number on buffer
2420 overflow, which isn't the case. Replaced with Exim's string_format()
2422 (b) There were several signed/unsigned issues. I just did the minimum
2423 hacking in of casts. There is scope for a larger refactoring.
2424 (c) The code used strcasecmp() which is not a standard C function.
2425 Replaced with Exim's strcmpic() function.
2426 (d) The code set only $1; it now sets $auth1 as well.
2427 (e) A simple test gave the error "authentication client didn't specify
2428 service in request". It would seem that Dovecot has changed its
2429 interface. Fortunately there's a specification; I followed it and
2430 changed what the client sends and it appears to be working now.
2432 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2435 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2436 address A is aliased to B and C, where B exists and C does not. Without
2437 -v the output is "A verified" because verification stops after a
2438 successful redirection if more than one address is generated. However,
2439 with -v the child addresses are also verified. Exim was outputting "A
2440 failed to verify" and then showing the successful verification for C,
2441 with its parentage. It now outputs "B failed to verify", showing B's
2442 parentage before showing the successful verification of C.
2444 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2445 look up a TXT record in a specific list after matching in a combined
2448 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2449 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2450 they consult the DNS. I had assumed they would set it the way they
2451 wanted; and indeed my experiments on Linux seem to show that in some
2452 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2453 To be on the safe side, however, I have now made the interface to
2454 host_find_byname() similar to host_find_bydns(), with an argument
2455 containing the DNS resolver options. The host_find_byname() function now
2456 sets these options at its start, just as host_find_bydns() does. The smtp
2457 transport options dns_qualify_single and dns_search_parents are passed to
2458 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2459 of host_find_byname() use the default settings of RES_DEFNAMES
2460 (qualify_single) but not RES_DNSRCH (search_parents).
2462 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2463 spool_read_header() do less string testing, by means of a preliminary
2464 switch on the second character of optional "-foo" lines. (This is
2465 overdue, caused by the large number of possibilities that now exist.
2466 Originally there were few.) While I was there, I also converted the
2467 str(n)cmp tests so they don't re-test the leading "-" and the first
2468 character, in the hope this might squeeze out yet more improvement.
2470 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2471 flag allowing group syntax was set by the header_syntax check but not
2472 turned off, possible causing trouble later; (2) The flag was not being
2473 set at all for the header_verify test, causing "group"-style headers to
2474 be rejected. I have now set it in this case, and also caused header_
2475 verify to ignore an empty address taken from a group. While doing this, I
2476 came across some other cases where the code for allowing group syntax
2477 while scanning a header line wasn't quite right (mostly, not resetting
2478 the flag correctly in the right place). These bugs could have caused
2479 trouble for malformed header lines. I hope it is now all correct.
2481 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2482 with the "reply" argument non-NULL. The code, however (which originally
2483 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2484 but it didn't always do it. This confused somebody who was copying the
2485 code for some other use. I have removed all the tests.
2487 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2488 feature that was used to support insecure browsers during the U.S. crypto
2489 embargo. It requires special client support, and Exim is probably the
2490 only MTA that supported it -- and would never use it because real RSA is
2491 always available. This code has been removed, because it had the bad
2492 effect of slowing Exim down by computing (never used) parameters for the
2493 RSA_EXPORT functionality.
2495 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2496 authenticator to fail if there's a tab character in the incoming data
2497 (there should never be unless someone is messing about, as it's supposed
2498 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2499 if the connection is using TLS or if the remote IP is the same as the
2500 local IP, and the "valid-client-cert option" if a client certificate has
2503 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2504 authenticators. This can be used for authorization after authentication
2505 succeeds. (In the case of plaintext, it servers for both authentication
2508 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2509 if any retry times were supplied.
2511 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2512 connection, where there is no client IP address to check. In this
2513 situation, the verify now always succeeds.
2515 PH/25 Applied John Jetmore's -Mset patch.
2517 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2519 PH/27 In a string expansion for a processed (not raw) header when multiple
2520 headers of the same name were present, leading whitespace was being
2521 removed from all of them, but trailing whitespace was being removed only
2522 from the last one. Now trailing whitespace is removed from each header
2523 before concatenation. Completely empty headers in a concatenation (as
2524 before) are ignored.
2526 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2527 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2529 PH/29 [Removed. This was a change that I later backed out, and forgot to
2530 correct the ChangeLog entry (that I had efficiently created) before
2531 committing the later change.]
2533 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2534 address errors (4xx response to RCPT) over the same connection as other
2535 messages routed to the same hosts. Such deliveries are always "forced",
2536 so retry times are not inspected. This resulted in far too many retries
2537 for the affected addresses. The effect occurred only when there were more
2538 hosts than the hosts_max_try setting in the smtp transport when it had
2539 the 4xx errors. Those hosts that it had tried were not added to the list
2540 of hosts for which the message was waiting, so if all were tried, there
2541 was no problem. Two fixes have been applied:
2543 (i) If there are any address or message errors in an SMTP delivery, none
2544 of the hosts (tried or untried) are now added to the list of hosts
2545 for which the message is waiting, so the message should not be a
2546 candidate for sending over the same connection that was used for a
2547 successful delivery of some other message. This seems entirely
2548 reasonable: after all the message is NOT "waiting for some host".
2549 This is so "obvious" that I'm not sure why it wasn't done
2550 previously. Hope I haven't missed anything, but it can't do any
2551 harm, as the worst effect is to miss an optimization.
2553 (ii) If, despite (i), such a delivery is accidentally attempted, the
2554 routing retry time is respected, so at least it doesn't keep
2555 hammering the server.
2557 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2558 in ${readsocket because some servers need this prod.
2560 PH/32 Added some extra debug output when updating a wait-xxx database.
2562 PH/33 The hint "could be header name not terminated by colon", which has been
2563 given for certain expansion errors for a long time, was not being given
2564 for the ${if def:h_colon_omitted{... case.
2566 PH/34 The spec says: "With one important exception, whenever a domain list is
2567 being scanned, $domain contains the subject domain." There was at least
2568 one case where this was not true.
2570 PH/35 The error "getsockname() failed: connection reset by peer" was being
2571 written to the panic log as well as the main log, but it isn't really
2572 panic-worthy as it just means the connection died rather early on. I have
2573 removed the panic log writing for the ECONNRESET error when getsockname()
2576 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2577 runs only) independently of the message's sender address. This meant
2578 that, if the 4xx error was in fact related to the sender, a different
2579 message to the same recipient with a different sender could confuse
2580 things. In particular, this can happen when sending to a greylisting
2581 server, but other circumstances could also provoke similar problems.
2582 I have changed the default so that the retry time for these errors is now
2583 based a combination of the sender and recipient addresses. This change
2584 can be overridden by setting address_retry_include_sender=false in the
2587 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2588 remote server are returned as part of bounce messages. This was not
2589 happening for LMTP over a pipe (the lmtp transport), but now it is the
2590 same for both kinds of LMTP.
2592 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2593 in header lines that were in fact CNAMEs. This is no longer the case.
2595 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2596 and queue runs started by the daemon processed all messages. This has
2597 been fixed so that -R and -S can now usefully be given with -q<time>.
2599 PH/40 Import PCRE release 6.7 (fixes some bugs).
2601 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2603 PH/42 Give an error if -q is specified more than once.
2605 PH/43 Renamed the variables $interface_address and $interface_port as
2606 $received_ip_address and $received_port, to make it clear that these
2607 values apply to message reception, and not to the outgoing interface when
2608 a message is delivered. (The old names remain recognized, of course.)
2610 PH/44 There was no timeout on the connect() call when using a Unix domain
2611 socket in the ${readsocket expansion. There now is.
2613 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2614 be meaningful with "accept".
2616 SC/01 Eximstats V1.43
2617 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2619 SC/02 Eximstats V1.44
2620 Use a glob alias rather than an array ref in the generated
2621 parser. This improves both readability and performance.
2623 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2624 Collect SpamAssassin and rejection statistics.
2625 Don't display local sender or destination tables unless
2626 there is data to show.
2627 Added average volumes into the top table text output.
2629 SC/04 Eximstats V1.46
2630 Collect data on the number of addresses (recipients)
2631 as well as the number of messages.
2633 SC/05 Eximstats V1.47
2634 Added 'Message too big' to the list of mail rejection
2635 reasons (thanks to Marco Gaiarin).
2637 SC/06 Eximstats V1.48
2638 Mainlog lines which have GMT offsets and are too short to
2639 have a flag are now skipped.
2641 SC/07 Eximstats V1.49 (Alain Williams)
2642 Added the -emptyok flag.
2644 SC/08 Eximstats V1.50
2645 Fixes for obtaining the IP address from reject messages.
2647 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2648 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2649 whitespace changes from 4.64-PH/27
2651 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2654 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2655 are found, allow negative numbers in numeric criteria)
2657 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2659 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2660 to match changes made in 4.64-PH/43
2662 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2664 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2665 "domain" is actually an IP address, and force a failure. This locks out
2666 those revolvers/nameservers that support "A-for-A" lookups, in
2667 contravention of the specifications.
2669 PH/48 When a host name was looked up from an IP address, and the subsequent
2670 forward lookup of the name timed out, the host name was left in
2671 $sender_host_name, contrary to the specification.
2673 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2674 restricted to single-key lookups, Exim was not diagnosing an error if
2675 * or *@ was used with a query-style lookup.
2677 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2679 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2680 long ago, but noone interested enough thought of it. Let's just say that
2681 the "1.1" means that there are some new functions that weren't there at
2682 some point in the past.
2684 PH/51 Error processing for expansion failure of helo_data from an smtp
2685 transport during callout processing was broken.
2687 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2688 tested/used via the -bh/-bhc/-bs options.
2690 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2691 bug, fixed in subsequent PCRE releases).
2693 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2694 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2696 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2702 SC/01 Use a glob alias rather than an array ref in eximstats generated
2703 parser. This improves both readability and performance.
2705 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2706 Don't display local sender or destination tables in eximstats unless
2707 there is data to show.
2708 Added average volumes into the eximstats top table text output.
2710 SC/03 Collect data on the number of addresses (recipients) as well
2711 as the number of messages in eximstats.
2713 TF/01 Correct an error in the documentation for the redirect router. Exim
2714 does (usually) call initgroups() when daemonizing.
2716 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2717 with consistent privilege compared to when running as a daemon.
2719 TF/03 Note in the spec that $authenticated_id is not set for local
2720 submissions from trusted users.
2722 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2723 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2725 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2726 by adding some example configuration directives to the default
2727 configuration file. A little bit of work is required to uncomment the
2728 directives and define how usernames and passwords are checked, but
2729 there is now a framework to start from.
2731 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2732 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2733 without this. I don't know how relevant this is to other LDAP libraries.
2735 PH/02 Add the verb name to the "unknown ACL verb" error.
2737 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2739 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2741 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2742 directory not expanded when it should be if an expanded home directory
2743 was set for the address (which is overridden by the transport).
2745 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2748 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2749 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2750 because it is too late at that time, and has no effect.
2752 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2753 security issue with \' (bugzilla #107). I could not use the
2754 PQescapeStringConn() function, because it needs a PGconn value as one of
2757 PH/08 When testing addresses using -bt, indicate those final addresses that
2758 are duplicates that would not cause an additional delivery. At least one
2759 person was confused, thinking that -bt output corresponded to deliveries.
2760 (Suppressing duplicates isn't a good idea as you lose the information
2761 about possibly different redirections that led to the duplicates.)
2763 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2764 systems where poll() doesn't work, in particular OS X.
2766 PH/10 Added more information to debugging output for retry time not reached.
2768 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2769 operations in malware.c.
2771 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2774 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2775 syslog_duplication set false, log lines that would normally be written
2776 both the the main log and to the reject log were not written to syslog at
2779 PH/14 In the default configuration, change the use of "message" in ACL warn
2780 statements to "add_header".
2782 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2783 not followed by a command (e.g. "seen endif").
2785 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2786 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2789 PH/17 Added extra conditions to the default value of delay_warning_condition
2793 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2794 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2795 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2798 The Auto-Submitted: and various List- headers are standardised, whereas I
2799 don't think Precedence: ever was.
2801 PH/18 Refactored debugging code in route_finduser() to show more information,
2802 in particular, the error code if getpwnam() issues one.
2804 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2805 This is apparently needed in addition to the PH/07 change above to avoid
2806 any possible encoding problems.
2808 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2809 but not after initializing Perl.
2811 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2812 output them only if debugging. By default they are written stderr,
2813 apparently, which is not desirable.
2815 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2818 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2821 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2823 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2824 authenticated or an ident call has been made. Suppress the default
2825 values for $authenticated_id and $authenticated_sender (but permit -oMai
2826 and -oMas) when testing with -bh.
2828 PH/24 Re-jigged the order of the tests in the default configuration so that the
2829 tests for valid domains and recipients precede the DNS black list and CSA
2830 tests, on the grounds that those ones are more expensive.
2832 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2833 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2834 This bug exists in every version of Exim that I still have, right back to
2837 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2838 However, an attempt to turn on case-sensitivity in a regex key by
2839 including (?-i) didn't work because the subject string was already
2840 lowercased, and the effects were non-intuitive. It turns out that a
2841 one-line patch can be used to allow (?-i) to work as expected.
2847 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2848 other effects) broke the use of negated acl sub-conditions.
2850 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2853 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2854 "Deny" causes Exim to reject the incoming connection with a 554 error.
2855 Unfortunately, if there is a major crisis, such as a disk failure,
2856 tcp-wrappers gives "deny", whereas what one would like would be some
2857 kind of temporary error. A kludge has been added to help with this.
2858 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2859 554 error is used if errno is still zero or contains ENOENT (which occurs
2860 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2863 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2865 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2866 errors. Otherwise a message that provokes a temporary error (when other
2867 messages do not) can cause a whole host to time out.
2869 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2870 addresses were routed directly to files or pipes from a redirect router.
2871 File deliveries just didn't batch; pipe deliveries might have suffered
2874 PH/06 A failure to get a lock for a hints database would erroneously always say
2875 "Failed to get write lock", even when it was really a read lock.
2877 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2878 of 0600. This has been changed to use the value of the lockfile_mode
2879 option (which defaults to 0600).
2881 PH/08 Applied small patch from the Sieve maintainer.
2883 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2884 folder from quota calculations, a direct delivery into this folder messed
2885 up the contents of the maildirsize file. This was because the regex was
2886 used only to exclude .Trash (or whatever) when the size of the mailbox
2887 was calculated. There was no check that a delivery was happening into an
2888 excluded directory. This bug has been fixed by ignoring all quota
2889 processing for deliveries into excluded directories.
2891 PH/10 Added the maildirfolder_create_regex option to appendfile.
2897 PH/01 The code for finding all the local interface addresses on a FreeBSD
2898 system running IPv6 was broken. This may well have applied to all BSD
2899 systems, as well as to others that have similar system calls. The broken
2900 code found IPv4 interfaces correctly, but gave incorrect values for the
2901 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2902 that it would not match correctly against @[] and not recognize the IPv6
2905 PH/02 The ipliteral router was not recognizing addresses of the form user@
2906 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2908 PH/03 Added disable_ipv6.
2910 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2911 decoded form, because it is most often used to construct To: headers
2912 lines in autoreplies, and the decoded form may well be syntactically
2913 invalid. However, $reply_address has leading white space removed, and all
2914 newlines turned into spaces so that the autoreply transport does not
2917 PH/05 If group was specified without a user on a router, and no group or user
2918 was specified on a transport, the group from the router was ignored.
2920 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2921 for visible compile-time settings that can be used to change these
2922 numbers, for those that want even more. Backwards compatibility with old
2923 spool files has been maintained. However, going back to a previous Exim
2924 release will lost any variables that are in spool files.
2926 PH/07 Two small changes when running in the test harness: increase delay when
2927 passing a TCP/IP connection to a new process, in case the original
2928 process has to generate a bounce, and remove special handling of
2929 127.0.0.2 (sic), which is no longer necessary.
2931 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2932 be the same on different OS.
2934 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2937 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2938 whether --show-vars was specified or not
2940 JJ/02 exipick: Added support for new ACL variable spool format introduced
2943 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2944 syntactically invalid From: or Reply-to: line, and a filter used this to
2945 generate an autoreply, and therefore failed to obtain an address for the
2946 autoreply, Exim could try to deliver to a non-existent relative file
2947 name, causing unrelated and misleading errors. What now happens is that
2948 it logs this as a hard delivery error, but does not attempt to create a
2951 PH/11 The exinext utility has a -C option for testing purposes, but although
2952 the given file was scanned by exinext itself; it wasn't being passed on
2953 when Exim was called.
2955 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2956 an end-of-file indication when reading a command response.
2958 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2959 compiled. In many other places in Exim, IPv6 addresses are always
2960 recognized, so I have changed this. It also means that IPv4 domain
2961 literals of the form [IPV4:n.n.n.n] are now always recognized.
2963 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2964 used if the router is not running as root, for example, when verifying at
2965 ACL time, or when using -bh. The debugging output from this situation was
2966 non-existent - all you got was a failure to exec. I have made two
2969 (a) Failures to set uid/gid, the current directory, or a process leader
2970 in a subprocess such as that created by queryprogram now generate
2971 suitable debugging output when -d is set.
2973 (b) The queryprogram router detects when it is not running as root,
2974 outputs suitable debugging information if -d is set, and then runs
2975 the subprocess without attempting to change uid/gid.
2977 PH/15 Minor change to Makefile for building test_host (undocumented testing
2980 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2981 additional section of a DNS packet that returns MX or SRV records.
2982 Instead, it always explicitly searches for A/AAAA records. This avoids
2983 major problems that occur when a DNS server includes only records of one
2984 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2985 fixed another bug: if SRV records were looked up and the corresponding
2986 address records were *not* found in the additional section, the port
2987 values from the SRV records were lost.
2989 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2990 using the correct key (the original address) when searching the retry
2991 rules in order to find which one to use for generating the retry hint.
2993 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2994 adding the default one. Similarly, if it contains a Reply-To: header, the
2995 errors_reply_to option, if set, is not used.
