1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Rework error string handling in TLS interface so that the caller in
12 more cases is responsible for logging. This permits library-sourced
13 string to be attached to addresses during delivery, and collapses
14 pairs of long lines into single ones.
16 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
17 during configuration. Wildcards are allowed and expanded.
19 JH/02 Rework error string handling in DKIM to pass more info back to callers.
20 This permits better logging.
22 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
23 do not close it down and have the child transport start it up again on
24 the passed-on TCP connection. Instead, proxy the child (and any
25 subsequent ones) for TLS via a unix-domain socket channel. Logging is
26 affected: the continued delivery log lines do not have any DNSSEC, TLS
27 Certificate or OCSP information. TLS cipher information is still logged.
29 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
30 identical IP addresses on different listening ports. Will also affect
33 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
34 add noisy ifdef guards to special-case this sillyness.
35 Patch from Bernd Kuhls.
37 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
38 than 255 are no longer allowed.
40 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
41 Disable the session-cache too, which might reduce our load. Since we
42 currrectly use a new context for every connection, both as server and
43 client, there is no benefit for these.
44 GnuTLS appears to not support tickets server-side by default (we don't
45 call gnutls_session_ticket_enable_server()) but client side is enabled
46 by default on recent versions (3.1.3 +) unless the PFS priority string
49 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
50 <https://reproducible-builds.org/specs/source-date-epoch/>.
52 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
53 the check for any unsuccessful recipients did not notice the limit, and
54 erroneously found still-pending ones.
56 JH/08 Pipeline CHUNKING command and data together, on kernels that support
57 MSG_MORE. Only in-clear (not on TLS connections).
59 JH/09 Avoid using a temporary file during transport using dkim. Unless a
60 transport-filter is involved we can buffer the headers in memory for
61 creating the signature, and read the spool data file once for the
62 signature and again for transmission.
64 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
65 4.77 as the kernel support then wasn't solid, having issues in 64bit
66 mode. Now, it's been long enough. Add support for FreeBSD also.
68 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
69 case where the routing stage had gathered several addresses to send to
70 a host before calling the transport for the first, we previously failed
71 to close down TLS in the old transport process before passing the TCP
72 connection to the new process. The new one sent a STARTTLS command
73 which naturally failed, giving a failed delivery and bloating the retry
74 database. Investigation and fix prototype from Wolfgang Breyha.
76 JH/12 Fix check on SMTP command input synchronisation. Previously there were
77 false-negatives in the check that the sender had not preempted a response
78 or prompt from Exim (running as a server), due to that code's lack of
79 awareness of the SMTP input buffering.
81 PP/04 Add commandline_checks_require_admin option.
82 Exim drops privileges sanely, various checks such as -be aren't a
83 security problem, as long as you trust local users with access to their
84 own account. When invoked by services which pass untrusted data to
85 Exim, this might be an issue. Set this option in main configuration
86 AND make fixes to the calling application, such as using `--` to stop
89 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
90 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
91 responses to those, into a single TLS record each way (this usually means
92 a single packet). As a side issue, smtp_enforce_sync now works on TLS
95 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
96 affects you only if you're dancing at the edge of the param size limits.
97 If you are, and this message makes sense to you, then: raise the
98 configured limit or use OpenSSL 1.1. Nothing we can do for older
105 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
106 than -2003 did; needs libidn2 in addition to libidn.
108 JH/02 The path option on a pipe transport is now expanded before use.
110 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
111 Patch provided by "Björn", documentation fix added too.
113 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
114 missing a wire-to-host endian conversion.
116 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
117 close after a BDAT command line could be taken as a following command,
118 giving a synch failure. Fix by only checking for synch immediately
119 before acknowledging the chunk.
121 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
122 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
123 macro. Patches provided by Josh Soref.
125 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
126 Previously we did not; the RFC seems ambiguous and VRFY is not listed
127 by IANA as a service extension. However, John Klensin suggests that we
130 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
131 the dkim code may be unix-mode line endings rather than smtp wire-format
132 CRLF, so prepend a CR to any bare LF.
134 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
135 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
137 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
138 the first were themselves being wrongly included in the feed into dkim
139 processing; with most chunk sizes in use this resulted in an incorrect
140 body hash calculated value.
142 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
143 DKIM signature block, for verification. Although advised against by
144 standards it is specifically not ruled illegal.
146 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
148 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
149 missing a body hash (the bh= tag).
151 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
152 It seems that HAProxy sends the Proxy Protocol information in clear and
153 only then does a TLS startup, so do the same.
155 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
156 TCP connections (such as for Spamd) unless the daemon successfully set
157 Fast Open mode on its listening sockets. This fixes breakage seen on
158 too-old kernels or those not configured for Fast Open, at the cost of
159 requiring both directions being enabled for TFO, and TFO never being used
160 by non-daemon-related Exim processes.
162 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
163 endings, at least on the first header line. Try to canonify any that get
164 past that check, despite the cost.
166 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
167 now limited to an arbitrary five deep, while parsing addresses with the
168 strip_excess_angle_brackets option enabled.
170 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
171 instead leave the unprompted TLS handshake in socket buffer for the
172 TLS library to consume.
174 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
176 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
178 JH/16 Drop variables when they go out of scope. Memory management drops a whole
179 region in one operation, for speed, and this leaves assigned pointers
180 dangling. Add checks run only under the testsuite which checks all
181 variables at a store-reset and panics on a dangling pointer; add code
182 explicitly nulling out all the variables discovered. Fixes one known
183 bug: a transport crash, where a dangling pointer for $sending_ip_address
184 originally assigned in a verify callout, is re-used.