2997 PH/19 When calculating a retry time, Exim used to measure the "time since
2998 failure" by looking at the "first failed" field in the retry record. Now
2999 it does not use this if it is later than than the arrival time of the
3000 message. Instead it uses the arrival time. This makes for better
3001 behaviour in cases where some deliveries succeed, thus re-setting the
3002 "first failed" field. An example is a quota failure for a huge message
3003 when small messages continue to be delivered. Without this change, the
3004 "time since failure" will always be short, possible causing more frequent
3005 delivery attempts for the huge message than are intended.
3006 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3008 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3009 $1, $2, $3) because the numerical variables can be reset during some
3010 expansion items (e.g. "match"), thereby losing the authentication data.
3012 PH/21 Make -bV show the size of off_t variables so that the test suite can
3013 decide whether to run tests for quotas > 2G.
3015 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3016 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3017 filecount value is greater than 2G or if a quota value is greater than 2G
3018 on a system where the size of off_t is not greater than 4, a panic error
3021 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3022 never match. The debug and -bh output now contains an explicit error
3023 message indicating a malformed IPv4 address or mask.
3025 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3026 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3027 PH/23 above applies.
3029 PH/25 Do not write to syslog when running in the test harness. The only
3030 occasion when this arises is a failure to open the main or panic logs
3031 (for which there is an explicit test).
3033 PH/26 Added the /no_tell option to "control=freeze".
3035 PH/27 If a host name lookup failed very early in a connection, for example, if
3036 the IP address matched host_lookup and the reverse lookup yielded a name
3037 that did not have a forward lookup, an error message of the form "no IP
3038 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3039 could be logged. Now it outputs the IP address instead of "NULL".
3041 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3042 allows the sender and the authenticated sender to be set when
3043 submitting a message from within Exim. Since child_open_exim() is
3044 documented for local_scan(), the new function should be too.
3046 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3047 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3048 results in an empty string is now treated as unset.
3050 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3052 PH/31 Added sender_verify_fail logging option.
3054 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3055 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3056 tidied the source and removed it altogether.
3058 PH/33 When a queue run was abandoned because the load average was too high, a
3059 log line was always written; now it is written only if the queue_run log
3060 selector is set. In addition, the log line for abandonment now contains
3061 information about the queue run such as the pid. This is always present
3062 in "start" and "stop" lines but was omitted from the "abandon" line.
3064 PH/34 Omit spaces between a header name and the colon in the error message that
3065 is given when verify = headers_syntax fails (if there are lots of them,
3066 the message gets confusing).
3068 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3069 names, as there are now some PTR records that contain slashes. This check
3070 is only to protect against broken name servers that fall over on strange
3071 characters, so the fact that it applies to all lookups doesn't matter.
3073 PH/36 Now that the new test suite is complete, we can remove some of the
3074 special code in Exim that was needed for the old test suite. For example,
3075 sorting DNS records because real resolvers return them in an arbitrary
3076 order. The new test suite's fake resolver always returns records in the
3079 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3080 bounces) except when queue_only is set, to avoid logging races between
3081 the different processes.
3083 PH/38 Panic-die if .include specifies a non-absolute path.
3085 PH/39 A tweak to the "H" retry rule from its user.
3087 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3088 a label. They prevented compilation on older perls.
3090 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3091 a warning to be raised on newish perls.
3093 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3094 on queue. Changes to match documented behaviour of showing count of
3095 messages matching specified criteria.
3097 PH/40 Changed the default ident timeout from 30s to 5s.
3099 PH/41 Added support for the use of login_cap features, on those BSD systems
3100 that have them, for controlling the resources used by pipe deliveries.
3102 PH/42 The content-scanning code uses fopen() to create files in which to put
3103 message data. Previously it was not paying any attention to the mode of
3104 the files. Exim runs with umask(0) because the rest of the code creates
3105 files with open(), and sets the required mode explicitly. Thus, these
3106 files were ending up world-writeable. This was not a big issue, because,
3107 being within the spool directory, they were not world-accessible. I have
3108 created a function called modefopen, which takes an additional mode
3109 argument. It sets umask(777), creates the file, chmods it to the required
3110 mode, then resets the umask. All the relevant calls to fopen() in the
3111 content scanning code have been changed to use this function.
3113 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3114 to 24 hours. This avoids potential overflow problems when processing G
3115 and H retry rules. I suspect nobody ever tinkers with this value.
3117 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3119 PH/45 When the plaintext authenticator is running as a client, the server's
3120 challenges are checked to ensure they are valid base64 strings. By
3121 default, the authentication attempt is cancelled if an invalid string is
3122 received. Setting client_ignore_invalid_base64 true ignores these errors.
3123 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3124 they are received. Thus, the responses can be made to depend on the
3125 challenges. If an invalid string is ignored, an empty string is placed in
3128 PH/46 Messages that are created by the autoreply transport now contains a
3129 References: header, in accordance with RFCs 2822 and 3834.
3131 PH/47 Added authenticated_sender_force to the smtp transport.
3133 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3135 PH/49 Installed latest patch from the Sieve maintainer.
3137 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3138 was also set, the appendfile transport was unnecessarily scanning a
3139 directory of message files (e.g. for maildir delivery) to find the count
3140 of files (along with the size), even though it did not need this
3141 information. It now does the scan only if it needs to find either the
3142 size of the count of files.
3144 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3146 PH/52 Two bugs concerned with error handling when the smtp transport is
3149 (i) Exim was not creating retry information for temporary errors given
3150 for individual recipients after the DATA command when the smtp transport
3151 was used in LMTP mode. This meant that they could be retried too
3152 frequently, and not timed out correctly.
3154 (ii) Exim was setting the flag that allows error details to be returned
3155 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3156 recipients that were returned after the DATA command.
3158 PH/53 This is related to PH/52, but is more general: for any failing address,
3159 when detailed error information was permitted to be returned to the
3160 sender, but the error was temporary, then after the final timeout, only
3161 "retry timeout exceeded" was returned. Now it returns the full error as
3162 well as "retry timeout exceeded".
3164 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3165 do this, and (what is worse) MTAs that accept it.
3167 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3168 will now be deprecated.
3170 PH/56 New os.c-cygwin from the Cygwin maintainer.
3172 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3173 formats (previously only available in exim formats via -bpr, -bpru,
3174 and -bpra. Now also available in native and exiqgrep formats)
3176 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3177 with very large, slow to parse queues
3179 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3181 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3183 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3184 responses to authentication challenges, though it was showing the
3185 challenges; (ii) I've removed the CR characters from the debug output for
3188 PH/58 Allow for the insertion of a newline as well as a space when a string
3189 is turned into more than one encoded-word during RFC 2047 encoding. The
3190 Sieve code now uses this.
3192 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3193 data_4xx, lost_connection, tls_required.
3195 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3196 message was being sent as an SMTP response.
3198 PH/61 Add -l and -k options to exicyclog.
3200 PH/62 When verifying, if an address was redirected to one new address, so that
3201 verification continued, and the new address failed or deferred after
3202 having set something in $address_data, the value of $address_data was not
3203 passed back to the ACL. This was different to the case when no
3204 redirection occurred. The value is now passed back in both cases.
3206 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3207 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3208 use login_cap.h, so on its own it isn't the distinguishing feature. The
3209 new name refers directly to the setclassresources() function.
3211 PH/65 Added configuration files for NetBSD3.
3213 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3215 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3216 is preferred over IPv4.
3218 PH/68 The bounce_return_message and bounce_return_body options were not being
3219 honoured for bounces generated during the reception of non-SMTP messages.
3220 In particular, this applied to messages rejected by the ACL. This bug has
3221 been fixed. However, if bounce_return_message is true and bounce_return_
3222 body is false, the headers that are returned for a non-SMTP message
3223 include only those that have been read before the error was detected.
3224 (In the case of an ACL rejection, they have all been read.)
3226 PH/69 The HTML version of the specification is now built in a directory called
3227 spec_html instead of spec.html, because the latter looks like a path with
3228 a MIME-type, and this confuses some software.
3230 PH/70 Catch two compiler warnings in sieve.c.
3232 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3233 function verify_get_ident() calls ip_connect() to connect a socket, but
3234 if the "connect()" function timed out, ip_connect() used to close the
3235 socket. However, verify_get_ident() also closes the socket later, and in
3236 between Exim writes to the log, which may get opened at this point. When
3237 the socket was closed in ip_connect(), the log could get the same file
3238 descriptor number as the socket. This naturally causes chaos. The fix is
3239 not to close the socket in ip_connect(); the socket should be closed by
3240 the function that creates it. There was only one place in the code where
3241 this was missing, in the iplookup router, which I don't think anybody now
3242 uses, but I've fixed it anyway.
3244 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3245 well as to direct DNS lookups. Otherwise the handling of names in host
3246 lists is inconsistent and therefore confusing.
3252 PH/01 Two changes to the default runtime configuration:
3254 (1) Move the checks for relay_from_hosts and authenticated clients from
3255 after to before the (commented out) DNS black list checks.
3257 (2) Add control=submission to the relay_from_hosts and authenticated
3258 clients checks, on the grounds that messages accepted by these
3259 statements are most likely to be submissions.
3261 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3263 (1) Generate an error if the third argument for the ${prvs expansion is
3266 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3269 (3) Reset the variables that are obtained from the first argument of
3270 ${prvscheck and used in the second argument before leaving the code,
3271 because their memory is reclaimed, so using them afterwards may do
3274 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3275 one (it's much easier than Tom thought :-).
3277 (5) Because of (4), we can now allow for the use of $prvscheck_result
3278 inside the third argument.
3280 PH/03 For some reason, the default setting of PATH when running a command from
3281 a pipe transport was just "/usr/bin". I have changed it to
3284 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3285 anything to be listed in the output from -bV.
3287 PH/05 When a filter generated an autoreply, the entire To: header line was
3288 quoted in the delivery log line, like this:
3290 => >A.N.Other <ano@some.domain> <original@ddress> ...
3292 This has been changed so that it extracts the operative address. There
3293 may be more than one such address. If so, they are comma-separated, like
3296 => >ano@some.domain,ona@other.domain <original@ddress> ...
3298 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3299 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3300 not being looked up in the rDNS to get a host name, Exim was showing the
3301 IP address twice in Received: lines, even though the IP addresses were
3302 identical. For example:
3304 Received: from [1.2.3.4] (helo=[1.2.3.4])
3306 However, if the real host name was known, it was omitting the HELO data
3307 if it matched the actual IP address. This has been tidied up so that it
3308 doesn't show the same IP address twice.
3310 PH/07 When both +timestamp and +memory debugging was on, the value given by
3311 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3312 called by the debug printing, thereby overwriting the timestamp buffer.
3313 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3315 PH/08 When the original message was included in an autoreply transport, it
3316 always said "this is a copy of the message, including all the headers",
3317 even if body_only or headers_only was set. It now gives an appropriate
3320 PH/09 Applied a patch from the Sieve maintainer which:
3322 o fixes some comments
3323 o adds the (disabled) notify extension core
3324 o adds some debug output for the result of if/elsif tests
3325 o points to the current vacation draft in the documentation
3326 and documents the missing references header update
3330 o fixes a bug in processing the envelope test (when testing
3331 multiple envelope elements, the last element determined the
3334 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3335 Electronic Mail") by including:
3337 Auto-submitted: auto-generated
3339 in the messages that it generates (bounce messages and others, such as
3340 warnings). In the case of bounce messages for non-SMTP messages, there was
3341 also a typo: it was using "Auto_submitted" (underscore instead of
3342 hyphen). Since every message generated by Exim is necessarily in response
3343 to another message, thes have all been changed to:
3345 Auto-Submitted: auto-replied
3347 in accordance with these statements in the RFC:
3349 The auto-replied keyword:
3351 - SHOULD be used on messages sent in direct response to another
3352 message by an automatic process,
3354 - MUST NOT be used on manually-generated messages,
3356 - MAY be used on Delivery Status Notifications (DSNs) and Message
3357 Disposition Notifications (MDNs),
3359 - MUST NOT be used on messages generated by automatic or periodic
3360 processes, except for messages which are automatic responses to
3363 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3364 to the default Received: header definition.
3366 PH/12 Added log selector acl_warn_skipped (default on).
3368 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3369 variables because (a) they are in the wrong storage pool and (b) even if
3370 they were copied, it wouldn't work properly because of the caching.
3372 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3373 checking when decoding. Apparently there are clients that generate
3374 overlong encoded strings. Why am I not surprised?
3376 PH/15 If the first argument of "${if match_address" was not empty, but did not
3377 contain an "@" character, Exim crashed. Now it writes a panic log message
3378 and treats the condition as false.
3380 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3382 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3383 extension "envelope-auth". The code is finished and in agreement with
3384 other implementations, but there is no documentation so far and in fact,
3385 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3386 not changing the active code.
3388 Print executed "if" and "elsif" statements when debugging is used. This
3389 helps a great deal to understand what a filter does.
3391 Document more things not specified clearly in RFC3028. I had all this
3392 sorted out, when out of a sudden new issues came to my mind. Oops."
3394 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3397 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3398 canonical form (as documented). However, after a host name lookup from
3399 the IP address, check_host() was doing a simple string comparison with
3400 addresses acquired from the DNS when checking that the found name did
3401 have the original IP as one of its addresses. Since any found IPv6
3402 addresses are likely to be in abbreviated form, the comparison could
3403 fail. Luckily, there already exists a function for doing the comparison
3404 by converting both addresses to binary, so now that is used instead of
3405 the text comparison.
3407 PH/20 There was another similar case to PH/19, when a complete host name was
3408 given in a host list; looking up its IP address could give an abbreviated
3409 form, whereas the current host's name might or might not be abbreviated.
3410 The same fix has been applied.
3416 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3417 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3420 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3421 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3423 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3425 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3426 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3427 Also, for "Auto-Submitted", treat anything other than "no" as
3428 non-personal, in accordance with RFC 3834. (Previously it treated
3429 anything starting "auto-" as non-personal.)
3431 TF/01 The control=submission/name=... option had a problem with syntax
3432 errors if the name included a slash character. The /name= option
3433 now slurps the rest of the string, so it can include any characters
3434 but it must come last in the list of options (after /sender_retain
3437 PH/05 Some modifications to the interface to the fake nameserver for the new
3445 TK/01 Added the "success_on_redirect" address verification option. See
3446 NewStuff for rationale and an example.
3448 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3450 PH/02 Patch to exigrep to allow it to work on syslog lines.
3452 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3453 fread() to skip over the body file's header line, because in Cygwin the
3454 header line is locked and is inaccessible.
3456 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3457 co-exist for some time) to make it clear that it is the Exim ID that is
3458 referenced, not the Message-ID: header line.
3460 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3461 string_format() function, because snprintf() does not exist on all
3464 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3465 ${stat: expansion item.
3467 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3468 protocol synchronization error", to keep the pedants happy.
3470 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3471 well as for IRIX systems, when gcc is being used. See the host.c source
3474 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3476 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3479 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3480 transport and to the smtp transport in LMTP mode.
3482 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3484 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3485 run a filter in a subprocess. This could lead to confusion in subsequent
3486 lookups in the parent process. There should also be a search_tidyup() at
3487 the end of the subprocess.
3489 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3490 only if the host matched helo_try_verify_hosts, which caused the
3491 verification to occur when the EHLO/HELO command was issued. The ACL just
3492 tested the remembered result. Now, if a previous verification attempt has
3493 not happened, "verify = helo" does it there and then.
3495 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3497 TK/03 Fix log output including CR from clamd.
3499 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3500 exist provoked a memory error which could cause a segfault.
3502 PH/15 Installed PCRE 6.2
3504 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3506 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3507 of the problem. Specifically, suggested +O2 rather than +O1 for the
3510 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3512 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3513 record, the port was not correctly logged when the outgoing_port log
3514 selector was set (it logged the transort's default port).
3516 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3517 fallback_hosts, and "hosts" in the smtp transport.
3519 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3520 host errors such as "Connection refused".
3522 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3523 authentication with radiusclient 0.4.9:
3525 - Error returned from rc_read_config was caught wrongly
3526 - Username/password not passed on to radius server due to wrong length.
3528 The presumption is that some radiusclient API changes for 4.51/PH/17
3529 were not taken care of correctly. The code is still untested by me (my
3530 Linux distribution still has 0.3.2 of radiusclient), but it was
3531 contributed by a Radius user.
3533 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3534 expanding the "port" option of the smtp transport.
3536 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3537 while reading a MIME header. Thanks to Tom Hughes for a patch.
3539 PH/24 Include config.h inside local_scan.h so that configuration settings are
3542 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3543 that in an ACL for RCPT (for example), you can examine exactly what was
3546 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3547 commands, but it was not correctly comparing the address with the actual
3548 client host address. Thus, it would show the EHLO address in Received:
3549 header lines when this was not necessary.
3551 PH/27 Added the % operator to ${eval:}.
3553 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3554 it should be ignoring failures (because with -C, for example, it has lost
3555 privilege). It wasn't ignoring creation failures other than "already
3558 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3561 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3562 because an input line was too long, either on its own, or by virtue of
3563 too many continuations, the temporary file was not being removed, and the
3564 return code was incorrect.
3566 PH/31 Missing "BOOL" in function definition in filtertest.c.
3568 PH/32 Applied Sieve patches from the maintainer.
3570 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3572 PH/33 Added "verify = not_blind".
3574 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3575 Local/Makefile (with some defaults set). These are used in built scripts
3576 such as exicyclog, but they have never been used in the exim_install
3577 script (though there are many overriding facilities there). I have
3578 arranged that the exim_install script now takes note of these two
3581 PH/35 Installed configuration files for Dragonfly.
3583 PH/36 When a locally submitted message by a trusted user did not contain a
3584 From: header, and the sender address was obtained from -f or from an SMTP
3585 MAIL command, and the trusted user did not use -F to supply a sender
3586 name, $originator_name was incorrectly used when constructing a From:
3587 header. Furthermore, $originator_name was used for submission mode
3588 messages from external hosts without From: headers in a similar way,
3589 which is clearly wrong.
3591 PH/37 Added control=suppress_local_fixups.
3593 PH/38 When log_selector = +received_sender was set, and the addition of the
3594 sender made the log line's construction buffer exactly full, or one byte
3595 less than full, an overflow happened when the terminating "\n" was
3598 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3599 when the result of a list match is failure because a DNS lookup failed.
3601 PH/40 RM_COMMAND is now used in the building process.
3603 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3604 the "build-* directories that it finds.
3606 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3607 address in a domain literal was a prefix of an interface address.
3609 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3610 when verifying a sender address, unless rewrite_headers is false.
3612 PH/44 Wrote a long comment about why errors_to addresses are verified as
3613 recipients, not senders.
3615 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3616 the ratelimit ACL was added.
3618 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3620 PH/46 Added extra information about PostgreSQL errors to the error string.
3622 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3623 avoiding the need to install special zones in a real server. This is
3624 backwards compatible; if it can't find the fake resolver, it drops back.
3625 Thus, both old and new test suites can be run.
3627 TF/02 Added util/ratelimit.pl
3629 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3630 clock is set back in time.
3632 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3633 Candler <B.Candler@pobox.com>.
3635 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3636 OK for addresses that are the result of redirections.
3638 PH/48 A number of further additions for the benefit of the new test suite,
3639 including a fake gethostbyname() that interfaces to the fake DNS resolver
3642 TF/06 The fix for widen_domains has also been applied to qualify_single and
3643 search_parents which are the other dnslookup options that can cause
3646 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3649 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3651 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3652 service" if it encountered an MX record with an empty target hostname.
3653 The message is now "an MX or SRV record indicated no SMTP service".
3655 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3656 if the DNS of the sending site is misconfigured. This is quite a
3657 common situation. This change restores the behaviour of treating a
3658 helo verification defer as a failure.
3660 PH/51 If self=fail was set on a router, the bounce message did not include the
3661 actual error message.
3667 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3669 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3670 command itself ended in error, the underlying message about the transport
3671 filter timeout was being overwritten with the pipe command error. Now the
3672 underlying error message should be appended to the second error message.
3674 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3676 PH/02 Reduce the amount of output that "make" produces by default. Full output
3677 can still be requested.
3679 PH/03 The warning log line about a condition test deferring for a "warn" verb
3680 was being output only once per connection, rather than after each
3681 occurrence (because it was using the same function as for successful
3682 "warn" verbs). This seems wrong, so I have changed it.
3684 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3685 it should not have, which might have caused a crash in the right
3686 circumstances, but probably never did.
3688 PH/04 Installed a modified version of Tony Finch's patch to make submission
3689 mode fix the return path as well as the Sender: header line, and to
3690 add a /name= option so that you can make the user's friendly name appear
3693 TF/03 Added the control = fakedefer ACL modifier.
3695 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3696 Mark Lowes for thorough testing.
3698 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3700 TK/03 Merged latest SRS patch from Miles Wilton.
3702 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3703 with the definition in sysexits.h (which is #included earlier).
3704 Fortunately, Exim does not actually use EX_OK. The code used to try to
3705 preserve the sysexits.h value, by assuming that macro definitions were
3706 scanned for macro replacements. I have been disabused of this notion,
3707 so now the code just undefines EX_OK before #including unistd.h.
3709 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3710 in the smtp transport. When a block could not be written in a single
3711 write() function, the timeout was being re-applied to each part-write.
3712 This seems wrong - if the receiver was accepting one byte at a time it
3713 would take for ever. The timeout is now adjusted when this happens. It
3714 doesn't have to be particularly precise.
3716 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3717 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3719 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3720 without a domain if the check to <postmaster@domain> fails.
3722 SC/01 Eximstats: added -xls and the ability to specify output files
3723 (patch written by Frank Heydlauf).
3725 SC/02 Eximstats: use FileHandles for outputting results.
3727 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3729 SC/04 Eximstats: fixed display of large numbers with -nvr option
3731 SC/05 Eximstats: fixed merging of reports with empty tables.
3733 SC/06 Eximstats: added the -include_original_destination flag
3735 SC/07 Eximstats: removed tabs and trailing whitespace.
3737 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3739 TK/06 MBOX spool code: Add real "From " MBOX separator line
3740 so the .eml file is really in mbox format (even though
3741 most programs do not really care). Patch from Alex Miller.
3743 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3744 The latter is generated from $received_to and is only set if the
3745 message has one envelope recipient. SA can use these headers,
3746 obviously out-of-the-box. Patch from Alex Miller.
3748 PH/08 The ${def test on a variable was returning false if the variable's
3749 value was "0", contrary to what the specification has always said!
3750 The result should be true unless the variable is empty.
3752 PH/09 The syntax error of a character other than { following "${if
3753 def:variable_name" (after optional whitespace) was not being diagnosed.
3754 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3755 accidental colon was present, for example, could give incorrect results.
3757 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3758 result is used (not including appendfile, where other changes are about
3761 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3762 This involved changing a lot of size variables from int to off_t. It
3763 should work with maildirs and everything.
3765 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3766 spamd dying while we are connected to it.
3768 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3771 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3772 function for BDB 4.3.
3774 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3776 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3777 the sources for PCRE have been re-arranged and more files are now
3780 PH/15 The code I had for printing potentially long long variables in PH/11
3781 above was not the best (it lost precision). The length of off_t variables
3782 is now inspected at build time, and an appropriate printing format (%ld
3783 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3784 to be "long long int" or "long int". This is needed for the internal
3785 formatting function string_vformat().
3787 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3788 the configuration file to be ":syslog", then the script "guesses" where
3789 the logs files are, rather than using the compiled in default. In our
3790 case the guess is not the same as the compiled default, so the script
3791 suddenly stopped working when I started to use syslog. The patch checks
3792 to see if log_file_path is "". If so, it attempts to read it from exim
3793 with no configuration file to get the compiled in version, before it
3794 falls back to the previous guessing code."
3796 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3797 implementing BATV in an Exim configuration. See NewStuff for the gory
3800 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3801 Makefile that are specific to HP-UX.
3803 PH/18 If the "use_postmaster" option was set for a recipient callout together
3804 with the "random" option, the postmaster address was used as the MAIL
3805 FROM address for the random test, but not for the subsequent recipient
3806 test. It is now used for both.
3808 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3809 patch removes a few documentation additions to RFC 3028, because the
3810 latest draft now contains them. It adds the new en;ascii-case comparator
3811 and a new error check for 8bit text in MIME parts. Comparator and
3812 require names are now matched exactly. I enabled the subaddress
3813 extension, but it is not well tested yet (read: it works for me)."
3815 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3816 rework some of the code of TK/09 above to avoid the hardwired use of
3817 "%lld" and "long long". Replaced the call to snprintf() with a call to
3820 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3821 records point to non-existent hosts", "retry timeout exceeded", and
3822 "retry time not reached for any host after a long failure period".
3824 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3825 experimental DomainKeys support:
3827 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3828 (2) On an error such as an illegally used "control", the wrong name for
3829 the control was given.
3831 These problems did NOT occur unless DomainKeys support was compiled.
3833 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3835 PH/24 Added ${if match_ip condition.
3837 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3838 databases so that it will be absolutely obvious if a crash occurs in the
3839 DB library. This is a regular occurrence (often caused by mis-matched
3842 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3843 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3844 picked up on a user's system that detects such things. There doesn't seem
3845 to be a gcc warning option for this - only an attribute that has to be
3846 put on the function's prototype. It seems that in Fedora Core 4 they have
3847 set this on a number of new functions. No doubt there will be more in due
3850 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3851 specify a transport. However, if an address that was verified by such a
3852 router was the subject of a callout, Exim crashed because it tried to
3853 read the rcpt_include_affixes from the non-existent transport. Now it
3854 just assumes that the setting of that option is false. This bug was
3855 introduced by 4.51/PH/31.
3857 PH/28 Changed -d+all to exclude +memory, because that information is very
3858 rarely of interest, but it makes the output a lot bigger. People tend to
3859 do -d+all out of habit.
3861 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3862 code in os-type was giving problems when libc.so lives in lib64, like on
3865 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3866 aren't the modern standard, and it seems that some systems' include files
3867 don't always have them. Exim was already checking for some of the newer
3868 ones like T_AAAA, and defining it itself. I've added checks for all the
3869 record types that Exim uses.
3871 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3872 not automatically generating a new one, as it is supposed to. This
3873 prevented TLS from working. If the file did exist, but contained invalid
3874 data, a new version was generated, as expected. It was only the case of a
3875 non-existent file that was broken.
3877 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3878 with a change in libdomainkeys > 0.64.
3880 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3881 from DNS. If the selector record carries the flag, it now has
3882 precedence over the domain-wide flag.
3884 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3886 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3887 the use of a port name that isn't defined in /etc/services) occurred, the
3888 message was deferred as in a normal delivery, and thus remained on the
3889 spool, instead of being failed because of the mua_wrapper setting. This
3890 is now fixed, and I tidied up some of the mua_wrapper messages at the
3893 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3894 the messages in a hash of arrays rather than using individual hashes.
3895 This is a bit cleaner and results in dramatic memory savings, albeit
3896 at a slight CPU cost.
3898 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3899 as requested by Marc Sherman.
3901 SC/10 Eximstats: added histograms for user specified patterns as requested
3904 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3906 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3907 fopen() in the content-scanning modules that did not already have it.
3913 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3914 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3916 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3918 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3920 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3921 to test Sieve filters that use "vacation".
3923 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3924 that changes the way the GnuTLS parameters are stored in the cache file.
3925 The new format can be generated externally. For backward compatibility,
3926 if the data in the cache doesn't make sense, Exim assumes it has read an
3927 old-format file, and it generates new data and writes a new file. This
3928 means that you can't go back to an older release without removing the
3931 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3932 work if there are any delivery delays because "one_time" forces the
3933 parent to be marked "delivered", so its unseen clone is never tried
3934 again. For this reason, Exim now forbids the simultaneous setting of
3937 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3938 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3939 entry if you want to know the details. The fix, however, neglected to
3940 consider the case when local delivery batching is involved. The test for
3941 "previously delivered" was not happening when checking to see if an
3942 address could be batched with a previous (undelivered) one; under
3943 certain circumstances this could lead to multiple deliveries to the same
3946 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3947 in its include files, and this causes problems building Exim.
3949 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3950 header_syntax) but Exim was just ignoring anything given after a slash.
3951 In particular, this caused confusion with an attempt to use "verify =
3952 reverse_host_lookup/defer_ok". An error is now given when options are
3953 supplied for verify items that do not have them. (Maybe reverse_host_
3954 lookup should have a defer_ok option, but that's a different point.)
3956 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3957 defined by RFC 821) to 2048, because there were problems with some AUTH
3958 commands, and RFC 1869 says the size should be increased for extended
3959 SMTP commands that take arguments.
3961 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3964 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3965 "unknown" error; now it says that the functionality isn't in the binary.
3967 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3968 an address' error message when a string expansion fails (syntax or
3969 whatever). Otherwise the password may appear in the log. Following change
3970 PH/42 below, there is no longer a chance of it appearing in a bounce
3973 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3975 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3976 of its addresses was ever tried. (Bugzilla bug #2.)
3978 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3979 the result incorrectly in the debug output. (It correctly added a newline
3980 to what was transported.)
3982 TF/01 Added $received_time.
3984 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3985 commented out examples of how to interface to a virus scanner and to
3986 SpamAssassin. Also added commented examples of av_scanner and
3987 spamd_address settings.
3989 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3990 and controls are allowed in which ACLs. There were a couple of minor
3991 errors. Some of the entries in the conditions table (which is a table of
3992 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3993 negation of where the condition IS allowed.
3995 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3997 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3998 header file does not have a version number, so I've had to invent a new
3999 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4000 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4001 radiusclient), but it was contributed by a Radius user.
4003 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4004 files or directories, for OpenSSL.
4006 PH/19 When an Exim process that is running as root has to create an Exim log
4007 file, it does so in a subprocess that runs as exim:exim so as to get the
4008 ownership right at creation (otherwise, other Exim processes might see
4009 the file with the wrong ownership). There was no test for failure of this
4010 fork() call, which would lead to the process getting stuck as it waited
4011 for a non-existent subprocess. Forks do occasionally fail when resources
4012 run out. I reviewed all the other calls to fork(); they all seem to check
4015 PH/20 When checking for unexpected SMTP input at connect time (before writing
4016 the banner), Exim was not dealing correctly with a non-positive return
4017 from the read() function. If the client had disconnected by this time,
4018 the result was a log entry for a synchronization error with an empty
4019 string after "input=" when read() returned zero. If read() returned -1
4020 (an event I could not check), uninitialized data bytes were printed.
4021 There were reports of junk text (parts of files, etc) appearing after
4024 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4026 PH/22 Added support for macro redefinition, and (re)definition in between
4027 driver and ACL definitions.
4029 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4030 forgetting to use the resulting value; it was using the unexpanded value.
4032 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4033 hadn't been configured. The fix is from Juergen Kreileder, who
4034 understands it better than I do:
4036 "Here's what I see happening with three configured cyrus_sasl
4037 authenticators configured (plain, login, cram-md5):
4039 On startup auth_cyrus_sasl_init() gets called for each of these.
4040 This means three calls to sasl_listmech() without a specified mech_list.
4041 => SASL tests which mechs of all available mechs actually work
4042 => three warnings about OTP not working
4043 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4045 With the patch, sasl_listmech() also gets called three times. But now
4046 SASL's mech_list option is set to the server_mech specified in the the
4047 authenticator. Or in other words, the answer from sasl_listmech()
4048 gets limited to just the mech you're testing for (which is different
4050 => the return list contains just 'plain' or 'login', 'cram-md5' or
4051 nothing depending on the value of ob->server_mech.
4053 I've just tested the patch: Authentication still works fine,
4054 unavailable mechs specified in the exim configuration are still
4055 caught, and the auth.log warnings about OTP are gone."
4057 PH/25 When debugging is enabled, the contents of the command line are added
4058 to the debugging output, even when log_selector=+arguments is not
4061 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4062 answer is "GNU", and only if the return is "GNU/something" is the answer
4065 PH/27 $acl_verify_message is now set immediately after the failure of a
4066 verification in an ACL, and so is available in subsequent modifiers. In
4067 particular, the message can be preserved by coding like this:
4069 warn !verify = sender
4070 set acl_m0 = $acl_verify_message
4072 Previously, $acl_verify_message was set only while expanding "message"
4073 and "log_message" when a very denied access.
4075 PH/28 Modified OS/os.c-Linux with
4077 -#ifndef OS_LOAD_AVERAGE
4078 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4080 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4081 nomenclature these days.)
4083 PH/29 Installed patch from the Sieve maintainer that adds the options
4084 sieve_useraddress and sieve_subaddress to the redirect router.
4086 PH/30 In these circumstances:
4087 . Two addresses routed to the same list of hosts;
4088 . First host does not offer TLS;
4089 . First host accepts first address;
4090 . First host gives temporary error to second address;
4091 . Second host offers TLS and a TLS session is established;
4092 . Second host accepts second address.
4093 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4094 and peerdn, if requested) that were in fact used only for the second
4097 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4098 attention to any local part prefix or suffix that was matched by the
4099 router that accepted the address. It now behaves in the same way as it
4100 does for delivery: the affixes are removed from the local part unless
4101 rcpt_include_affixes is set on the transport.
4103 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4104 timeout during the DATA phase of an incoming message.
4106 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4107 applied a patch sanctioned by the Sieve maintainer.
4109 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4110 the uid or gid is negative. A case of a negative gid caused this to be
4111 noticed. The fix allows for either to be negative.
4113 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4114 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4117 PH/36 The change PH/12 above was broken. Fixed it.
4119 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4120 the grounds that routing the same address twice would always produce the
4121 same answer. This might have been true once, but it is certainly no
4122 longer true now. Routing a child address may depend on the previous
4123 routing that produced that child. Some complicated redirection strategies
4124 went wrong when messages had multiple recipients, and made Exim's
4125 behaviour dependent on the order in which the addresses were given.
4127 I have moved the duplicate checking until after the routing is complete.
4128 Exim scans the addresses that are assigned to local and remote
4129 transports, and removes any duplicates. This means that more work will be
4130 done, as duplicates will always all be routed, but duplicates are
4131 presumably rare, so I don't expect this is of any significance.
4133 For deliveries to pipes, files, and autoreplies, the duplicate checking
4134 still happens during the routing process, since they are not going to be
4137 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4138 It corrects a timeout issue with spamd. This is Ian's comment: "The
4139 background is that sometimes spamd either never reads data from a
4140 connection it has accepted, or it never writes response data. The exiscan
4141 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4142 blindly assumes that writes won't block so it may never time out."
4144 PH/39 Allow G after quota size as well as K and M.
4146 PH/40 The value set for $authenticated_id in an authenticator may not contain
4147 binary zeroes or newlines because the value is written to log lines and
4148 to spool files. There was no check on this. Now the value is run through
4149 the string_printing() function so that such characters are converted to
4150 printable escape sequences.
4152 PH/41 $message_linecount is a new variable that contains the total number of
4153 lines in the message. Compare $body_linecount, which is the count for the
4156 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4157 bounce and delay warning messages, except in certain special cases, which
4160 (a) An SMTP error message from a remote host;
4161 (b) A message specified in a :fail: redirection;
4162 (c) A message specified in a "fail" command in a system filter;
4163 (d) A message specified in a FAIL return from the queryprogram router;
4164 (e) A message specified by the cannot_route_message router option.
4166 In these cases only, Exim does include the error details in bounce and
4167 warning messages. There are also a few cases where bland messages such
4168 as "unrouteable address" or "local delivery error" are given.