186 PP/06 Drop '.' from @INC in various Perl scripts.
188 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
190 PP/08 Reduce a number of compilation warnings under clang; building with
191 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
192 should be warning-free.
194 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
196 HS/01 Fix portability problems introduced by PP/08 for platforms where
197 realloc(NULL) is not equivalent to malloc() [SunOS et al].
199 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
200 chunk. This allows us to accept broken chunked messages. We need a more
201 general solution here.
203 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
204 already-broken messages in the queue.
206 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
208 JH/19 Fix reference counting bug in routing-generated-address tracking.
214 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
215 supports it and a size is available (ie. the sending peer gave us one).
217 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
218 years of being deprecated). The replacements are the ACLs
219 acl_smtp_mime and acl_not_smtp_mime.
221 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
222 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
223 or even an in-clear connection were permitted. Now, if the host lookup
224 was dnssec and dane was requested then the host is only used if the
225 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
226 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
227 if one fails this test.
228 This means that a poorly-configured remote DNS will make it incommunicado;
229 but it protects against a DNS-interception attack on it.
231 JH/04 Bug 1810: make continued-use of an open smtp transport connection
232 non-noisy when a race steals the message being considered.
234 JH/05 If main configuration option tls_certificate is unset, generate a
235 self-signed certificate for inbound TLS connections.
237 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
238 in rewrites and routers.
240 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
241 and logged a warning sing 4.83; now they are a configuration file error.
243 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
244 (lacking @domain). Apply the same qualification processing as RCPT.
246 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
248 JH/10 Support ${sha256:} applied to a string (as well as the previous
251 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
252 a cutthrough deliver is pending, as we always want to make a connection.
253 This also avoids re-routing the message when later placing the cutthrough
254 connection after a verify cache hit.
255 Do not update it with the verify result either.
257 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
258 when routing results in more than one destination address.
260 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
261 signing (which inhibits the cutthrough capability). Previously only
262 the presence of an option was tested; now an expansion evaluating as
263 empty is permissible (obviously it should depend only on data available
264 when the cutthrough connection is made).
266 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
267 the relevant preceding SMTP command did not note the pipelining mode.
269 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
270 Previously they were not counted.
272 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
273 as one having no matching records. Previously we deferred the message
274 that needed the lookup.
276 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
277 distinguished as "(=".
279 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
280 for missing MX records. Previously it only worked for missing A records.
282 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
284 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
285 after the data-go-ahead and data-ack. Patch from Jason Betts.
287 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
288 even for a "none" policy. Patch from Tony Meyer.
290 JH/22 Fix continued use of a connection for further deliveries. If a port was
291 specified by a router, it must also match for the delivery to be
294 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
295 When one of the recipients of a message was unsuitable for the connection
296 (has no matching addresses), we lost track of needing to mark it
297 deferred. As a result mail would be lost.
299 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
301 JH/25 Decoding ACL controls is now done using a binary search; the source code
302 takes up less space and should be simpler to maintain. Merge the ACL
303 condition decode tables also, with similar effect.
305 JH/26 Fix problem with one_time used on a redirect router which returned the
306 parent address unchanged. A retry would see the parent address marked as
307 delivered, so not attempt the (identical) child. As a result mail would
310 JH/27 Fix a possible security hole, wherein a process operating with the Exim
311 UID can gain a root shell. Credit to http://www.halfdog.net/ for
312 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
315 JH/28 Enable {spool,log} filesystem space and inode checks as default.
316 Main config options check_{log,spool}_{inodes,space} are now
317 100 inodes, 10MB unless set otherwise in the configuration.
319 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
320 Previously it only applied to the main-section connection policy
323 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
325 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
326 by me. Added RFC7919 DH primes as an alternative.
328 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
329 are not in the system include path.
331 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
332 GnuTLS, when a session startup failed (eg because the client disconnected)
333 Exim did stdio operations after fclose. This was exposed by a recent
334 change which nulled out the file handle after the fclose.
336 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
337 signed directly by the cert-signing cert, rather than an intermediate
338 OCSP-signing cert. This is the model used by LetsEncrypt.
340 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
342 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
343 an incoming connection.
345 HS/02 Bug 1802: Do not half-close the connection after sending a request
348 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
349 fallback to "prime256v1".
351 JH/34 SECURITY: Use proper copy of DATA command in error message.
352 Could leak key material. Remotely exploitable. CVE-2016-9963.
358 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
359 and 3.4.4 - once the server is enabled to respond to an OCSP request
360 it does even when not requested, resulting in a stapling non-aware
361 client dropping the TLS connection.
363 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
364 support variable-length bit vectors. No functional change.
366 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
367 The I= interface field on outgoing lines is now after the H= remote
368 host field, same as incoming lines. There is a separate
369 outgoing_interface log selector which allows you to disable the
372 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
373 If not running log_selector +smtp_connection the mainlog would be held
374 open indefinitely after a "too many connections" event, including to a
375 deleted file after a log rotate. Leave the per net connection logging
376 leaving it open for efficiency as that will be quickly detected by the
377 check on the next write.
379 HS/01 Bug 1671: Fix post transport crash.
380 Processing the wait-<transport> messages could crash the delivery
381 process if the message IDs didn't exist for some reason. When
382 using 'split_spool_directory=yes' the construction of the spool
383 file name failed already, exposing the same netto behaviour.
385 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
386 mime_regex ACL conditions.
388 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
389 to DSN fail messages (bounces): remote IP, remote greeting, remote response
390 to HELO, local diagnostic string.
392 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
393 log line to debug. Even when configured with a tls authenticator many
394 client connections are expected to not authenticate in this way, so
395 an authenticate fail is not an error.
397 HS/02 Add the Exim version string to the process info. This way exiwhat
398 gives some more detail about the running daemon.
400 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
401 matter for fast-change records such as DNSBLs.
403 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
404 retry record, even if constant. There may be multiple transports with
405 different interface settings and the retry behaviour needs to be kept
408 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
410 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
412 JH/10 Bug 840: fix log_defer_output option of pipe transport
414 JH/11 Bug 830: use same host for all RCPTS of a message, even under
415 hosts_randomize. This matters a lot when combined with mua_wrapper.
417 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
418 ${quote_pgsql:<string>} operator.
420 JH/13 Bug 1708: avoid misaligned access in cached lookup.
422 JH/14 Change header file name for freeradius-client. Relevant if compiling
423 with Radius support; from the Gentoo tree and checked under Fedora.
425 JH/15 Bug 1712: Introduce $prdr_requested flag variable
427 JH/16 Bug 1714: Permit an empty string as expansion result for transport
428 option transport_filter, meaning no filtering.
430 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
432 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
433 defaults to "*" (all hosts). The variable is now available when not built
434 with TLS, default unset, mainly to enable keeping the testsuite sane.
435 If a server certificate is not supplied (via tls_certificate) an error is
436 logged, and clients will find TLS connections fail on startup. Presumably
437 they will retry in-clear.
438 Packagers of Exim are strongly encouraged to create a server certificate
439 at installation time.
441 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
442 with the $config_file variable.
444 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
445 in transport context, after the attempt, and per-recipient. The latter type
446 is per host attempted. The event data is the error message, and the errno
447 information encodes the lookup type (A vs. MX) used for the (first) host,
448 and the trailing two digits of the smtp 4xx response.
450 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
451 to write to mainlog (or rejectlog, paniclog) in the window between file
452 creation and permissions/ownership being changed. Particularly affects
453 installations where exicyclog is run as root, rather than exim user;
454 result is that the running daemon panics and dies.
456 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
458 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
459 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
460 "pri" and "weight". Note that the previous implicit priority given by the
461 list order is no longer honoured.
463 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
466 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
467 by defining SUPPORT_SOCKS.
469 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
470 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
471 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
472 variables are renamed to proxy_{local,external}_{address,port}.
474 JH/27 Move Internationalisation support from Experimental to mainline, enabled
475 for a build by defining SUPPORT_I18N
477 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
478 of the query string, and make ${quote_redis:} do that quoting.
480 JH/29 Move Events support from Experimental to mainline, enabled by default
481 and removable for a build by defining DISABLE_EVENT.
483 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
485 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
486 cached by the daemon.
488 JH/32 Move Redis support from Experimental to mainline, enabled for a build
489 by defining LOOKUP_REDIS. The libhiredis library is required.
491 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
492 keys are given for lookup.
494 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
495 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
496 only supported when built with TLS support. The PolarSSL SHA routines
497 are still used when the TLS library is too old for convenient support.
499 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
500 openssl_options), for security. OpenSSL forces this from version 1.1.0
501 server-side so match that on older versions.
503 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
504 allocation for $value could be released as the expansion processing
505 concluded, but leaving the global pointer active for it.
507 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
508 and to use the domains and local_parts ACL conditions.
510 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
511 incorrectly not doubled on cutthrough transmission, hence seen as a
512 body-termination at the receiving system - resulting in truncated mails.
513 Commonly the sender saw a TCP-level error, and retransmitted the message
514 via the normal store-and-forward channel. This could result in duplicates
515 received - but deduplicating mailstores were liable to retain only the
516 initial truncated version.
518 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
520 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
522 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
523 we're in there, support oversigning also; bug 1309.
525 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
527 HS/04 Add support for keep_environment and add_environment options.
529 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
530 either intentional arithmetic overflow during PRNG, or testing config-
533 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
534 delivery resulted in actual delivery. Cancel cutthrough before DATA
537 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
538 rejects a recipient: pass the reject to the originator.
540 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
541 Many were false-positives and ignorable, but it's worth fixing the
544 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
545 for the new environment-manipulation done at startup. Move the routines
546 from being local to tls.c to being global via the os.c file.
548 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
549 an extract embedded as result-arg for a map, the first arg for extract
550 is unavailable so we cannot tell if this is a numbered or keyed
551 extraction. Accept either.
557 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
560 JH/02 The smtp transport option "multi_domain" is now expanded.
562 JH/03 The smtp transport now requests PRDR by default, if the server offers
565 JH/04 Certificate name checking on server certificates, when exim is a client,
566 is now done by default. The transport option tls_verify_cert_hostnames
567 can be used to disable this per-host. The build option
568 EXPERIMENTAL_CERTNAMES is withdrawn.
570 JH/05 The value of the tls_verify_certificates smtp transport and main options
571 default to the word "system" to access the system default CA bundle.
572 For GnuTLS, only version 3.0.20 or later.
574 JH/06 Verification of the server certificate for a TLS connection is now tried
575 (but not required) by default. The verification status is now logged by
576 default, for both outbound TLS and client-certificate supplying inbound
579 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
582 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
583 Status Notification (bounce) messages are now MIME format per RFC 3464.
584 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
585 under the control of the dsn_advertise_hosts option, and routers may
586 have a dsn_lasthop option.
588 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
589 default, modifiable by a malware= option. The list separator for
590 the options can now be changed in the usual way. Bug 68.