4170 PH/43 $value is now also set for the "else" part of a ${run expansion.
4172 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4173 being worked on, but at least Exim now implements the latest version to
4176 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4177 process to complete was treated as a delivery failure, a timeout while
4178 writing the message to the pipe was logged, but erroneously treated as a
4179 successful delivery. Such timeouts include transport filter timeouts. For
4180 consistency with the overall process timeout, these timeouts are now
4181 treated as errors, giving rise to delivery failures by default. However,
4182 there is now a new Boolean option for the pipe transport called
4183 timeout_defer, which, if set TRUE, converts the failures into defers for
4184 both kinds of timeout. A transport filter timeout is now identified in
4187 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4188 systems where "make" and "gmake" are different, calling "gmake" at top
4189 level broke things. I've arranged for the value of $(MAKE) to be passed
4190 from the Makefile to this script so that it can call the same version of
4194 A note about Exim versions 4.44 and 4.50
4195 ----------------------------------------
4197 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4198 changes of various kinds. As a consequence, a big documentation update was
4199 needed. This delayed the release for rather longer than seemed good, especially
4200 in the light of a couple of (minor) security issues. Therefore, the changes
4201 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4202 release. So 4.44 and 4.50 are in effect two different branches that both start
4205 I have left the 4.50 change log unchanged; it contains all the changes since
4206 4.43. The change log for 4.44 is below; many of its items are identical to
4207 those for 4.50. This seems to be the most sensible way to preserve the
4208 historical information.
4214 1. Minor wording change to the doc/README.SIEVE file.
4216 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4217 computation of the current number of files was incorrect.
4219 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4220 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4223 4. Give more explanation in the error message when the command for a transport
4224 filter fails to execute.
4226 5. There are several places where Exim runs a non-Exim command in a
4227 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4228 was being done only for the command run by the queryprogram router. It is
4229 now done for all such subprocesses. The other cases are: ${run, transport
4230 filters, and the commands run by the lmtp and pipe transports.
4232 6. Added CONFIGURE_GROUP build-time option.
4234 7. Some older OS have a limit of 256 on the maximum number of file
4235 descriptors. Exim was using setrlimit() to set 1000 as a large value
4236 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4237 systems. I've change it so that if it can't get 1000, it tries for 256.
4239 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4240 was an oversight, and furthermore, ever since the addition of extra
4241 controls (e.g. 4.43/32), the checks on when to allow different forms of
4242 "control" were broken. There should now be diagnostics for all cases when a
4243 control that does not make sense is encountered.
4245 9. Added the /retain_sender option to "control=submission".
4247 10. $recipients is now available in the predata ACL (oversight).
4249 11. Tidy the search cache before the fork to do a delivery from a message
4250 received from the command line. Otherwise the child will trigger a lookup
4251 failure and thereby defer the delivery if it tries to use (for example) a
4252 cached ldap connection that the parent has called unbind on.
4254 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4255 of $address_data from the recipient verification was clobbered by the
4256 sender verification.
4258 13. The value of address_data from a sender verification is now available in
4259 $sender_address_data in subsequent conditions in the ACL statement.
4261 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4263 15. Added a new option "connect=<time>" to callout options, to set a different
4266 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4267 was its contents. (It was OK if the option was not defined at all.)
4269 17. A "Completed" log line is now written for messages that are removed from
4270 the spool by the -Mrm option.
4272 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4273 information about exactly what failed.
4275 19. Added -dd to debug only the daemon process.
4277 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4278 handles timeouts, both on the server side and network timeouts. Renamed the
4279 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4281 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4282 It is now set to "smtps".
4284 22. $host_address is now set to the target address during the checking of
4285 ignore_target_hosts.
4287 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4288 being passed; this would have caused $sender_host_name to have been used if
4289 matching the list had actually called for a host name (not very likely,
4290 since this list is usually IP addresses). A host name is now passed as
4293 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4294 code that specifies a non-restarting handler (typically sigaction() in
4295 modern systems) in an attempt to fix a rare and obscure crash bug.
4297 25. Narrowed the window for a race in the daemon that could cause it to ignore
4298 SIGCHLD signals. This is not a major problem, because they are used only to
4299 wake it up if nothing else does.
4301 26. A malformed maildirsize file could cause Exim to calculate negative values
4302 for the mailbox size or file count. Odd effects could occur as a result.
4303 The maildirsize information is now recalculated if the size or filecount
4306 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4307 support for a long time. Removed HAVE_SYS_VFS_H.
4309 28. Installed the latest version of exipick from John Jetmore.
4311 29. In an address list, if the pattern was not a regular expression, an empty
4312 subject address (from a bounce message) matched only if the pattern was an
4313 empty string. Non-empty patterns were not even tested. This was the wrong
4314 because it is perfectly reasonable to use an empty address as part of a
4315 database query. An empty address is now tested by patterns that are
4316 lookups. However, all the other forms of pattern expect the subject to
4317 contain a local part and a domain, and therefore, for them, an empty
4318 address still always fails if the pattern is not itself empty.
4320 30. Exim went into a mad DNS loop when attempting to do a callout where the
4321 host was specified on an smtp transport, and looking it up yielded more
4322 than one IP address.
4324 31. Re-factored the code for checking spool and log partition space into a
4325 function that finds that data and another that does the check. The former
4326 is then used to implement four new variables: $spool_space, $log_space,
4327 $spool_inodes, and $log_inodes.
4329 32. The RFC2047 encoding function was originally intended for short strings
4330 such as real names; it was not keeping to the 75-character limit for
4331 encoded words that the RFC imposes. It now respects the limit, and
4332 generates multiple encoded words if necessary. To be on the safe side, I
4333 have increased the buffer size for the ${rfc2047: expansion operator from
4336 33. It is now permitted to omit both strings after an "if" condition; if the
4337 condition is true, the result is "true". As before, when the second string
4338 is omitted, a false condition yields an empty string. This makes it less
4339 cumbersome to write custom ACL and router conditions.
4341 34. Failure to deliver a bounce message always caused it to be frozen, even if
4342 there was an errors_to setting on the router. The errors_to setting is now
4345 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4346 canonical form (fully expanded) before being placed in
4347 $sender_host_address.
4349 36. The table in the code that translates DNS record types into text (T_A to
4350 "A" for instance) was missing entries for NS and CNAME. It is just possible
4351 that this could have caused confusion if both these types were looked up
4352 for the same domain, because the text type is used as part of Exim's
4353 per-process caching. But the chance of anyone hitting this buglet seems
4356 37. The dnsdb lookup has been extended in a number of ways.
4358 (1) There is a new type, "zns", which walks up the domain tree until it
4359 finds some nameserver records. It should be used with care.
4361 (2) There is a new type, "mxh", which is like "mx" except that it returns
4362 just the host names, not the priorities.
4364 (3) It is now possible to give a list of domains (or IP addresses) to be
4365 looked up. The behaviour when one of the lookups defers can be
4366 controlled by a keyword.
4368 (4) It is now possible to specify the separator character for use when
4369 multiple records are returned.
4371 38. The dnslists ACL condition has been extended: it is now possible to supply
4372 a list of IP addresses and/or domains to be looked up in a particular DNS
4375 39. Added log_selector=+queue_time_overall.
4377 40. When running the queue in the test harness, wait just a tad after forking a
4378 delivery process, to get repeatability of debugging output.
4380 41. Include certificate and key file names in error message when GnuTLS fails
4381 to set them up, because the GnuTLS error message doesn't include the name
4382 of the failing file when there is a problem reading it.
4384 42. Allow both -bf and -bF in the same test run.
4386 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4388 44. The "Exiscan patch" is now merged into the mainline Exim source.
4390 45. Sometimes the final signoff response after QUIT could fail to get
4391 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4392 < 0 before doing a fflush(). This bug looks as though it goes back to the
4393 introduction of TLS in release 3.20, but "sometimes" must have been rare
4394 because the tests only now provoked it.
4396 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4397 (this can affect the format of dates).
4399 47. exim_tidydb, when checking for the continued existence of a message for
4400 which it has found a message-specific retry record, was not finding
4401 messages that were in split spool directories. Consequently, it was
4402 deleting retry records that should have stayed in existence.
4404 48. Steve fixed some bugs in eximstats.
4406 49. The SPA authentication driver was not abandoning authentication and moving
4407 on to the next authenticator when an expansion was forced to fail,
4408 contradicting the general specification for all authenticators. Instead it
4409 was generating a temporary error. It now behaves as specified.
4411 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4412 (the order specifies the preference for clients). The order is now AES256,
4413 AES128, 3DES, ARCFOUR128.
4415 51. Small patch to Sieve code - explicitly set From: when generating an
4418 52. Exim crashed if a remote delivery caused a very long error message to be
4419 recorded - for instance if somebody sent an entire SpamAssassin report back
4420 as a large number of 550 error lines. This bug was coincidentally fixed by
4421 increasing the size of one of Exim's internal buffers (big_buffer) that
4422 happened as part of the Exiscan merge. However, to be on the safe side, I
4423 have made the code more robust (and fixed the comments that describe what
4426 53. Now that there can be additional text after "Completed" in log lines (if
4427 the queue_time_overall log selector is set), a one-byte patch to exigrep
4428 was needed to allow it to recognize "Completed" as not the last thing in
4431 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4432 patch that reportedly fixes this has been added. I am not expert enough to
4433 create a test for it. This is what the patch creator wrote:
4435 "I found a little strange behaviour of ldap code when working with
4436 Windows 2003 AD Domain, where users was placed in more than one
4437 Organization Units. When I tried to give exim partial DN, the exit code
4438 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4439 But simultaneously result of request was absolutely normal ldap result,
4440 so I produce this patch..."
4442 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4443 REFERENCE, so I have modified the code to exclude the patch when that macro
4446 55. Some experimental protocols are using DNS PTR records for new purposes. The
4447 keys for these records are domain names, not reversed IP addresses. The
4448 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4449 leaves it alone. Component reversal etc. now happens only for IP addresses.
4452 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4454 57. Double the size of the debug message buffer (to 2048) so that more of very
4455 long debug lines gets shown.
4457 58. The exicyclog utility now does better if the number of log files to keep
4458 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4460 59. Two changes related to the smtp_active_hostname option:
4462 (1) $smtp_active_hostname is now available as a variable.
4463 (2) The default for smtp_banner uses $smtp_active_hostname instead
4464 of $primary_hostname.
4466 60. The host_aton() function is supposed to be passed a string that is known
4467 to be a valid IP address. However, in the case of IPv6 addresses, it was
4468 not checking this. This is a hostage to fortune. Exim now panics and dies
4469 if the condition is not met. A case was found where this could be provoked
4470 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4471 components; fortuitously, this particular loophole had already been fixed
4472 by change 4.50/55 above.
4474 If there are any other similar loopholes, the new check in host_aton()
4475 itself should stop them being exploited. The report I received stated that
4476 data on the command line could provoke the exploit when Exim was running as
4477 exim, but did not say which command line option was involved. All I could
4478 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4479 running as the user.
4482 61. There was a buffer overflow vulnerability in the SPA authentication code
4483 (which came originally from the Samba project). I have added a test to the
4484 spa_base64_to_bits() function which I hope fixes it.
4487 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4488 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4490 63. The daemon start-up calls getloadavg() while still root for those OS that
4491 need the first call to be done as root, but it missed one case: when
4492 deliver_queue_load_max is set with deliver_drop_privilege. This is
4493 necessary for the benefit of the queue runner, because there is no re-exec
4494 when deliver_drop_privilege is set.
4496 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4497 This has been fixed.
4499 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4500 was in use, was not putting the data itself into the right store pool;
4501 consequently, it could be overwritten for a subsequent message in the same
4502 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4505 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4507 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4508 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4509 the function were treating the return as a boolean value, which happened to
4510 work because 0=false and not-0=true, but is not correct code.
4512 68. The host_aton() function was not handling scoped IPv6 addresses (those
4513 with, for example, "%eth0" on the end) correctly.
4515 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4516 negated items (that is, ~something) in unsigned ints. Some compilers
4517 apparently mutter when there is no cast.
4519 70. If an address verification called from an ACL failed, and did not produce a
4520 user-specific message (i.e. there was only a "system" message), nothing was
4521 put in $acl_verify_message. In this situation, it now puts the system
4524 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4525 session; change 4.31/43 added the unwanted input to the log line - except
4526 that it did not do this in the start of session case. It now does.
4528 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4529 This is wrong and can cause the other end to generate a synchronization
4530 error if it is another Exim or anything else that does the synchronization
4531 check. A QUIT command is no longer sent after a timeout.
4533 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4534 during host lookups.
4536 74. The defer_ok option of callout verification was not working if it was used
4537 when verifying addresses in header lines, that is, for this case:
4539 verify = header_sender/callout=defer_ok
4541 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4542 those file descriptors could be used for SMTP connections. If anything
4543 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4544 could be sent to the SMTP client, causing chaos. The daemon now opens
4545 stdin, stdout, and stderr to /dev/null when it puts itself into the
4548 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4549 log by default. The user can override this with suitable Perl magic.
4551 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4552 the log message when discard triggers, was not working for the DATA ACL or
4553 for the non-SMTP ACL.
4555 78. Error message wording change in sieve.c.
4557 79. If smtp_accept_max_per_host was set, the number of connections could be
4558 restricted to fewer than expected, because the daemon was trying to set up
4559 a new connection before checking whether the processes handling previous
4560 connections had finished. The check for completed processes is now done
4561 earlier. On busy systems, this bug wouldn't be noticed because something
4562 else would have woken the daemon, and it would have reaped the completed
4565 80. If a message was submitted locally by a user whose login name contained one
4566 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4567 It caused a spool format error. I have fixed the spool reading code. A
4568 related problem was that the "from" clause in the Received: line became
4569 illegal because of the space(s). It is now covered by ${quote_local_part.
4571 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4574 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4576 83. Patch from Sieve maintainer to fix unterminated string problem in
4577 "vacation" handling.
4579 84. Some minor changes to the Linux configuration files to help with other
4580 OS variants using glibc.
4582 85. One more patch for Sieve to update vacation handling to latest spec.
4585 ----------------------------------------------------
4586 See the note above about the 4.44 and 4.50 releases.
4587 ----------------------------------------------------
4593 1. Change 4.43/35 introduced a bug that caused file counts to be
4594 incorrectly computed when quota_filecount was set in an appendfile
4597 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4598 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4601 3. Give more explanation in the error message when the command for a transport
4602 filter fails to execute.
4604 4. There are several places where Exim runs a non-Exim command in a
4605 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4606 was being done only for the command run by the queryprogram router. It is
4607 now done for all such subprocesses. The other cases are: ${run, transport
4608 filters, and the commands run by the lmtp and pipe transports.
4610 5. Some older OS have a limit of 256 on the maximum number of file
4611 descriptors. Exim was using setrlimit() to set 1000 as a large value
4612 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4613 systems. I've change it so that if it can't get 1000, it tries for 256.
4615 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4616 was an oversight, and furthermore, ever since the addition of extra
4617 controls (e.g. 4.43/32), the checks on when to allow different forms of
4618 "control" were broken. There should now be diagnostics for all cases when a
4619 control that does not make sense is encountered.
4621 7. $recipients is now available in the predata ACL (oversight).
4623 8. Tidy the search cache before the fork to do a delivery from a message
4624 received from the command line. Otherwise the child will trigger a lookup
4625 failure and thereby defer the delivery if it tries to use (for example) a
4626 cached ldap connection that the parent has called unbind on.
4628 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4629 of $address_data from the recipient verification was clobbered by the
4630 sender verification.
4632 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4633 was its contents. (It was OK if the option was not defined at all.)
4635 11. A "Completed" log line is now written for messages that are removed from
4636 the spool by the -Mrm option.
4638 12. $host_address is now set to the target address during the checking of
4639 ignore_target_hosts.
4641 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4642 being passed; this would have caused $sender_host_name to have been used if
4643 matching the list had actually called for a host name (not very likely,
4644 since this list is usually IP addresses). A host name is now passed as
4647 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4648 code that specifies a non-restarting handler (typically sigaction() in
4649 modern systems) in an attempt to fix a rare and obscure crash bug.
4651 15. Narrowed the window for a race in the daemon that could cause it to ignore
4652 SIGCHLD signals. This is not a major problem, because they are used only to
4653 wake it up if nothing else does.
4655 16. A malformed maildirsize file could cause Exim to calculate negative values
4656 for the mailbox size or file count. Odd effects could occur as a result.
4657 The maildirsize information is now recalculated if the size or filecount
4660 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4661 support for a long time. Removed HAVE_SYS_VFS_H.
4663 18. Updated exipick to current release from John Jetmore.
4665 19. Allow an empty sender to be matched against a lookup in an address list.
4666 Previously the only cases considered were a regular expression, or an
4669 20. Exim went into a mad DNS lookup loop when doing a callout where the
4670 host was specified on the transport, if the DNS lookup yielded more than
4673 21. The RFC2047 encoding function was originally intended for short strings
4674 such as real names; it was not keeping to the 75-character limit for
4675 encoded words that the RFC imposes. It now respects the limit, and
4676 generates multiple encoded words if necessary. To be on the safe side, I
4677 have increased the buffer size for the ${rfc2047: expansion operator from
4680 22. Failure to deliver a bounce message always caused it to be frozen, even if
4681 there was an errors_to setting on the router. The errors_to setting is now
4684 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4685 canonical form (fully expanded) before being placed in
4686 $sender_host_address.
4688 24. Updated eximstats to version 1.33
4690 25. Include certificate and key file names in error message when GnuTLS fails
4691 to set them up, because the GnuTLS error message doesn't include the name
4692 of the failing file when there is a problem reading it.
4694 26. Expand error message when OpenSSL has problems setting up cert/key files.
4697 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4698 (this can affect the format of dates).
4700 28. exim_tidydb, when checking for the continued existence of a message for
4701 which it has found a message-specific retry record, was not finding
4702 messages that were in split spool directories. Consequently, it was
4703 deleting retry records that should have stayed in existence.
4705 29. eximstats updated to version 1.35
4706 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4707 1.35 - bugfix such that pie charts by volume are generated correctly
4709 30. The SPA authentication driver was not abandoning authentication and moving
4710 on to the next authenticator when an expansion was forced to fail,
4711 contradicting the general specification for all authenticators. Instead it
4712 was generating a temporary error. It now behaves as specified.
4714 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4715 (the order specifies the preference for clients). The order is now AES256,
4716 AES128, 3DES, ARCFOUR128.
4718 31. Small patch to Sieve code - explicitly set From: when generating an
4721 32. Exim crashed if a remote delivery caused a very long error message to be
4722 recorded - for instance if somebody sent an entire SpamAssassin report back
4723 as a large number of 550 error lines. This bug was coincidentally fixed by
4724 increasing the size of one of Exim's internal buffers (big_buffer) that
4725 happened as part of the Exiscan merge. However, to be on the safe side, I
4726 have made the code more robust (and fixed the comments that describe what
4729 33. Some experimental protocols are using DNS PTR records for new purposes. The
4730 keys for these records are domain names, not reversed IP addresses. The
4731 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4732 leaves it alone. Component reversal etc. now happens only for IP addresses.