592 JH/10 The smtp_receive_timeout main option is now expanded before use.
594 JH/11 The incoming_interface log option now also enables logging of the
595 local interface on delivery outgoing connections.
597 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
598 if the interface and destination host and port all match.
600 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
603 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
604 Patch from Andrew Lewis.
606 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
607 now supports optional time-restrictions, weighting, and priority
608 modifiers per server. Patch originally by <rommer@active.by>.
610 JH/16 The spamd_address main option now supports a mixed list of local
611 and remote servers. Remote servers can be IPv6 addresses, and
612 specify a port-range.
614 JH/17 Bug 68: The spamd_address main option now supports an optional
615 timeout value per server.
617 JH/18 Bug 1581: Router and transport options headers_add/remove can
618 now have the list separator specified.
620 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
623 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
626 JH/21 Support for the A6 type of dns record is withdrawn.
628 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
629 rather than the verbs used.
631 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
632 from 255 to 1024 chars.
634 JH/24 Verification callouts now attempt to use TLS by default.
636 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
637 are generic router options now. The defaults didn't change.
639 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
640 Original patch from Alexander Shikoff, worked over by JH.
642 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
643 files if ZCAT_COMMAND is not executable.
645 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
647 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
649 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
650 Normally benign, it bites when the pair was led to by a CNAME;
651 modern usage is to not canonicalize the domain to a CNAME target
652 (and we were inconsistent anyway for A-only vs AAAA+A).
654 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
656 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
657 when evaluating $sender_host_dnssec.
659 JH/31 Check the HELO verification lookup for DNSSEC, adding new
660 $sender_helo_dnssec variable.
662 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
664 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
666 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
668 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
669 documented as working, but never had. Support all but $spam_report.
671 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
672 added for tls authenticator.
674 HS/03 Add perl_taintmode main config option
680 TL/01 When running the test suite, the README says that variables such as
681 no_msglog_check are global and can be placed anywhere in a specific
682 test's script, however it was observed that placement needed to be near
683 the beginning for it to behave that way. Changed the runtest perl
684 script to read through the entire script once to detect and set these
685 variables, reset to the beginning of the script, and then run through
686 the script parsing/test process like normal.
688 TL/02 The BSD's have an arc4random API. One of the functions to induce
689 adding randomness was arc4random_stir(), but it has been removed in
690 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
691 function when detected.
693 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
694 cause callback expansion.
696 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
697 syntax errors in an expansion can be treated as a string instead of
698 logging or causing an error, due to the internal use of bool_lax
699 instead of bool when processing it.
701 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
702 server certificates when making smtp deliveries.
704 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
706 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
708 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
710 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
711 Merged patch from Sebastian Wiedenroth.
713 JH/05 Fix results-pipe from transport process. Several recipients, combined
714 with certificate use, exposed issues where response data items split
715 over buffer boundaries were not parsed properly. This eventually
716 resulted in duplicates being sent. This issue only became common enough
717 to notice due to the introduction of connection certificate information,
718 the item size being so much larger. Found and fixed by Wolfgang Breyha.
720 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
721 size buffer was used, resulting in syntax errors when an expansion
724 JH/07 Add support for directories of certificates when compiled with a GnuTLS
725 version 3.3.6 or later.
727 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
728 is EXPERIMENTAL_EVENT, the main-configuration and transport options
729 both become "event_action", the variables become $event_name, $event_data
730 and $event_defer_errno. There is a new variable $verify_mode, usable in
731 routers, transports and related events. The tls:cert event is now also
732 raised for inbound connections, if the main configuration event_action
735 TL/06 In test suite, disable OCSP for old versions of openssl which contained
736 early OCSP support, but no stapling (appears to be less than 1.0.0).
738 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
739 server certificate names available under the smtp transport option
740 "tls_verify_cert_hostname" now do not permit multi-component wildcard
743 JH/10 Time-related extraction expansions from certificates now use the main
744 option "timezone" setting for output formatting, and are consistent
745 between OpenSSL and GnuTLS compilations. Bug 1541.
747 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
748 encoded parameter in the incoming message. Bug 1558.
750 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
751 include certificate info, eximon was claiming there were spoolfile
754 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
756 JH/14 Log delivery-related information more consistently, using the sequence
757 "H=<name> [<ip>]" wherever possible.
759 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
760 are problematic for Debian distribution, omit them from the release
763 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
765 JH/16 Fix string representation of time values on 64bit time_t architectures.
768 JH/17 Fix a null-indirection in certextract expansions when a nondefault
769 output list separator was used.
774 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
775 checkers that were complaining about end of non-void function with no
778 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
779 This was a regression introduced in 4.83 by another bugfix.
781 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
783 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
784 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
790 TF/01 Correctly close the server side of TLS when forking for delivery.
792 When a message was received over SMTP with TLS, Exim failed to clear up
793 the incoming connection properly after forking off the child process to
794 deliver the message. In some situations the subsequent outgoing
795 delivery connection happened to have the same fd number as the incoming
796 connection previously had. Exim would try to use TLS and fail, logging
797 a "Bad file descriptor" error.
799 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
800 utilities have not been installed.
802 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
803 temporary space as the ACL may create new global variables.
805 TL/01 LDAP support uses per connection or global context settings, depending
806 upon the detected version of the libraries at build time.
808 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
809 to extract and use the src ip:port in logging and expansions as if it
810 were a direct connection from the outside internet. PPv2 support was
811 updated based on HAProxy spec change in May 2014.
813 JH/02 Add ${listextract {number}{list}{success}{fail}}.
815 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
816 Properly escape header and check for NULL return.
818 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
821 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
823 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
824 characters in header names, implemented as a verify condition.
825 Contributed by Michael Fischer v. Mollard.
827 TL/05 Rename SPF condition results err_perm and err_temp to standardized
828 results permerror and temperror. Previous values are deprecated but
829 still accepted. In a future release, err_perm and err_temp will be
830 completely removed, which will be a backward incompatibility if the
831 ACL tests for either of these two old results. Patch contributed by
832 user bes-internal on the mailing list.
834 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
836 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
837 selectors, in both main and reject logs.
839 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
842 JH/07 Add malware type "sock" for talking to simple daemon.
844 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
846 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
847 routers/transports under cutthrough routing.
849 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
850 numbers. Touch up "bool" conditional to keep the same definition.
852 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
854 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
856 JH/12 Expand items in router/transport headers_add or headers_remove lists
857 individually rather than the list as a whole. Bug 1452.
859 Required for reasonable handling of multiple headers_ options when
860 they may be empty; requires that headers_remove items with embedded
861 colons must have them doubled (or the list-separator changed).
863 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
864 view the policy declared in the DMARC record. Currently, $dmarc_status
865 is a combined value of both the record presence and the result of the
868 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
870 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
871 dnslookup router and the smtp transport (applying to the forward
874 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
875 of ldap servers used for a specific lookup. Patch provided by Heiko
878 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
879 New variable $lookup_dnssec_authenticated for observability.
881 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
882 Patch submitted by Lars Timman.
884 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
886 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
887 Requires trusted mode and valid format message id, aborts otherwise.
888 Patch contributed by Heiko Schlichting.
890 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
891 certextract with support for various fields. Bug 1358.
893 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
894 is requested by default, modifiable by smtp transport option
897 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
898 operate on certificate variables to give certificate fingerprints
899 Also new ${sha256:cert_variable}.
901 JH/23 The PRDR feature is moved from being Experimental into the mainline.
903 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
904 Christian Aistleitner.
906 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
908 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
909 file. Patch from Wolfgang Breyha.
911 JH/25 Expand the coverage of the delivery $host and $host_address to
912 client authenticators run in verify callout. Bug 1476.
914 JH/26 Port service names are now accepted for tls_on_connect_ports, to
915 align with daemon_smtp_ports. Bug 72.
917 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
918 support and error reporting did not work properly.
920 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
921 and is readable. Patch from Andrew Colin Kissa.
923 TL/14 Enhance documentation of ${run expansion and how it parses the
924 commandline after expansion, particularly in the case when an
925 unquoted variable expansion results in an empty value.
927 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
929 PP/02 Fix internal collision of T_APL on systems which support RFC3123
930 by renaming away from it. Addresses GH issue 15, reported by
933 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
935 TL/15 SECURITY: prevent double expansion in math comparison functions
936 (can expand unsanitized data). Not remotely exploitable.
943 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
945 PP/02 Make -n do something, by making it not do something.
946 When combined with -bP, the name of an option is not output.
948 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
951 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
952 $sender_host_name and config options to manage this, and basic check
955 PP/05 DSCP support for outbound connections and control modifier for inbound.
957 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
958 (Only plugin which currently uses this is kerberos4, which nobody should
959 be using, but we should make it available and other future plugins might
960 conceivably use it, even though it would break NAT; stuff *should* be
961 using channel bindings instead).
963 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
964 name; added for Sendmail compatibility; requires admin caller.
965 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
966 just ignore it); requires trusted caller.
967 Also parse but ignore: -Ac -Am -X<logfile>
970 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
972 TL/02 Add +smtp_confirmation as a default logging option.
974 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
975 Patch by Magnus Holmgren from 2007-02-20.
977 TL/04 Bugzilla 1281 - Spec typo.
978 Bugzilla 1283 - Spec typo.
979 Bugzilla 1290 - Spec grammar fixes.
981 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
983 TL/06 Add Experimental DMARC support using libopendmarc libraries.
985 TL/07 Fix an out of order global option causing a segfault. Reported to dev
986 mailing list by by Dmitry Isaikin.
988 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
990 JH/02 Support "G" suffix to numbers in ${if comparisons.
992 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
994 NM/01 Bugzilla 1197 - Spec typo
995 Bugzilla 1196 - Spec examples corrections
997 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
999 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1000 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1003 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1004 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1006 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1007 "acl {{name}{arg}...}", and optional args on acl condition
1010 JH/05 Permit multiple router/transport headers_add/remove lines.
1012 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1014 JH/07 Avoid using a waiting database for a single-message-only transport.
1015 Performance patch from Paul Fisher. Bugzilla 1262.
1017 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1020 JH/09 Add $headers_added variable, with content from use of ACL modifier
1021 add_header (but not yet added to the message). Bugzilla 199.
1023 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1024 Pulled from Bugzilla 817 by Wolfgang Breyha.
1026 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1028 (nb: this is the same fix as in Exim 4.80.1)
1030 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1033 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1036 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1038 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1039 advertise SMTP AUTH mechanism to us, instead of a generic
1040 protocol violation error. Also, make Exim more robust to bad
1041 data from the Dovecot auth socket.
1043 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1045 When a queue runner is handling a message, Exim first routes the
1046 recipient addresses, during which it prunes them based on the retry
1047 hints database. After that it attempts to deliver the message to
1048 any remaining recipients. It then updates the hints database using
1051 So if a recipient address works intermittently, it can get repeatedly
1052 deferred at routing time. The retry hints record remains fresh so the
1053 address never reaches the final cutoff time.