4735 34. The host_aton() function is supposed to be passed a string that is known
4736 to be a valid IP address. However, in the case of IPv6 addresses, it was
4737 not checking this. This is a hostage to fortune. Exim now panics and dies
4738 if the condition is not met. A case was found where this could be provoked
4739 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4740 components; fortuitously, this particular loophole had already been fixed
4741 by change 4.50/55 or 4.44/33 above.
4743 If there are any other similar loopholes, the new check in host_aton()
4744 itself should stop them being exploited. The report I received stated that
4745 data on the command line could provoke the exploit when Exim was running as
4746 exim, but did not say which command line option was involved. All I could
4747 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4748 running as the user.
4751 35. There was a buffer overflow vulnerability in the SPA authentication code
4752 (which came originally from the Samba project). I have added a test to the
4753 spa_base64_to_bits() function which I hope fixes it.
4756 36. The daemon start-up calls getloadavg() while still root for those OS that
4757 need the first call to be done as root, but it missed one case: when
4758 deliver_queue_load_max is set with deliver_drop_privilege. This is
4759 necessary for the benefit of the queue runner, because there is no re-exec
4760 when deliver_drop_privilege is set.
4762 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4763 was in use, was not putting the data itself into the right store pool;
4764 consequently, it could be overwritten for a subsequent message in the same
4765 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4768 38. Sometimes the final signoff response after QUIT could fail to get
4769 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4770 < 0 before doing a fflush(). This bug looks as though it goes back to the
4771 introduction of TLS in release 3.20, but "sometimes" must have been rare
4772 because the tests only now provoked it.
4778 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4779 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4780 Now it returns an integer. A number of calls were still expecting a T/F
4781 return. Fortuitously, in all cases, the tests worked in OK situations,
4782 which is the norm. However, things would have gone wrong on any write
4783 failures on the smtp file descriptor. This function is used when sending
4784 messages over SMTP and also when doing verify callouts.
4786 2. When Exim is called to do synchronous delivery of a locally submitted
4787 message (the -odf or -odi options), it no longer closes stderr before doing
4790 3. Implemented the mua_wrapper option.
4792 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4794 5. Implemented the functions header_remove(), header_testname(),
4795 header_add_at_position(), and receive_remove_recipient(), and exported them
4798 6. If an ACL "warn" statement specified the addition of headers, Exim already
4799 inserted X-ACL-Warn: at the start if there was no header name. However, it
4800 was not making this test for the second and subsequent header lines if
4801 there were newlines in the string. This meant that an invalid header could
4802 be inserted if Exim was badly configured.
4804 7. Allow an ACL "warn" statement to add header lines at the start or after all
4805 the Received: headers, as well as at the end.
4807 8. Added the rcpt_4xx retry error code.
4809 9. Added postmaster_mailfrom=xxx to callout verification option.
4811 10. Added mailfrom=xxxx to the callout verification option, for verify=
4814 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4815 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4817 12. Inserted some casts to stop certain compilers warning when using pointer
4818 differences as field lengths or precisions in printf-type calls (mostly
4819 affecting debugging statements).
4821 13. Added optional readline() support for -be (dynamically loaded).
4823 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4824 same clock tick as a message's arrival, so that its received time was the
4825 same as the "first fail" time on the retry record, and that message
4826 remained on the queue past the ultimate address timeout, every queue runner
4827 would try a delivery (because it was past the ultimate address timeout) but
4828 after another failure, the ultimate address timeout, which should have then
4829 bounced the address, did not kick in. This was a "< instead of <=" error;
4830 in most cases the first failure would have been in the next clock tick
4831 after the received time, and all would be well.
4833 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4834 being recognized when the domain list was tested by the match_domain
4835 condition in an expansion string.
4837 16. Added the ${str2b64: operator.
4839 17. Exim was always calling setrlimit() to set a large limit for the number of
4840 processes, without checking whether the existing limit was already
4841 adequate. (It did check for the limit on file descriptors.) Furthermore,
4842 errors from getrlimit() and setrlimit() were being ignored. Now they are
4843 logged to the main and panic logs, but Exim does carry on, to try to do its
4844 job under whatever limits there are.
4846 18. Imported PCRE 5.0.
4848 19. Trivial typo in log message " temporarily refused connection" (the leading
4851 20. If the log selector return_path_on_delivery was set and an address was
4852 redirected to /dev/null, the delivery process crashed because it assumed
4853 that a return path would always be set for a "successful" delivery. In this
4854 case, the whole delivery is bypassed as an optimization, and therefore no
4857 21. Internal re-arrangement: the function for sending a challenge and reading
4858 a response while authentication was assuming a zero-terminated challenge
4859 string. It's now changed to take a pointer and a length, to allow for
4860 binary data in such strings.
4862 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4864 23. Exim was not respecting finduser_retries when seeking the login of the
4865 uid under which it was called; it was always trying 10 times. (The default
4866 setting of finduser_retries is zero.) Also, it was sleeping after the final
4867 failure, which is pointless.
4869 24. Implemented tls_on_connect_ports.
4871 25. Implemented acl_smtp_predata.
4873 26. If the domain in control=submission is set empty, Exim assumes that the
4874 authenticated id is a complete email address when it generates From: or
4875 Sender: header lines.
4877 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4878 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4879 chown and chgrp in /bin and hostname in /usr/bin.
4881 28. Exim was keeping the "process log" file open after each use, just as it
4882 does for the main log. This opens the possibility of it remaining open for
4883 long periods when the USR1 signal hits a daemon. Occasional processlog
4884 errors were reported, that could have been caused by this. Anyway, it seems
4885 much more sensible not to leave this file open at all, so that is what now
4888 29. The long-running daemon process does not normally write to the log once it
4889 has entered its main loop, and it closes the log before doing so. This is
4890 so that log files can straightforwardly be renamed and moved. However,
4891 there are a couple of unusual error situations where the daemon does write
4892 log entries, and I had neglected to close the log afterwards.
4894 30. The text of an SMTP error response that was received during a remote
4895 delivery was being truncated at 512 bytes. This is too short for some of
4896 the long messages that one sometimes sees. I've increased the limit to
4899 31. It is now possible to make retry rules that apply only when a message has a
4900 specific sender, in particular, an empty sender.
4902 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4903 it possible to be selective about when SMTP synchronization is enforced.
4905 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4907 32. Implemented hosts_connection_nolog.
4909 33. Added an ACL for QUIT.
4911 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4914 35. Added mailbox_size and mailbox_filecount to appendfile.
4916 36. Added control = no_multiline_responses to ACLs.
4918 37. There was a bug in the logic of the code that waits for the clock to tick
4919 in the case where the clock went backwards by a substantial amount such
4920 that the microsecond fraction of "now" was more than the microsecond
4921 fraction of "then" (but the whole seconds number was less).
4923 38. Added support for the libradius Radius client library this is found on
4924 FreeBSD (previously only the radiusclient library was supported).
4930 1. When certain lookups returned multiple values in the form name=value, the
4931 quoting of the values was not always being done properly. Specifically:
4932 (a) If the value started with a double quote, but contained no whitespace,
4934 (b) If the value contained whitespace other than a space character (i.e.
4935 tabs or newlines or carriage returns) it was not quoted.
4936 This fix has been applied to the mysql and pgsql lookups by writing a
4937 separate quoting function and calling it from the lookup code. The fix
4938 should probably also be applied to nisplus, ibase and oracle lookups, but
4939 since I cannot test any of those, I have not disturbed their existing code.
4941 2. A hit in the callout cache for a specific address caused a log line with no
4942 reason for rejecting RCPT. Now it says "Previous (cached) callout
4943 verification failure".
4945 3. There was an off-by-one bug in the queryprogram router. An over-long
4946 return line was truncated at 256 instead of 255 characters, thereby
4947 overflowing its buffer with the terminating zero. As well as fixing this, I
4948 have increased the buffer size to 1024 (and made a note to document this).
4950 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4951 when Exim is waiting for an SMTP response from a remote server, Exim
4952 restarts its select() call on the socket, thereby resetting its timeout.
4953 This is not a problem when such interrupts are rare. Somebody set up a cron
4954 job to run exiwhat every 2 minutes, which is less than the normal select()
4955 timeout (5 or 10 minutes). This meant that the select() timeout never
4956 kicked in because it was always reset. I have fixed this by comparing the
4957 time when an interrupt arrives with the time at the start of the first call
4958 to select(). If more time than the timeout has elapsed, the interrupt is
4959 treated as a timeout.
4961 5. Some internal re-factoring in preparation for the addition of Sieve
4962 extensions (by MH). In particular, the "personal" test is moved to a
4963 separate function, and given an option for scanning Cc: and Bcc: (which is
4964 not set for Exim filters).
4966 6. When Exim created an email address using the login of the caller as the
4967 local part (e.g. when creating a From: or Sender: header line), it was not
4968 quoting the local part when it contained special characters such as @.
4970 7. Installed new OpenBSD configuration files.
4972 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4973 try to make them clearer.
4975 9. Callout options, other than the timeout value, were being ignored when
4976 verifying sender addresses in header lines. For example, when using
4978 verify = header_sender/callout=no_cache
4980 the cache was (incorrectly) being used.
4982 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4983 only the Cygwin environment.
4985 11. When return_path_on_delivery was set as a log selector, if different remote
4986 addresses in the same message used different return paths and parallel
4987 remote delivery occurred, the wrong values would sometimes be logged.
4988 (Whenever a remote delivery process finished, the return path value from
4989 the most recently started remote delivery process was logged.)
4991 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4992 lines when AUTH and/or TLS are in use. This is the "received protocol"
4993 field. Exim used to use "asmtp" for authenticated SMTP, without any
4994 indication (in the protocol name) for TLS use. Now it follows the RFC and
4995 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4996 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4997 names appear in log lines as well as in Received: header lines.
4999 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5000 extensions, and comparison tests, and to fix some bugs.
5002 14. Changes to the "personal" filter test:
5004 (1) The test was buggy in that it was just doing the equivalent of
5005 "contains" tests on header lines. For example, if a user's address was
5006 anne@some.where, the "personal" test would incorrectly be true for
5008 To: susanne@some.where
5010 This test is now done by extracting each address from the header in turn,
5011 and checking the entire address. Other tests that are part of "personal"
5012 are now done using regular expressions (for example, to check local parts
5013 of addresses in From: header lines).
5015 (2) The list of non-personal local parts in From: addresses has been
5016 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5017 taken from the Sieve specification recommendations.
5019 (3) If the message contains any header line starting with "List-" it is
5020 treated as non-personal.
5022 (4) The test for "circular" in the Subject: header line has been removed
5023 because it now seems ill-conceived.
5025 15. Minor typos in src/EDITME comments corrected.
5027 16. Installed latest exipick from John Jetmore.
5029 17. If headers_add on a router specified a text string that was too long for
5030 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5031 of string_sprintf() is now avoided.
5033 18. $message_body_size was not set (it was always zero) when running the DATA
5034 ACL and the local_scan() function.
5036 19. For the "mail" command in an Exim filter, no default was being set for
5037 the once_repeat time, causing a random time value to be used if "once" was
5038 specified. (If the value happened to be <= 0, no repeat happened.) The
5039 default is now 0s, meaning "never repeat". The "vacation" command was OK
5040 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5041 (I found it when inspecting the code).
5043 20. There is now an overall timeout for performing a callout verification. It
5044 defaults to 4 times the callout timeout, which applies to individual SMTP
5045 commands during the callout. The overall timeout applies when there is more
5046 than one host that can be tried. The timeout is checked before trying the
5047 next host. This prevents very long delays if there are a large number of
5048 hosts and all are timing out (e.g. when the network connections are timing
5049 out). The value of the overall timeout can be changed by specifying an
5050 additional sub-option for "callout", called "maxwait". For example:
5052 verify = sender/callout=5s,maxwait=20s
5054 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5055 to the end before writing, but this should make it even safer).
5057 22. Exim was forgetting that it had advertised PIPELINING for the second and
5058 subsequent messages on an SMTP connection. It was also not resetting its
5059 memory on STARTTLS and an internal HELO.
5061 23. When Exim logs an SMTP synchronization error within a session, it now
5062 records whether PIPELINING has been advertised or not.
5064 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5065 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5066 rather than long int.
5068 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5070 26. Added the never_mail option to autoreply.
5076 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5077 crash if the getsockname() call failed; this can happen if a connection is
5078 closed very soon after it is established. The problem was simply in the
5079 order in which certain operations were done, causing Exim to try to write
5080 to the SMTP stream before it had set up the file descriptor. The bug has
5081 been fixed by making things happen in the correct order.
5087 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5088 before the connection was closed, thus losing the rejection response.
5090 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5091 some early Solaris releases, but causes trouble in current releases where
5092 socklen_t is defined.
5094 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5097 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5100 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5101 that releasing the top of it at the end releases what was used for sub-
5102 expansions (unless the block got too big). However, discard this block if
5103 the first thing is a variable or header, so that we can use its block when
5104 it is dynamic (useful for very large $message_headers, for example).
5106 6. Lookups now cache *every* query, not just the most recent. A new, separate
5107 store pool is used for this. It can be recovered when all lookup caches are
5108 flushed. Lookups now release memory at the end of their result strings.
5109 This has involved some general refactoring of the lookup sources.
5111 7. Some code has been added to the store_xxx() functions to reduce the amount
5112 of flapping under certain conditions.
5114 8. log_incoming_interface used to affect only the <= reception log lines. Now
5115 it causes the local interface and port to be added to several more SMTP log
5116 lines, for example "SMTP connection from", and rejection lines.
5118 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5120 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5122 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5123 could be overwritten at the end of the current message (or the start of a
5124 new message if it was set in a HELO ACL). The value is now preserved for
5125 the duration of the SMTP connection.
5127 12. If a transport had a headers_rewrite setting, and a matching header line
5128 contained an unqualified address, that address was qualified, even if it
5129 did not match any rewriting rules. The underlying bug was that the values
5130 of the flags that permit the existence of unqualified sender and recipient
5131 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5132 non-local messages, and by -bnq for local messages) were not being
5133 preserved with the message after it was received.
5135 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5136 "next input=" as part of the text comprising the host identity instead of
5137 the correct text. The code was using the same buffer for two different
5138 strings. However, depending on which order the printing function evaluated
5139 its arguments, the bug did not always show up. Under Linux, for example, my
5140 test suite worked just fine.
5142 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5143 doesn't work with some older versions of Perl. It has been changed to "my",
5144 which in any case is probably the better facility to use.
5146 15. A really picky compiler found some instances of statements for creating
5147 error messages that either had too many or two few arguments for the format
5150 16. The size of the buffer for calls to the DNS resolver has been increased
5151 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5152 for addresses that have a lot of PTR records. This alleviates a problem; it
5153 does not fully solve it.
5155 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5156 buffer now truncates the list and logs the incident, which is the same
5157 action as happens when Exim is looking up a host name and its aliases.
5158 Previously in this situation something unpredictable would happen;
5159 sometimes it was "internal error: store_reset failed".
5161 18. If a server dropped the connection unexpectedly when an Exim client was
5162 using GnuTLS and trying to read a response, the client delivery process
5163 crashed while trying to generate an error log message.
5165 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5166 string, for example:
5168 warn message = H1: something\nH2: something
5170 the text was added as a single header line from Exim's point of view
5171 though it ended up OK in the delivered message. However, searching for the
5172 second and subsequent header lines using $h_h2: did not work. This has been
5173 fixed. Similarly, if a system filter added multiple headers in this way,
5174 the routers could not see them.
5176 20. Expanded the error message when iplsearch is called with an invalid key to
5177 suggest using net-iplsearch in a host list.
5179 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5180 ACLs are no longer actually imposed (and a message to that effect is
5183 22. If a "gecos" field in a passwd entry contained escaped characters, in
5184 particular, if it contained a \" sequence, Exim got it wrong when building
5185 a From: or a Sender: header from that name. A second bug also caused
5186 incorrect handling when an unquoted " was present following a character
5187 that needed quoting.
5189 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5190 was not being matched caselessly.
5192 24. Arranged for all hyphens in the exim.8 source to be escaped with
5195 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5196 properly. Recipient callout cache records were still being keyed to include
5197 the sender, even when use_sender was set false. This led to far more
5198 callouts that were necessary. The sender is no longer included in the key
5199 when use_sender is false.
5201 26. Added "control = submission" modifier to ACLs.
5203 27. Added the ${base62d: operator to decode base 62 numbers.
5205 28. dnsdb lookups can now access SRV records.
5207 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5208 the configuration file.
5210 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5211 (-v) mode. This makes the output for a verbose queue run more intelligible.
5213 31. Added a use_postmaster feature to recipient callouts.
5215 32. Added the $body_zerocount variable, containing the number of binary zero
5216 bytes in the message body.
5218 33. The time of last modification of the "new" subdirectory is now used as the
5219 "mailbox time last read" when there is a quota error for a maildir
5222 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5224 35. Added +ignore_unknown as a special item in host lists.
5226 36. Code for decoding IPv6 addresses in host lists is now included, even if
5227 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5228 address was recognized as an IP address, but was then not correctly decoded
5229 into binary, causing unexpected and incorrect effects when compared with
5236 1. Very minor rewording of debugging text in manualroute to say "list of
5237 hosts" instead of "hostlist".
5239 2. If verify=header_syntax was set, and a header line with an unqualified
5240 address (no domain) and a large number of spaces between the end of the
5241 name and the colon was received, the reception process suffered a buffer
5242 overflow, and (when I tested it) crashed. This was caused by some obsolete
5243 code that should have been removed. The fix is to remove it!
5245 3. When running in the test harness, delay a bit after writing a bounce
5246 message to get a bit more predictability in the log output.
5248 4. Added a call to search_tidyup() just before forking a reception process. In
5249 theory, someone could use a lookup in the expansion of smtp_accept_max_
5250 per_host which, without the tidyup, could leave open a database connection.
5252 5. Added the variables $recipient_data and $sender_data which get set from a
5253 lookup success in an ACL "recipients" or "senders" condition, or a router
5254 "senders" option, similar to $domain_data and $local_part_data.
5256 6. Moved the writing of debug_print from before to after the "senders" test
5259 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5260 problems for message scanning, either using a data ACL, or using
5261 local_scan() because the Received: header was not generated till after they
5262 were called (in order to set the time as the time of reception completion).
5263 I have revised the way this works. The header is now generated after the
5264 body is received, but before the ACL or local_scan() are called. After they
5265 are run, the timestamp in the header is updated.
5271 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5272 before starting a queue runner without re-exec. This happened only when
5273 deliver_drop_privilege was set or when the Exim user was set to root. The
5274 effect of the bug was that timeouts during subsequent deliveries caused
5275 crashes instead of being properly handled. The handler is now left at its
5276 default (and expected) setting.
5278 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5279 message, again when deliver_drop_privilege is set or Exim is run as root.
5280 The bug described in (1) was not present in this case, but the tidying up
5281 of the other signals was missing. I have made the two cases consistent.
5283 3. The ignore_target_hosts setting on a manualroute router was being ignored
5284 for hosts that were looked up using the /MX notation.
5286 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5289 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5290 operated on the sender address. After changing the $sender_address to <>
5291 for the sender address verify, Exim was re-instated it as the original
5292 (before rewriting) address, but remembering that it had rewritten it, so it
5293 wasn't rewriting it again. This bug also had the effect of breaking the
5294 sender address verification caching when the sender address was rewritten.
5296 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5297 This has been changed so that if the ip literal address matches
5298 ignore_target_hosts, the router declines.
5300 7. Added expansion conditions match_domain, match_address, and match_local_
5301 part (NOT match_host).
5303 8. The placeholder for the Received: header didn't have a length field set.
5305 9. Added code to Exim itself and to exim_lock to test for a specific race
5306 condition that could lead to file corruption when using MBX delivery. The
5307 issue is with the lockfile that is created in /tmp. If this file is removed
5308 after a process has opened it but before that process has acquired a lock,
5309 there is the potential for a second process to recreate the file and also
5310 acquire a lock. This could lead to two Exim processes writing to the file
5311 at the same time. The added code performs the same test as UW imapd; it
5312 checks after acquiring the lock that its file descriptor still refers to
5313 the same named file.
5315 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5316 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5319 11. Added the smtp_active_hostname option. If used, this will typically be made
5320 to depend on the incoming interface address. Because $interface_address is
5321 not set up until the daemon has forked a reception process, error responses
5322 that can happen earlier (such as "too many connections") no longer contain
5325 12. If an expansion in a condition on a "warn" statement fails because a lookup
5326 defers, the "warn" statement is abandoned, and the next ACL statement is
5327 processed. Previously this caused the whole ACL to be aborted.
5329 13. Added the iplsearch lookup type.
5331 14. Added ident_timeout as a log selector.
5333 15. Added tls_certificate_verified as a log selector.
5335 16. Added a global option tls_require_ciphers (compare the smtp transport
5336 option of the same name). This controls incoming TLS connections.
5338 17. I finally figured out how to make tls_require_ciphers do a similar thing
5339 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5340 before starting the TLS session.
5342 18. Tabs are now shown as \t in -bP output.
5344 19. If the log selector return_path_on_delivery was set, Exim crashed when
5345 bouncing a message because it had too many Received: header lines.
5347 20. If two routers both had headers_remove settings, and the first one included
5348 a superfluous trailing colon, the final name in the first list and the
5349 first name in the second list were incorrectly joined into one item (with a
5350 colon in the middle).
5356 1. Added -C and -D options to the exinext utility, mainly to make it easier
5357 to include in the automated testing, but these could be helpful when
5358 multiple configurations are in use.
5360 2. The exinext utility was not formatting the output nicely when there was
5361 an alternate port involved in the retry record key, nor when there was a
5362 message id as well (for retries that were specific to a specific message
5363 and a specific host). It was also confused by IPv6 addresses, because of
5364 the additional colons they contain. I have fixed the IPv4 problem, and
5365 patched it up to do a reasonable job for IPv6.
5367 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5368 delivery, the log line now contains "pipelined" if PIPELINING was used.
5370 4. An SMTP transport process used to panic and die if the bind() call to set
5371 an explicit outgoing interface failed. This has been changed; it is now
5372 treated in the same way as a connect() failure.
5374 5. A reference to $sender_host_name in the part of a conditional expansion
5375 that was being skipped was still causing a DNS lookup. This no longer
5378 6. The def: expansion condition was not recognizing references to header lines
5379 that used bh_ and bheader_.
5381 7. Added the _cache feature to named lists.
5383 8. The code for checking quota_filecount in the appendfile transport was
5384 allowing one more file than it should have been.
5386 9. For compatibility with Sendmail, the command line option
5394 and sets the incoming protocol and host name (for trusted callers). The
5395 host name and its colon can be omitted when only the protocol is to be set.
5396 Note the Exim already has two private options, -pd and -ps, that refer to
5397 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5398 "s", but I don't think that's a major issue.
5400 10. A number of refactoring changes to the code, none of which should affect
5403 (a) The number of logging options was getting close to filling up the
5404 32-bit word that was used as a bit map. I have split them into two classes:
5405 those that are passed in the argument to log_write(), and those that are
5406 only ever tested independently outside of that function. These are now in
5407 separate 32-bit words, so there is plenty of room for expansion again.
5408 There is no change in the user interface or the logging behaviour.
5410 (b) When building, for example, log lines, the code previously used a
5411 macro that called string_cat() twice, in order to add two strings. This is
5412 not really sufficiently general. Furthermore, there was one instance where
5413 it was actually wrong because one of the argument was used twice, and in
5414 one call a function was used. (As it happened, calling the function twice
5415 did not affect the overall behaviour.) The macro has been replaced by a
5416 function that can join an arbitrary number of extra strings onto a growing
5419 (c) The code for expansion conditions now uses a table and a binary chop
5420 instead of a serial search (which was left over from when there were very
5421 few conditions). Also, it now recognizes conditions like "pam" even when
5422 the relevant support is not compiled in: a suitably worded error message is
5423 given if an attempt is made to use such a condition.
5425 11. Added ${time_interval:xxxxx}.
5427 12. A bug was causing one of the ddress fields not to be passed back correctly
5428 from remote delivery subprocesses. The field in question was not being
5429 subsequently used, so this caused to problems in practice.
5431 13. Added new log selectors queue_time and deliver_time.
5433 14. Might have fixed a bug in maildirsizefile handling that threw up
5434 "unexpected character" debug warnings, and recalculated the data
5435 unnecessarily. In any case, I expanded the warning message to give more
5438 15. Added the message "Restricted characters in address" to the statements in
5439 the default ACL that block characters like @ and % in local parts.
5441 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5442 Three changes have been made:
5444 (a) There was a serious bug; a negative response to MAIL caused the whole
5445 recipient domain to be cached as invalid, thereby blocking all messages
5446 to all local parts at the same domain, from all senders. This bug has
5447 been fixed. The domain is no longer cached after a negative response to
5448 MAIL if the sender used is not empty.
5450 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5453 (c) A new callout option, "use_sender" has been added for people who want
5454 the modified behaviour.
5460 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5463 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5464 indeed breaks things for older releases.
5466 3. Added additional logging to the case where there is a problem reading data
5467 from a filter that is running in a subprocess using a pipe, in order to
5468 try to track down a specific problem.
5470 4. Testing facility fudge: when running in the test harness and attempting
5471 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5472 getting "No route to host". Convert this to a timeout.
5474 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5477 6. Some OS don't have socklen_t but use size_t instead. This affects the
5478 fifth argument of getsockopt() amongst other things. This is now
5479 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5480 can be set for individual OS. I have set it for SunOS5, OSF1, and
5481 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5482 some earlier ones do not.
5484 7. Change 4.30/15 was not doing the test caselessly.
5486 8. The standard form for an IPv6 address literal was being rejected by address
5487 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5488 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5489 this, as well as the form without the "IPv6" on the front (but only when
5490 address literals are enabled, of course).
5492 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5494 10. Exim crashed if a message with an empty sender address specified by -f
5495 encountered a router with an errors_to setting. This could be provoked only
5496 by a command such as
5500 where an empty string was supplied; "<>" did not hit this bug.
5502 11. Installed PCRE release 4.5.
5504 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5505 remained set. It is now erased.
5507 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5508 times from message ids (which are base 36 rather than the normal 62).
5510 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5511 were being counted as actual protocol errors, and logged if the log
5512 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5513 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5514 and DATA following a set of rejected RCPTs do not count as protocol errors.
5515 In other words, Exim assumes they were pipelined, though this may not
5516 actually be the case. Of course, in all cases the client gets an
5517 appropriate error code.
5519 15. If a lookup fails in an ACL condition, a message about the failure may
5520 be available; it is used if testing the ACL cannot continue, because most
5521 such messages specify what the cause of the deferral is. However, some
5522 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5523 that caused an old message to be retained and used if a later statement
5524 caused a defer, replacing the real cause of the deferral.
5526 16. If an IP address had so many PTR records that the DNS lookup buffer
5527 was not large enough to hold them, Exim could crash while trying to process
5528 the truncated data. It now detects and logs this case.
5530 17. Further to 4.21/58, another change has been made: if (and only if) the
5531 first line of a message (the first header line) ends with CRLF, a bare LF
5532 in a subsequent header line has a space inserted after it, so as not to
5533 terminate the header.
5535 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5536 unnecessarily, used atoi() instead of strtol(), and didn't check the
5537 termination when getting file sizes from file names by regex.
5539 19. Completely re-implemented the support for maildirsize files, in the light
5540 of a number of problems with the previous contributed implementation
5541 (4.30/29). In particular:
5543 . If the quota is zero, the maildirsize file is maintained, but no quota is
5546 . If the maildir directory does not exist, it is created before any attempt
5547 to write a maildirsize file.
5549 . The quota value in the file is just a cache; if the quota is changed in
5550 the transport, the new value overrides.
5552 . A regular expression is available for excluding directories from the
5555 20. The autoreply transport checks the characters in options that define the
5556 message's headers; it allows continued headers, but it was checking with
5557 isspace() after an embedded newline instead of explicitly looking for a
5560 21. If all the "regular" hosts to which an address was routed had passed their
5561 expiry times, and had not reached their retry times, the address was
5562 bounced, even if fallback hosts were defined. Now Exim should go on to try
5565 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5566 equivalent code in the SMTP transport. Some hosts send humungous responses
5567 to HELO/EHLO, more than 1024 it seems.
5569 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5570 gives compiler warnings in some environments. I've now done it "properly",
5573 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5574 (because of problems with the built-in one) was declared to return uschar *
5575 instead of char *, causing compiler failure.
5577 25. Fixed a file descriptor leak when processing alias/forward files.
5579 26. Fixed a minor format string issue in dbfn.c.
5581 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5583 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5584 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5587 29. When a sender address is verified, it is cached, to save repeating the test
5588 when there is more than one recipient in a message. However, when the
5589 verification involves a callout, it is possible for different callout
5590 options to be set for different recipients. It is too complicated to keep
5591 track of this in the cache, so now Exim always runs a verification when a
5592 callout is required, relying on the callout cache for the optimization.
5593 The overhead is duplication of the address routing, but this should not be
5596 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5597 to be verified with callout=postmaster, and the main callout worked but the
5598 postmaster check failed, the verification correctly failed. However, if a
5599 subsequent RCPT command asked for sender verification *without* the
5600 postmaster check, incorrect caching caused this verification also to fail,
5603 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5604 it was not caching the DNS options (qualify_single, search_parents) that
5605 were used when the lookup failed. A subsequent lookup with different
5606 options therefore always gave the same answer, though there were cases
5607 where it should not have. (Example: a "domains = !$mx_any" option on a
5608 dnslookup router: the "domains" option is always processed without any
5609 widening, but the router might have qualify_single set.) Now Exim uses the
5610 cached value only when the same options are set.
5612 32. Added John Jetmore's "exipick" utility to the distribution.
5614 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5615 than a timeout (e.g. a certificate is required, and is not provided), an
5616 Exim server now closes the connection immediately. Previously it waited for
5617 the client to close - but if the client is SSL, it seems that they each
5618 wait for each other, leading to a delay before one of them times out.
5620 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5621 maintained 0.8.x compatibility because I don't think many are using it, and
5622 it is clearly obsolete.
5624 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5627 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5628 client certificate was expired. A simple patch fixes this, though I don't
5629 understand the full logic of why the verify callback is called multiple
5632 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5633 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5634 which causes problems with some clients (such as the Certicom SSL Plus
5635 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5636 disables the coutermeasure allowing Eudora to connect."
5638 38. Exim was not checking that a write() to a log file succeeded. This could
5639 lead to Bad Things if a log got too big, in particular if it hit a file
5640 size limit. Exim now panics and dies if it cannot write to a log file, just
5641 as it does if it cannot open a log file.
5643 39. Modified OS/Makefile-Linux so that it now contains
5645 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5647 The two -D definitions ensure that Exim is compiled with large file
5648 support, which makes it possible to handle log files that are bigger than
5651 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5652 instance) a domain was checked against a named list that involved a lookup,
5653 causing $domain_data to be set, then another domain was checked against the
5654 same list, then the first domain was re-checked, the value of $domain_data
5655 after the final check could be wrong. In particular, if the second check
5656 failed, it could be set empty. This bug probably also applied to
5659 41. The strip_trailing_dot option was not being applied to the address given
5660 with the -f command-line option.
5662 42. The code for reading a message's header from the spool was incrementing
5663 $received_count, but never initializing it. This meant that the value was
5664 incorrect (doubled) while delivering a message in the same process in which
5665 it was received. In the most common configuration of Exim, this never
5666 happens - a fresh exec is done - but it can happen when
5667 deliver_drop_privilege is set.
5669 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5670 it now includes up to 150 characters of the unexpected data in the log
5673 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5674 and building data strings. The size of both of these buffers was 10 000
5675 bytes - far larger than anybody would *ever* want, thought I. Needless to
5676 say, somebody hit the limit. I have increased the maximum line length to
5677 20 000 and the maximum data length of concatenated lines to 100 000. I have
5678 also fixed two bugs, because there was no checking on these buffers. Tsk,
5679 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5680 buffer is too small.
5682 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5683 lsearch lookups. Now it does.
5685 46. When parsing a route_list item in a manualroute router, a fixed-length
5686 buffer was used for the list of hosts. I made this 1024 bytes long,
5687 thinking that nobody would ever have a list of hosts that long. Wrong.
5688 Somebody had a whole pile of complicated expansion conditions, and the
5689 string was silently truncated, leading to an expansion error. It turns out
5690 that it is easier to change to an unlimited length (owing to other changes
5691 that have happened since this code was originally written) than to build
5692 structure for giving a limitation error. The length of the item that
5693 expands into the list of hosts is now unlimited.
5695 47. The lsearch lookup could not handle data where the length of text line was
5696 more than 4095 characters. Such lines were truncated, leading to shortened
5697 data being returned. It should now handle lines of any length.
5699 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5700 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5703 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5704 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5705 no longer happens in comment lines. A list of replacements is now placed
5706 at the head of all of the source files, except those whose only change is
5707 to replace PERL_COMMAND in the very first #! line.
5709 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5710 messages on the queue, with a bottom-up merge sort, using code contributed
5711 by Michael Haardt. This should make operations like -bp somewhat faster on
5712 large queues. It won't affect queue runners, except when queue_run_in_order
5715 51. Installed eximstats 1.31 in the distribution.
5717 52. Added support for SRV lookups to the dnslookup router.
5719 53. If an ACL referred to $message_body or $message_body_end, the value was not
5720 reset for any messages that followed in the same SMTP session.
5722 54. The store-handling optimization for building very long strings was not
5723 differentiating between the different store pools. I don't think this
5724 actually made any difference in practice, but I've tidied it.
5726 55. While running the routers to verify a sender address, $sender_address
5727 was still set to the sender address. This is wrong, because when routing to
5728 send a bounce to the sender, it would be empty. Therefore, I have changed
5729 it so that, while verifying a sender address, $sender_address is set to <>.
5730 (There is no change to what happens when verifying a recipient address.)
5732 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5733 target A or AAAA records (if not already returned) without resetting the
5734 qualify_single or search_parents options of the DNS resolver. These are
5735 inappropriate in this case because the targets of MX and SRV records must
5736 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5737 target that, when qualified, matched something in the local domain. These
5738 two options are now turned off when doing these lookups.
5740 57. It seems that at least some releases of Reiserfs (which does not have the
5741 concept of a fixed number of inodes) returns zero and not -1 for the
5742 number of available inodes. This interacted badly with check_spool_inodes,
5743 which assumed that -1 was the "no such thing" setting. What I have done is
5744 to check that the total number of inodes is greater than zero before doing
5745 the test of how many are available.
5747 58. When a "warn" ACL statement has a log_message modifier, the message is
5748 remembered, and not repeated. This is to avoid a lot of repetition when a
5749 message has many recipients that cause the same warning to be written.
5750 However, Exim was preserving the list of already written lines for an
5751 entire SMTP session, which doesn't seem right. The memory is now reset if a
5752 new message is started.
5754 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5755 debugging output unless log_rewrite was also set.
5757 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5758 of (int)(handle) when we know that handle contains (void *)(-1).
5760 61. The Exim daemon panic-logs an error return when it closes the incoming
5761 connection. However "connection reset by peer" seems to be common, and
5762 isn't really an error worthy of noting specially, so that particular error
5765 62. When Exim is trying to find all the local interfaces, it used to panic and
5766 die if the ioctl to get the interface flags failed. However, it seems that
5767 on at least one OS (Solaris 9) it is possible to have an interface that is
5768 included in the list of interfaces, but for which you get a failure error
5769 for this call. This happens when the interface is not "plumbed" into a
5770 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5771 failure of the "get flags" call assumes that the interface is down.
5773 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5774 makes life easier for people who are doing arithmetic on fields extracted
5775 from dates, where you often get leading zeros that should not be
5776 interpreted as octal.