1055 This is a fairly common occurrence when a user is bumping up against
1056 their storage quota. Exim had some logic in its local delivery code
1057 to deal with this. However it did not apply to per-recipient defers
1058 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1060 This change adds a proper retry rule check during routing so that the
1061 final cutoff time is checked against the message's age. We only do
1062 this check if there is an address retry record and there is not a
1063 domain retry record; this implies that previous attempts to handle
1064 the address had the retry_use_local_parts option turned on. We use
1065 this as an approximation for the destination being like a local
1066 delivery, as in LMTP.
1068 I suspect this new check makes the old local delivery cutoff check
1069 redundant, but I have not verified this so I left the code in place.
1071 TF/02 Correct gecos expansion when From: is a prefix of the username.
1073 Test 0254 submits a message to Exim with the header
1077 When I ran the test suite under the user fanf2, Exim expanded
1078 the header to contain my full name, whereas it should have added
1079 a Resent-Sender: header. It erroneously treats any prefix of the
1080 username as equal to the username.
1082 This change corrects that bug.
1084 GF/01 DCC debug and logging tidyup
1085 Error conditions log to paniclog rather than rejectlog.
1086 Debug lines prefixed by "DCC: " to remove any ambiguity.
1088 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1090 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1091 Bug spotted by Jeremy Harris; was flawed since initial commit.
1092 Would have resulted in OCSP responses post-SNI triggering an Exim
1093 NULL dereference and crash.
1095 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1097 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1098 Bug detection, analysis and fix by Samuel Thibault.
1099 Bugzilla 1331, Debian bug #698092.
1101 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1103 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1104 Server implementation by Todd Lyons, client by JH.
1105 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1106 config variable "prdr_enable" controls whether the server
1107 advertises the facility. If the client requests PRDR a new
1108 acl_data_smtp_prdr ACL is called once for each recipient, after
1109 the body content is received and before the acl_smtp_data ACL.
1110 The client is controlled by both of: a hosts_try_prdr option
1111 on the smtp transport, and the server advertisement.
1112 Default client logging of deliveries and rejections involving
1113 PRDR are flagged with the string "PRDR".
1115 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1116 fclose(). Diagnosis by Todd Lyons.
1118 PP/17 Update configure.default to handle IPv6 localhost better.
1119 Patch by Alain Williams (plus minor tweaks).
1122 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1123 This is now consistent with GnuTLS, and is now documented: the
1124 previous undocumented portable approach to treating the option as
1125 unset was to force an expansion failure. That still works, and
1126 an empty string is now equivalent.
1128 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1129 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1130 not performing validation itself.
1132 PP/20 Added force_command boolean option to pipe transport.
1133 Patch from Nick Koston, of cPanel Inc.
1135 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1138 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1140 PP/21 Fix eximon continuous updating with timestamped log-files.
1141 Broken in a format-string cleanup in 4.80, missed when I repaired the
1142 other false fix of the same issue.
1143 Report and fix from Heiko Schlichting.
1146 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1147 Report from Prashanth Katuri.
1149 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1150 It's SecureTransport, so affects any MacOS clients which use the
1151 system-integrated TLS libraries, including email clients.
1153 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1154 using a MIME ACL for non-SMTP local injection.
1155 Report and assistance in diagnosis by Warren Baker.
1157 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1159 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1161 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1162 last id that failed so it may be referenced in subsequent ACL's.
1164 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1167 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1168 ldap library initialization, allowing self-signed CA's to be
1169 used. Also properly sets require_cert option later in code by
1170 using NULL (global ldap config) instead of ldap handle (per
1171 session). Bug diagnosis and testing by alxgomz.
1173 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1174 the src/util/ subdirectory.
1176 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1177 renamed to Transport Post Delivery Action by Jeremy Harris, as
1180 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1181 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1182 redis_servers = needs to be configured which will be used by the redis
1183 lookup. Patch from Warren Baker, of The Packet Hub.
1185 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1187 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1188 hostname or reverse DNS when processing a host list. Used suggestions
1189 from multiple comments on this bug.
1191 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1193 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1194 Missed a few lines, added it to make the runtest require no keyboard
1197 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1198 contains upper case chars. Make router use caseful_local_part.
1200 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1201 support when GnuTLS has been built with p11-kit.
1207 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1209 This, or similar/improved, will also be change PP/11 of 4.82.
1215 PP/01 Handle short writes when writing local log-files.
1216 In practice, only affects FreeBSD (8 onwards).
1217 Bugzilla 1053, with thanks to Dmitry Isaikin.
1219 NM/01 Bugzilla 949 - Documentation tweak
1221 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1224 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1226 PP/02 Implemented gsasl authenticator.
1228 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1230 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1231 `pkg-config foo` for cflags/libs.
1233 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1234 with rest of GSASL and with heimdal_gssapi.
1236 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1237 `pkg-config foo` for cflags/libs for the TLS implementation.
1239 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1240 properties get this fed in as external SSF. A number of robustness
1241 and debugging improvements to the cyrus_sasl authenticator.
1243 PP/08 cyrus_sasl server now expands the server_realm option.
1245 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1246 Patch by Jeremy Harris.
1248 PP/10 Added dbmjz lookup type.
1250 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1252 PP/12 MAIL args handles TAB as well as SP, for better interop with
1253 non-compliant senders.
1254 Analysis and variant patch by Todd Lyons.
1256 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1257 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1258 Patch from Dirk Mueller <dmueller@suse.com>
1260 PP/13 tls_peerdn now print-escaped for spool files.
1261 Observed some $tls_peerdn in wild which contained \n, which resulted
1262 in spool file corruption.