5778 64. Added qualify_domain to the redirect router, to override the global
5781 65. If a pathologically long header line contained very many addresses (the
5782 report of this problem mentioned 10 000) and each of them was rewritten,
5783 Exim could use up a very large amount of memory. (It kept on making new
5784 copies of the header line as it rewrote, and never released the old ones.)
5785 At the expense of a bit more processing, the header rewriting function has
5786 been changed so that it no longer eats memory in this way.
5788 66. The generation of the Received: header has been moved from the time that a
5789 message starts to be received, to the time that it finishes. The timestamp
5790 in the Received: header should now be very close to that of the <= log
5791 line. There are two side-effects of this change:
5793 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5794 logged header lines no longer include the local Received: line, because
5795 it has not yet been created. The same applies to a copy of the message
5796 that is returned to a non-SMTP sender when a message is rejected.
5798 (b) When a filter file is tested using -bf, no additional Received: header
5799 is added to the test message. After some thought, I decided that this
5802 This change does not affect the value of $received_for. It is still set
5803 after address rewriting, but before local_scan() is called.
5805 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5807 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5808 gave an unhelpful panic error message, and a defer error. I have managed to
5809 change this behaviour so that it now rejects any supplied certificate,
5810 which seems right, as the list of acceptable certificates is empty.
5812 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5813 gave an unhelpful defer error. I have not managed to make this reject any
5814 supplied certificates, but the error message it gives is "no certificate
5815 supplied", which is not helpful.
5817 70. exigrep's output now also includes lines that are not associated with any
5818 message, but which match the given pattern. Implemented by a patch from
5819 Martin Sluka, which also tidied up the Perl a bit.
5821 71. Recipient callout verification, like sender verification, was using <> in
5822 the MAIL FROM command. This isn't really the right thing, since the actual
5823 sender may affect whether the remote host accepts the recipient or not. I
5824 have changed it to use the actual sender in the callout; this means that
5825 the cache record is now keyed on a recipient/sender pair, not just the
5826 recipient address. There doesn't seem to be a real danger of callout loops,
5827 since a callout by the remote host to check the sender would use <>.
5828 [SEE ABOVE: changed after hitting problems.]
5830 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5831 temporary errors. However, in the case of such a code being given after
5832 the end of a data transmission (i.e. after ".") Exim was failing to write
5833 a retry record for the message. (Yes, there was some broken host that was
5834 actually sending 8xx at this point.)
5836 73. An unknown lookup type in a host list could cause Exim to panic-die when
5837 the list was checked. (An example that provoked this was putting <; in the
5838 middle of a list instead of at the start.) If this happened during a DATA
5839 ACL check, a -D file could be left lying around. This kind of configuration
5840 error no longer causes Exim to die; instead it causes a defer error. The
5841 incident is still logged to the main and panic logs.
5843 74. Buglet left over from Exim 3 conversion. The message "too many messages
5844 in one connection" was written to the rejectlog but not the mainlog, except
5845 when address rewriting (yes!) was being logged.
5847 75. Added write_rejectlog option.
5849 76. When a system filter was run not as root (that is, when system_filter_user
5850 was set), the values of the $n variables were not being returned to the
5851 main process; thus, they were not subsequently available in the $sn
5854 77. Added +return_path_on_delivery log selector.
5856 78. A connection timeout was being treated differently from recipients deferred
5857 when testing hosts_max_try with a message that was older than the host's
5858 retry timeout. (The host should not be counted, thus allowing all hosts to
5859 be tried at least once before bouncing.) This may have been the cause of an
5860 occasionally reported bug whereby a message would remain on the queue
5861 longer than the retry timeout, but would be bounced if a delivery was
5862 forced. I say "may" because I never totally pinned down the problem;
5863 setting up timeout/retry tests is difficult. See also the next item.
5865 79. The ultimate address timeout was not being applied to errors that involved
5866 a combination of host plus message (for example, a timeout on a MAIL
5867 command). When an address resolved to a number of possible hosts, and they
5868 were not all tried for each delivery (e.g. because of hosts_max_try), a
5869 message could remain on the queue longer than the retry timeout.
5871 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5874 81. Fixed an obscure SMTP outgoing bug which required at least the following
5875 conditions: (a) there was another message waiting for the same server;
5876 (b) the server returned 5xx to all RCPT commands in the first message so
5877 that the message was not completed; (c) the server dropped the connection
5878 or gave a negative response to the RSET that Exim sends to abort the
5879 transaction. The observed case was a dropped connection after DATA that had
5880 been sent in pipelining mode. That is, the server had advertised PIPELINING
5881 but was not implementing it correctly. The effect of the bug was incorrect
5882 behaviour, such as trying another host, and this could lead to a crash.
5888 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5889 and daemon.c were passed as pointers to ints; they should have been
5890 pointers to socklen_t variables (which are typically unsigned ints).
5892 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5895 3. Fixed a really odd bug that affected only the testing scheme; patching a
5896 certain fixed string in the binary changed the value of another string that
5897 happened to be identical to the end of the original first string.
5899 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5900 name", it returns that address as the IP address. On some operating
5901 systems (e.g. Solaris), it also passes back the IP address string as the
5902 "host name". However, on others (e.g. Linux), it passes back an empty
5903 string. Exim wasn't checking for this, and was changing the host name to an
5904 empty string, assuming it had been canonicalized.
5906 5. Although rare, it is permitted to have more than one PTR record for a given
5907 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5908 all the names associated with an address, because they do in Solaris.
5909 However, it seems that they do not in Linux for data that comes from the
5910 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5911 I found this out when I moved to a new Linux workstation and tried to run
5912 the Exim test suite.
5914 To get round this problem I have changed the code so that it now does its
5915 own call to the DNS to look up PTR records when searching for a host name.
5916 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5917 addresses that are only in /etc/hosts are still found.
5919 This behaviour is, however, controlled by an option called host_lookup_
5920 order, which defaults to "bydns:byaddr". If people want to use the other
5921 order, or indeed, just use one or the other means of lookup, they can
5922 specify it in this variable.
5924 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5925 some operating systems, this comes back from gethostbyaddr() as an empty
5926 string, and this is what Exim used to test for. However, it seems that in
5927 other systems, "." is yielded. Exim now tests for this case too.
5929 7. The values of check_spool_space and check_log_space are now held internally
5930 as a number of kilobytes instead of an absolute number of bytes. If a
5931 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5932 kilobyte. This means that much larger values can be stored.
5934 8. Exim monitor: an attempt to get the action menu when not actually pointing
5935 at a message produces an empty menu entitled "No message selected". This
5936 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5937 no entries in it ("Shell widget menu has zero width and/or height"). So I
5938 have added a single, blank menu entry in this case.
5940 9. Added ${quote_local_part.
5942 10. MIME decoding is now applied to the contents of Subject: header lines when
5945 11. Now that a reference to $sender_host_address automatically causes a reverse
5946 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5947 host lookup before query-style lookups in lists that might use this
5948 variable. This has therefore been abolished, and the "net-" prefix is no
5949 longer necessary for query-style lookups.
5951 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5952 have been a typo for LFLAGS, so it has been changed.
5954 13. The install script calls Exim with "-C /dev/null" in order to find the
5955 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5956 to be output. However, since Exim outputs its version number before the
5957 error, it didn't break the script. It just looked ugly. I fixed this by
5958 always allowing "-C /dev/null" if the caller is root.
5960 14. Ignore overlarge ACL variable number when reading spool file - insurance
5961 against a later release with more variables having written the file.
5963 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5964 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5965 this, as well as the form without the "IPv6" on the front.
5967 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5968 OS/Makefile-Darwin file.
5970 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5971 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5973 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5974 variable. However, some OS use ints for inodes. I've added cast to long int
5975 to get rid of the compiler warning.
5977 19. I had forgotten to lock out "/../" in configuration file names when
5978 ALT_CONFIG_PREFIX was set.
5980 20. Routers used for verification do not need to specify transports. However,
5981 if such a router generated a host list, and callout was configured, Exim
5982 crashed, because it could not find a port number from the (non-existent)
5983 transport. It now assumes port 25 in this circumstance.
5985 21. Added the -t option to exigrep.
5987 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5988 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5989 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5990 them is set, LOOKUP_LSEARCH is forced.
5992 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5994 24. Added sender and host information to the "rejected by local_scan()" log
5995 line; previously there was no indication of these.
5997 25. Added .include_if_exists.
5999 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6000 Linux. It turns out that not all file systems support this. Apparently some
6001 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6002 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6003 not supported on the file descriptor, is now ignored when Exim is trying to
6004 sync a directory. This applies only to Linux.
6006 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6008 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6009 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6012 29. Added support for maildirsize files from supplied patch (modified a bit).
6014 30. The use of :fail: followed by an empty string could lead Exim to respond to
6015 sender verification failures with (e.g.):
6017 550 Verification failed for <xxx>
6018 550 Sender verify failed
6020 where the first response line was missing the '-' that indicates it is not
6021 the final line of the response.
6023 31. The loop for finding the name of the user that called Exim had a hardwired
6024 limit of 10; it now uses the value of finduser_retries, which is used for
6025 all other user lookups.
6027 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6030 33. Exim was neglecting to zero errno before one call of strtol() when
6031 expanding a string and expecting an integer value. On some systems this
6032 resulted in spurious "integer overflow" errors. Also, it was casting the
6033 result into an int without checking.
6035 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6036 did not work. The code looks as if it has *never* worked, though it appears
6037 to have been documented since at least release 1.62. I have made it work.
6039 35. The "timeout_DNS" error in retry rules, also documented since at least
6040 1.62, also never worked. As it isn't clear exactly what this means, and
6041 clearly it isn't a major issue, I have abolished the feature by treating it
6042 as "timeout", and writing a warning to the main and panic logs.
6044 36. The display of retry rules for -brt wasn't always showing the error code
6047 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6048 timeout_connect_A, timeout_connect_MX.
6050 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6051 to the empty sender.
6053 39. The daemon was not analysing the content of -oX till after it had closed
6054 stderr and disconnected from the controlling terminal. This meant that any
6055 syntax errors were only noted on the panic log, and the return code from
6056 the command was 0. By re-arranging the code a little, I've made the
6057 decoding happen first, so such errors now appear on stderr, and the return
6058 code is 1. However, the actual setting up of the sockets still happens in
6059 the disconnected process, so errors there are still only recorded on the
6062 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6063 connections (as happens on some IP stacks) was logged at start up time as
6064 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6065 it from "IPv6 and IPv4", which means that two separate sockets are being
6068 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6069 says whether AF_INET or AF_INET6 was passed as an argument.
6071 42. Exiwhat output was messed up when time zones were included in log
6074 43. Exiwhat now gives more information about the daemon's listening ports,
6075 and whether -tls-on-connect was used.
6077 44. The "port" option of the smtp transport is now expanded.
6079 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6080 silently ignored. Now an error message is written to the main and panic
6083 46. There's a new ACL modifier called "logwrite" which writes to a log file
6084 as soon as it is encountered.
6086 47. Added $local_user_uid and $local_user_gid at routing time.
6088 48. Exim crashed when trying to verify a sender address that was being
6091 49. Exim was recognizing only a space character after ".include". It now also
6092 recognizes a tab character.
6094 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6095 extracting the relevant information from the specification. The man page no
6096 longer contains scrambled data for the -d option, and I've added a section
6097 at the front about calling Exim under different names.
6099 51. Added "extra_headers" argument to the "mail" command in filter files.
6101 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6104 53. Installed eximstats 1.29.
6106 54. Added transport_filter_timeout as a generic transport option.
6108 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6109 Cc: header lines. This was required by RFC 822, but it not required by RFC
6112 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6113 incoming messages that did not have them. Now it does so only if the
6114 message originates locally, that is, if there is no associated remote host
6115 address. When Resent- header lines are present, this applies to the Resent-
6116 lines rather than the non-Resent- lines.
6118 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6119 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6121 58. Messages for configuration errors now include the name of the main
6122 configuration file - useful now that there may be more than one file in a
6123 list (.included file names were always shown).
6125 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6126 for those rare installations that do not start the daemon as root or run it
6127 setuid root. I've cut out the call to initgroups() if the daemon is not
6130 60. The Exim user and group can now be bound into the binary as text strings
6131 that are looked up at the start of Exim's processing.
6133 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6135 62. Added $mailstore_basename variable.
6137 63. Installed patch to sieve.c from Michael Haardt.
6139 64. When Exim failed to open the panic log after failing to open the main log,
6140 the original message it was trying to log was written to stderr and debug
6141 output, but if they were not available (the usual case in production), it
6142 was lost. Now it is written to syslog before the two lines that record the
6143 failures to open the logs.
6145 65. Users' Exim filters run in subprocesses under the user's uid. It is
6146 possible for a "deliver" command or an alias in a "personal" command to
6147 provoke an address rewrite. If logging of address rewriting is configured,
6148 this fails because the process is not running as root or exim. There may be
6149 a better way of dealing with this, but for the moment (because 4.30 needs
6150 to be released), I have disabled address rewrite logging when running a
6151 filter in a non-root, non-exim process.
6157 1. The buildconfig auxiliary program wasn't quoting the value set for
6158 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6159 not defined. This bug was masked in 4.22 by the effect that was fixed in
6162 2. Some messages that were rejected after a message id was allocated were
6163 shown as "incomplete" by exigrep. It no longer does this for messages that
6164 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6166 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6167 have allow_domain_literals set, the ID did not get logged in the <= line.
6168 Domain literals are now always recognized in Message-ID: header lines.
6170 4. The first argument for a ${extract expansion item is the key name or field
6171 number. Leading and trailing spaces in this item were not being ignored,
6172 causing some misleading effects.
6174 5. When deliver_drop_privilege was set, single queue runner processes started
6175 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6176 same command in the process it spins off) were not dropping privilege.
6178 6. When the daemon running as "exim" started a queue runner, it always
6179 re-executed Exim in the spun-off process. This is a waste of effort when
6180 deliver_drop_privilege is set. The new process now just calls the
6181 queue-runner function directly.
6187 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6190 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6191 ignored. Though the use of -oP was forcing the writing of a pid file, it
6192 was always written to the default place.
6194 3. If the message "no IP address found for host xxxx" is generated during
6195 incoming verification, it is now followed by identification of the incoming
6196 connection (so you can more easily find what provoked it).
6198 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6200 5. Added some features to "harden" Exim a bit more against certain attacks:
6202 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6203 be put in Local/Makefile. This is like the never_users runtime option,
6204 but it cannot be overridden. The default setting is "root".
6206 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6207 prefix string with which any file named in a -C command line option
6210 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6211 is retained for -C and -D only if the caller of Exim is root. Without
6212 it, the exim user may also use -C and -D and retain privilege.
6214 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6215 command line option is disabled.
6217 6. Macro names set by the -D option must start with an upper case letter, just
6218 like macro names defined in the configuration file.
6220 7. Added "dereference=" facility to LDAP.
6222 8. Two instances of the typo "uknown" in the source files are fixed.
6224 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6225 the Configure-Makefile script screwed up while processing it.
6227 10. Incorporated PCRE 4.4.
6229 11. The SMTP synchronization check was not operating right at the start of an
6230 SMTP session. For example, it could not catch a HELO sent before the client
6231 waited for the greeting. There is now a check for outstanding input at the
6232 point when the greeting is written. Because of the duplex, asynchronous
6233 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6234 way, but not yet received, when the check is performed.
6236 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6237 on TCP/IP sockets, because this apparently causes some broken clients to
6240 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6241 unchanged) from the Cygwin maintainer.
6243 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6244 was supported instead of testing for mbx. Effectively a typo.
6246 15. The spa authenticator server code was not checking that the input it
6247 received was valid base64.
6249 16. The debug output line for the "set" modifier in ACLs was not showing the
6250 name of the variable that was being set.
6252 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6254 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6255 lookup on its own. Something else was needed to trigger the lookup. For
6256 example, a match in host_lookup or the need for a host name in a host list.
6257 Now, if $sender_host_name is referenced and the host name has not yet been
6258 looked up, a lookup is performed. If the lookup fails, the variable remains
6259 empty, and $host_lookup_failed is set to "1".
6261 19. Added "eqi" as a case-independent comparison operator.
6263 20. The saslauthd authentication condition could segfault if neither service
6264 nor realm was specified.
6266 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6267 error message that was logged was misleading, and incoming SMTP
6268 connections were dropped. The message is now more accurate, and temporary
6269 errors are given to SMTP connections.
6271 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6272 (except RSET) with a 421 error, until QUIT is received. However, it was
6273 failing to send a response to QUIT.
6275 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6276 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6277 ACL is now run after the helo_try_verify_hosts code.
6279 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6280 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6281 case-independent, so other case variants are also recognized). Apparently
6282 some systems use these upper case variants.
6284 25. If more than two messages were waiting for the same host, and a transport
6285 filter was specified for the transport, Exim sent two messages over the
6286 same TCP/IP connection, and then failed with "socket operation on non-
6287 socket" when it tried to send the third.
6289 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6291 27. The extern definition of crypt16() in expand.c was not being excluded when
6292 the OS had its own crypt16() function.
6294 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6295 as a preferred synonym for return_size_limit, both as an option and as an
6298 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6300 30. Changed the default configuration ACL to relax the local part checking rule
6301 for addresses that are not in any local domains. For these addresses,
6302 slashes and pipe symbols are allowed within local parts, but the sequence
6303 /../ is explicitly forbidden.
6305 31. SPA server authentication was not clearing the challenge buffer before
6308 32. log_message in a "warn" ACL statement was writing to the reject log as
6309 well as to the main log, which contradicts the documentation and doesn't
6310 seem right (because no rejection is happening). So I have stopped it.
6312 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6313 However, I am unable to do any testing of this.
6315 34. Fixed an infelicity in the appendfile transport. When checking directories
6316 for a mailbox, to see if any needed to be created, it was accidentally
6317 using path names with one or more superfluous leading slashes; tracing
6318 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6320 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6321 contents are added to the log line that is written for every discarded
6322 recipient. (Previously a log_message setting was ignored.)
6324 36. The ${quote: operator now quotes the string if it is empty.