1264 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1265 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1266 or write after TLS renegotiation, which otherwise led to messages
1269 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1270 as a tracking header (ie: a signed header comes before the signature).
1271 Patch from Wolfgang Breyha.
1273 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1274 comma-sep list; embedded commas doubled.
1276 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1278 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1280 Report and patch from Dmitry Banschikov.
1282 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1283 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1284 failures. We appear to now support TLS1.1+ with Exim.
1286 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1287 lets Exim select keys and certificates based upon TLS SNI from client.
1288 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1289 before an outbound SMTP session. New log_selector, +tls_sni.
1291 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1292 NULL dereference. Report and patch from Alun Jones.
1294 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1295 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1296 Not seeing resolver debug output on NetBSD, but suspect this is a
1297 resolver implementation change.
1299 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1300 Left warnings. Added "eximon gdb" invocation mode.
1302 PP/21 Defaulting "accept_8bitmime" to true, not false.
1304 PP/22 Added -bw for inetd wait mode support.
1306 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1307 locate the relevant includes and libraries. Made this the default.
1309 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1310 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1312 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1313 This may cause build issues on older platforms.
1315 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1316 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1317 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1318 Added SNI support via GnuTLS too.
1319 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1321 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1323 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1324 Applied second patch from Janne, implementing suggestion to default
1325 multiple-strings-in-record handling to match SPF spec.
1327 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1329 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1330 read-only, out of scope).
1331 Patch from Wolfgang Breyha, report from Stuart Northfield.
1333 PP/29 Fix three issues highlighted by clang analyser static analysis.
1334 Only crash-plausible issue would require the Cambridge-specific
1335 iplookup router and a misconfiguration.
1336 Report from Marcin Mirosław.
1338 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1340 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1341 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1342 As part of this, removing so much warning spew let me fix some minor
1343 real issues in debug logging.
1345 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1346 assignment on my part. Fixed.
1348 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1349 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1350 Janne Snabb (who went above and beyond: thank you).
1352 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1353 string otherwise requires a connection and a bunch more work and it's
1354 relatively easy to get wrong. Should also expose TLS library linkage
1357 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1358 64-bit ${eval} (JH/03).
1360 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1361 GNU libc to support some of the 64-bit stuff, should not lead to
1362 conflicts. Defined before os.h is pulled in, so if a given platform
1363 needs to override this, it can.
1365 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1366 protection layer was required, which is not implemented.
1367 Bugzilla 1254, patch from Wolfgang Breyha.
1369 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1370 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1371 tls_dhparam take prime identifiers. Also unbreak combination of
1372 OpenSSL+DH_params+TLSSNI.
1374 PP/39 Disable SSLv2 by default in OpenSSL support.
1380 PP/01 Solaris build fix for Oracle's LDAP libraries.
1381 Bugzilla 1109, patch from Stephen Usher.
1383 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1385 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1388 TF/02 Fix a couple more cases where we did not log the error message
1389 when unlink() failed. See also change 4.74-TF/03.
1391 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1392 lock up or crash if it happened to be inside a call to libc when it
1393 got a SIGUSR1 from exiwhat.
1395 The SIGUSR1 handler appends the current process status to the process
1396 log which is later printed by exiwhat. It used to use the general
1397 purpose logging code to do this, but several functions it calls are
1398 not safe for signals.
1400 The new output code in the SIGUSR1 handler is specific to the process
1401 log, and simple enough that it's easy to inspect for signal safety.
1402 Removing some special cases also simplifies the general logging code.
1403 Removing the spurious timestamps from the process log simplifies
1406 TF/04 Improved ratelimit ACL condition.
1408 The /noupdate option has been deprecated in favour of /readonly which
1409 has clearer semantics. The /leaky, /strict, and /readonly update modes
1410 are mutually exclusive. The update mode is no longer included in the
1411 database key; it just determines when the database is updated. (This
1412 means that when you upgrade Exim will forget old rate measurements.)
1414 Exim now checks that the per_* options are used with an update mode that
1415 makes sense for the current ACL. For example, when Exim is processing a
1416 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1417 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1418 must specify per_mail/readonly. If you omit the update mode it defaults to
1419 /leaky where that makes sense (as before) or /readonly where required.
1421 The /noupdate option is now undocumented but still supported for
1422 backwards compatibility. It is equivalent to /readonly except that in
1423 ACLs where /readonly is required you may specify /leaky/noupdate or
1424 /strict/noupdate which are treated the same as /readonly.
1426 A useful new feature is the /count= option. This is a generalization
1427 of the per_byte option, so that you can measure the throughput of other
1428 aggregate values. For example, the per_byte option is now equivalent
1429 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1431 The per_rcpt option has been generalized using the /count= mechanism
1432 (though it's more complicated than the per_byte equivalence). When it is
1433 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1434 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1435 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1436 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1437 Note that using per_rcpt with a non-readonly update mode in more than
1438 one ACL will cause the recipients to be double-counted. (The per_mail
1439 and per_byte options don't have this problem.)
1441 The handling of very low rates has changed slightly. If the computed rate
1442 is less than the event's count (usually one) then this event is the first
1443 after a long gap. In this case the rate is set to the same as this event's
1444 count, so that the first message of a spam run is counted properly.
1446 The major new feature is a mechanism for counting the rate of unique
1447 events. The new per_addr option counts the number of different
1448 recipients that someone has sent messages to in the last time period. It
1449 behaves like per_rcpt if all the recipient addresses are different, but
1450 duplicate recipient addresses do not increase the measured rate. Like
1451 the /count= option this is a general mechanism, so the per_addr option
1452 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1453 example, measure the rate that a client uses different sender addresses
1454 with the options per_mail/unique=$sender_address. There are further
1455 details in the main documentation.