6326 37. The install script runs exim in order to find its version number. If for
6327 some reason other than non-existence or emptiness, which it checks, it
6328 could not run './exim', it was installing it with an empty version number,
6329 i.e. as "exim-". This error state is now caught, and the installation is
6332 38. An argument was missing from the function that creates an error message
6333 when Exim fails to connect to the socket for saslauthd authentication.
6334 This could cause Exim to crash, or give a corrupted message.
6336 39. Added isip, isip4, and isip6 to ${if conditions.
6338 40. The ACL variables $acl_xx are now saved with the message, and can be
6339 accessed later in routers, transports, and filters.
6341 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6342 strings in the file are not string-expanded.
6344 42. If a MAIL command specified a SIZE value that was too large to fit into an
6345 int variable, the check against message_size_limit failed. Such values are
6346 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6347 day this will have to be increased, but I don't think I want to be around
6348 when emails are that large.
6355 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6356 iconv() is not standard in FreeBSD.
6358 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6359 IPv6 enabled. The observed symptom was a segmentation fault on return from
6360 the function os_common_find_running_interfaces() in src/os.c.
6362 3. In the check_special_case() function in daemon.c I had used "errno" as an
6363 argument name, which causes warnings on some systems. This was basically a
6364 typo, since it was named "eno" in the comments!
6366 4. The code that waits for the clock to tick (at a resolution of some fraction
6367 of a second) so as to ensure message-id uniqueness was always waiting for
6368 at least one whole tick, when it could have waited for less. [This is
6369 almost certainly not relevant at current processor speeds, where it is
6370 unlikely to ever wait at all. But we try to future-proof.]
6372 5. The function that sleeps for a time interval that includes fractions of a
6373 second contained a race. It did not block SIGALRM between setting the
6374 timer, and suspending (a couple of lines later). If the interval was short
6375 and the sigsuspend() was delayed until after it had expired, the suspension
6376 never ended. On busy systems this could lead to processes getting stuck for
6379 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6380 process, before it forks any delivery processes. The open lookup caching
6381 mechanism meant that the open file or database connection was passed into
6382 the delivery process. The problem was that delivery processes always tidy
6383 up cached lookup data. This could cause a problem for the next delivery
6384 process started by the queue runner, because the external queue runner
6385 process does not know about the closure. So the next delivery process
6386 still has data in the lookup cache. In the case of a file lookup, there was
6387 no problem because closing a file descriptor in a subprocess doesn't affect
6388 the parent. However, if the lookup was caching a connection to a database,
6389 the connection was closed, and the second delivery process was likely to
6390 see errors such as "PGSQL: query failed: server closed the connection
6391 unexpectedly". The problem has been fixed by closing all cached lookups
6392 in a queue runner before running a delivery process.
6394 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6395 seem to have the "const" qualifier which it has on other OS. I've
6398 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6399 *of the same type* (client or server) with the same public name that an
6400 error should be diagnosed.
6402 9. When Exim looked up a host name for an IP address, but failed to find the
6403 original IP address when looking up the host name (a safety check), it
6404 output the message "<ip address> does not match any IP for NULL", which was
6405 confusing, to say the least. The bug was that the host name should have
6406 appeared instead of "NULL".
6408 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6409 user that is built into the binary, and the -C or -D options is used, root
6410 privilege is dropped before the configuration file is read. In addition,
6411 logging is switched to stderr instead of the normal log files. If the
6412 configuration then re-defines the Exim user, the unprivileged environment
6413 is probably not what is expected, so Exim logs a panic warning message (but
6416 However, if deliver_drop_privilege is set, the unprivileged state may well
6417 be exactly what is intended, so the warning has been cut out in that case,
6418 and Exim is allowed to try to write to its normal log files.
6424 1. smtp_return_error_details was not giving details for temporary sender
6425 or receiver verification errors.
6427 2. Diagnose a configuration error if two authenticators have the same public
6430 3. Exim used not to create the message log file for a message until the first
6431 delivery attempt. This could be confusing when incoming messages were held
6432 for policy or load reasons. The message log file is now created at the time
6433 the message is received, and an initial "Received" line is written to it.
6435 4. The automatically generated man page for command line options had a minor
6436 bug that caused no ill effects; however, a more serious problem was that
6437 the procedure for building the man page automatically didn't always
6438 operate. Consequently, release 4.20 contains an out-of-date version. This
6439 shouldn't happen again.
6441 5. When building Exim with embedded Perl support, the script that builds the
6442 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6443 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6445 6. The freeze_tell option was not being used for messages that were frozen on
6446 arrival, either by an ACL or by local_scan().
6448 7. Added the smtp_incomplete_transaction log selector.
6450 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6451 was accepting AUTH without a new EHLO.
6453 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6454 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6457 10. Logging of TCP/IP connections (when configured) now happens in the main
6458 daemon process instead of the child process, so that the TCP/IP connection
6459 count is more accurate (but it can never be perfect).
6461 11. The use of "drop" in a nested ACL was not being handled correctly in the
6462 outer ACL. Now, if condition failure induced by the nested "drop" causes
6463 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6464 or "require"), the connection is dropped.
6466 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6467 that yield "discard" can now be used with an "accept" or a "discard" verb,
6468 but an error is generated for any others (because I can't see a useful way
6469 to define what should happen).
6471 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6472 are now processed in the same way as lines in the Exim configuration file.
6473 In particular, continuation lines are supported.
6475 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6477 15. Added -ti meaning -t -i.
6479 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6480 domains, and warn by logging if others are found.
6482 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6483 returned from GIFCONF when Exim is trying to find the list of interfaces on
6484 a host. The code in os.c has been modified to copy each ifreq to an aligned
6485 structure in all cases.
6487 Also, in some cases, the returned ifreq's were being copied to a 'struct
6488 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6489 means the last couple of bytes of an IPv6 address could be chopped if the
6490 ifreq contained only a normal sockaddr (14 bytes storage).
6492 18. Named domain lists were not supported in the hosts_treat_as_local option.
6493 An entry such as +xxxx was not recognized, and was treated as a literal
6496 19. Ensure that header lines added by a DATA ACL are included in the reject log
6497 if the ACL subsequently rejects the message.
6499 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6500 MD5 (which is deprecated).
6502 21. When testing a filter file using -bf, Exim was writing a message when it
6503 took the sender from a "From " line in the message, but it was not doing so
6504 when it took $return_path from a Return-Path: header line. It now does.
6506 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6507 with a valid header line field name (a series of printing characters
6508 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6510 23. Changed "disc" in the source to "disk" to conform to the documentation and
6511 the book and for uniformity.
6513 24. Ignore Sendmail's -Ooption=value command line item.
6515 25. When execve() failed while trying to run a command in a pipe transport,
6516 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6517 could be confused with a return value of 69 from the command itself. This
6518 has been changed to 127, the value the shell returns if it is asked to run
6519 a non-existent command. The wording for the related log line suggests a
6520 non-existent command as the problem.
6522 26. If received_header_text expands to an empty string, do not add a Received:
6523 header line to the message. (Well, it adds a token one on the spool, but
6524 marks it "old" so that it doesn't get used or transmitted.)
6526 27. Installed eximstats 1.28 (addition of -nt option).
6528 28. There was no check for failure on the call to getsockname() in the daemon
6529 code. This can fail if there is a shortage of resources on the system, with
6530 ENOMEM, for example. A temporary error is now given on failure.
6532 29. Contrary to the C standard, it seems that in some environments, the
6533 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6534 program. Exim now does this explicitly; it affects the formatting of
6535 timestamps using strftime().
6537 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6538 complaints. I've now initialized all the variables, to avoid this.
6540 32. Header lines added by a system filter were not being "seen" during
6541 transport-time rewrites.
6543 33. The info_callback() function passed to OpenSSL is set up with type void
6544 (*)(SSL *, int, int), as described somewhere. However, when calling the
6545 function (actually a macro) that sets it up, the type void(*)() is
6546 expected. I've put in a cast to prevent warnings from picky compilers.
6548 34. If a DNS black list lookup found a CNAME record, but there were no A
6549 records associated with the domain it pointed at, Exim crashed.
6551 35. If a DNS black list lookup returned more than one A record, Exim ignored
6552 all but the first. It now scans all returned addresses if a particular IP
6553 value is being sought. In this situation, the contents of the
6554 $dnslist_value variable are a list of all the addresses, separated by a
6557 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6558 to accept a host name and all its aliases if the forward lookup for any of
6559 them yielded the IP address of the incoming connection. Now it accepts only
6560 those names whose forward lookup yields the correct IP address. Any other
6561 names are discarded. This closes a loophole whereby a rogue DNS
6562 administrator could create reverse DNS records to break through a
6563 wildcarded host restriction in an ACL.
6565 37. If a user filter or a system filter that ran in a subprocess used any of
6566 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6567 the wrong values were passed to the pipe command ($thisaddress had the
6568 value of $0, $0 had the value of $1, etc). This bug was introduced by
6569 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6571 38. Improved the line breaking for long SMTP error messages from ACLs.
6572 Previously, if there was no break point between 40 and 75 characters, Exim
6573 left the rest of the message alone. Two changes have been made: (a) I've
6574 reduced the minimum length to 35 characters; (b) if it can't find a break
6575 point between 35 and 75 characters, it looks ahead and uses the first one
6576 that it finds. This may give the occasional overlong line, but at least the
6577 remaining text gets split now.
6579 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6580 file descriptors might be low, and that setting 1000 would always raise it.
6581 It turns out that in some environments, the limit is already over 1000 and
6582 that lowering it causes trouble. So now Exim takes care not to decrease it.
6584 40. When delivering a message, the value of $return_path is set to $sender_
6585 address at the start of routing (routers may change the value). By an
6586 oversight, this default was not being set up when an address was tested by
6587 -bt or -bv, which affected the outcome if any router or filter referred to
6590 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6591 log only when "message" or "log_message" are set. However, if one of the
6592 conditions was an address verification, or a call to a nested ACL, the
6593 messages generated by the underlying test were being passed through. This
6594 no longer happens. The underlying message is available in $acl_verify_
6595 message for both "message" and "log_message" expansions, so it can be
6596 passed through if needed.
6598 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6599 new expansion $bh_ to give the encoded byte string without charset
6600 translation. Translation happens only if iconv() is available; HAVE_ICONV
6601 indicates this at build time. HEADERS_CHARSET gives the charset to
6602 translate to; headers_charset can change it in the configuration, and
6603 "headers charset" can change it in an individual filter file.
6605 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6606 that creates RFC 2047 encoded "words" labels them as that charset instead
6607 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6608 expansion operator; (ii) when Exim creates a From: line for a local
6609 message; (iii) when a header line is rewritten to include a "phrase" part.
6611 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6612 buggy, causing it to skip the first lines of messages whose message ID
6613 ended in 'D'. This would not have bitten before Exim release 4.14, because
6614 message IDs were unlikely to end in 'D' before then. The effect was to have
6615 incorrect size information for certain domains.
6617 45. #include "config.h" was missing at the start of the crypt16.c module. This
6618 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6621 46. If there was a timeout during a "random" callout check, Exim treated it as
6622 a failure of the random address, and carried on sending RSET and the real
6623 address. If the delay was just some slowness somewhere, the response to the
6624 original RCPT would be taken as a response to RSET and so on, causing
6625 mayhem of various kinds.
6627 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6628 when I implemented it. It didn't allow for the fact that some option values
6629 may legitimately be negative (e.g. size_addition), and it didn't even do
6630 the right test for positive values.
6632 48. Domain names in DNS records are case-independent. Exim always looks them up
6633 in lower case. Some resolvers return domain names in exactly the case they
6634 appear in the zone file, that is, they may contain uppercase letters. Not
6635 all resolvers do this - some return always lower case. Exim was treating a
6636 change of case by a resolver as a change of domain, similar to a widening
6637 of a domain abbreviation. This triggered its re-routing code and so it was
6638 trying to route what was effectively the same domain again. This normally
6639 caused routing to fail (because the router wouldn't handle the domain
6640 twice). Now Exim checks for this case specially, and just changes the
6641 casing of the domain that it ultimately uses when it transmits the message
6644 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6647 50. If a filter generated a file delivery with a non-absolute name (possible if
6648 no home directory exists for the router), the forbid_file option was not
6651 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6652 the existing equality matching.
6654 52. Exim was using ints instead of ino_t variables in some places where it was
6655 dealing with inode numbers.
6657 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6658 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6659 called TMPDIR, and if it finds it is different, it changes its value.
6661 54. The smtp_printf() function is now made available to local_scan() so
6662 additional output lines can be written before returning. There is also an
6663 smtp_fflush() function to enable the detection of a dropped connection.
6664 The variables smtp_input and smtp_batched_input are exported to
6667 55. Changed the default runtime configuration: the message "Unknown user"
6668 has been removed from the ACL, and instead placed on the localuser router,
6669 using the cannot_route_message feature. This means that any verification
6670 failures that generate their own messages won't get overridden. Similarly,
6671 the "Unrouteable address" message that was in the ACL for unverifiable
6672 relay addresses has also been removed.
6674 56. Added hosts_avoid_esmtp to the smtp transport.
6676 57. The exicyclog script was not checking for the esoteric option
6677 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6678 will work only if exicyclog is run under the appropriate euid.
6680 58. Following a discussion on the list, the rules by which Exim recognises line
6681 endings on incoming messages have been changed. The -dropcr and drop_cr
6682 options are now no-ops, retained only for backwards compatibility. The
6683 following line terminators are recognized: LF CRLF CR. However, special
6684 processing applies to CR:
6686 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6687 nor a local message in the state where . is a terminator.
6689 (ii) If a bare CR is encountered in a header line, an extra space is added
6690 after the line terminator so as not to end the header. The reasoning
6691 behind this is that bare CRs in header lines are most likely either
6692 to be mistakes, or people trying to play silly games.
6694 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6695 was being incorrectly given as 18 bytes larger than it should have been.
6696 This is a VOB (very old bug).
6698 60. This may never have affected anything current, but just in case it has:
6699 When the local host is found other than at the start of a list of hosts,
6700 the local host, those with the same MX, and any that follow, are discarded.
6701 When the list in question was part of a longer list of hosts, the following
6702 hosts (not currently being processed) were also being discarded. This no
6703 longer happens. I'm not sure if this situation could ever has previously
6706 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6709 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6710 header. This is something that is recommended in a new Internet Draft, and
6711 is something that is documented as being done by Sendmail. There are two
6712 possible values. For messages generated by the autoreply transport, Exim
6715 Auto-Submitted: auto-replied
6717 whereas for all other generated messages (e.g. bounces) it adds
6719 Auto-Submitted: auto-generated
6721 63. The "personal" condition in filters now includes a test for the
6722 Auto-Submitted: header. If it contains the string "auto-" the message it
6723 not considered personal.
6725 64. Added rcpt_include_affixes as a generic transport option.
6727 65. Added queue_only_override (default true).
6729 66. Added the syslog_duplication option.
6731 67. If what should have been the first header line of a message consisted of
6732 a space followed by a colon, Exim was mis-interpreting it as a header line.
6733 It isn't of course - it is syntactically invalid and should therefore be
6734 treated as the start of the message body. The misbehaviour could have
6735 caused a number of strange effects, including loss of data in subsequent
6736 header lines, and spool format errors.
6738 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6739 client host had authenticated. This control can now be exercised by an ACL
6740 for more flexibility.
6742 69. By default, callouts do not happen when testing with -bh. There is now a
6743 variant, -bhc, which does actually run the callout code, including
6744 consulting and updating the callout cache.
6746 70. Added support for saslauthd authentication, courtesy of Alexander
6749 71. If statvfs() failed on the spool or log directories while checking their
6750 size for availability, Exim confusingly gave the error "space shortage".
6751 Furthermore, in debugging mode it crashed with a floating point exception.
6752 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6753 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6754 serious problem, Exim now writes to the main and panic logs when this
6755 happens, with details of the failure. It then refuses to accept the
6756 incoming message, giving the message "spool directory problem" or "log
6757 directory problem" with a 421 code for SMTP messages.
6759 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6760 0, 1, and 2 exist, because some OS complain for execs without them (see
6761 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6762 descriptors. However, the code omitted to check that the open succeeded,
6763 causing mysterious errors if for some reason the permissions on /dev/null
6764 got screwed. Now Exim writes a message to the main and panic logs, and
6765 bombs out if it can't open /dev/null.
6767 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6768 interact so that it is all more flexible. It is supposed to remain
6769 backwards compatible. Also added extra_local_interfaces.
6771 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6772 to bomb out with an assertion failure - to the client this appears as a
6773 connection drop. This problem occurs in the part of the code that was taken
6774 from the Samba project. Fortunately, the assertion is in a very simple
6775 function, so I have fixed this by reproducing the function inline in the
6776 one place where it is called, and arranging for authentication to fail
6777 instead of killing the process with assert().
6779 75. The SPA client code was not working when the server requested OEM rather
6780 than Unicode encoding.
6782 76. Added code to make require_files with a specific uid setting more usable in
6783 the case where statting the file as root fails - usually a non-root-mounted
6784 NFS file system. When this happens and the failure is EACCES, Exim now
6785 forks a subprocess and does the per-uid checking as the relevant uid.
6787 77. Added process_log_path.
6789 78. If log_file_path was not explicitly set, a setting of check_log_space or
6790 check_log_inodes was ignored.
6792 79. If a space check for the spool or log partitions fails, the incident is now
6793 logged. Of course, in the latter case the data may get lost...
6795 80. Added the %p formatting code to string_format() so that it can be used to
6796 print addresses in debug_print(). Adjusted all the address printing in the
6797 debugging in store.c to use %p rather than %d.
6799 81. There was a concern that a line of code in smtp_in.c could overflow a
6800 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6801 initially expressed, the concern was not well-founded, because trailing
6802 spaces are removed early. However, if the trailing spaces were followed by
6803 a NULL, they did not get removed, so the overflow was possible. Two fixes
6806 (a) I re-wrote the offending code in a cleaner fashion.
6807 (b) If an incoming SMTP command contains a NULL character, it is rejected
6810 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6811 runs initgroups(), so that if the Exim user is in any additional groups,
6812 they will be used during message reception.
6818 The change log for 4.20 and earlier releases has been archived.