1457 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1459 TF/06 Removed a few PCRE remnants.
1461 TF/07 Automatically extract Exim's version number from tags in the git
1462 repository when doing development or release builds.
1464 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1465 Bugzilla 879. Patch from Paul Fisher.
1467 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1468 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1471 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1473 PP/05 Variable $av_failed, true if the AV scanner deferred.
1474 Bugzilla 1078. Patch from John Horne.
1476 PP/06 Stop make process more reliably on build failure.
1477 Bugzilla 1087. Patch from Heiko Schlittermann.
1479 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1480 Bugzilla 1089. Patch from Heiko Schlittermann.
1482 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1483 Bugzilla 1131. Patch from Holger Weiß.
1485 PP/09 Handle IPv6 addresses with SPF.
1486 Bugzilla 860. Patch from Wolfgang Breyha.
1488 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1490 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1493 PP/11 match_* no longer expand right-hand-side by default.
1494 New compile-time build option, EXPAND_LISTMATCH_RHS.
1495 New expansion conditions, "inlist", "inlisti".
1497 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1499 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1501 PP/14 fix log_write() format string regression from TF/03.
1502 Bugzilla 1152. Patch from Dmitry Isaikin.
1508 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1510 PP/02 Harmonised TLS library version reporting; only show if debugging.
1511 Layout now matches that introduced for other libraries in 4.74 PP/03.
1513 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1515 PP/04 New "dns_use_edns0" global option.
1517 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1520 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1521 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1523 TK/01 Updated PolarSSL code to 0.14.2.
1524 Bugzilla 1097. Patch from Andreas Metzler.
1526 PP/07 Catch divide-by-zero in ${eval:...}.
1527 Fixes bugzilla 1102.
1529 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1532 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1533 format-string attack -- SECURITY: remote arbitrary code execution.
1535 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1536 time unintentionally subject to list matching rules, letting the header
1537 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1538 arbitrary string expansion). This allowed for information disclosure.
1540 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1541 INT_MIN/-1 -- value coerced to INT_MAX.
1547 NM/01 Workaround for PCRE version dependency in version reporting
1550 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1551 This fixes portability to compilers other than gcc, notably
1552 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1554 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1555 makefiles for portability to HP-UX and POSIX correctness.
1557 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1558 Also via indented variable definition in the Makefile.
1559 (Debugging by Oliver Heesakkers).
1561 PP/02 Restore caching of spamd results with expanded spamd_address.
1562 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1564 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1565 Improves build reliability. Fix from: Frank Elsner
1567 NM/02 Fix wide character breakage in the rfc2047 coding
1568 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1570 NM/03 Allow underscore in dnslist lookups
1571 Fixes bug 1026. Patch from Graeme Fowler
1573 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1574 Code patches from Adam Ciarcinski of NetBSD.
1576 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1579 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1580 is logged, to avoid truncation. Patch from John Horne.
1582 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1583 Patch from Jakob Hirsch.
1585 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1586 SQL string expansion failure details.
1587 Patch from Andrey Oktyabrski.
1589 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1590 Patch from Simon Arlott.
1592 PP/09 New lookups functionality failed to compile on old gcc which rejects
1593 extern declarations in function scope.
1594 Patch from Oliver Fleischmann
1596 PP/10 Use sig_atomic_t for flags set from signal handlers.
1597 Check getgroups() return and improve debugging.
1598 Fixed developed for diagnosis in bug 927 (which turned out to be
1601 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1602 Patch from Mark Zealey.
1604 PP/12 Bugzilla 1056: Improved spamd server selection.
1605 Patch from Mark Zealey.
1607 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1608 Based on patch from Heiko Schlittermann.
1610 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1611 Patch from Uwe Doering, sign-off by Michael Haardt.
1613 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1614 variable declaration deep within a block. Bug and patch from
1617 PP/15 lookups-Makefile IRIX compatibility coercion.
1619 PP/16 Make DISABLE_DKIM build knob functional.
1621 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1622 Patch by Simon Arlott
1624 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1625 variable argument macros. Our copy now differs from upstream.
1631 TF/01 Failure to get a lock on a hints database can have serious
1632 consequences so log it to the panic log.
1634 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1635 controlled using the smtp_confirmation log selector.
1637 TF/03 Include the error message when we fail to unlink a spool file.
1639 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1640 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1641 for maintaining out-of-tree patches for some time.
1643 PP/01 Bugzilla 139: Documentation and portability issues.
1644 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1645 Handle per-OS dynamic-module compilation flags.
1647 PP/02 Let /dev/null have normal permissions.
1648 The 4.73 fixes were a little too stringent and complained about the
1649 permissions on /dev/null. Exempt it from some checks.
1650 Reported by Andreas M. Kirchwitz.
1652 PP/03 Report version information for many libraries, including
1653 Exim version information for dynamically loaded libraries. Created
1654 version.h, now support a version extension string for distributors
1655 who patch heavily. Dynamic module ABI change.
1657 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1658 privilege escalation vulnerability whereby the Exim run-time user
1659 can cause root to append content of the attacker's choosing to
1662 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1665 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1666 If dropping privileges for untrusted macros, we disabled normal logging
1667 on the basis that it would fail; for the Exim run-time user, this is not
1668 the case, and it resulted in successful deliveries going unlogged.
1669 Fixed. Reported by Andreas Metzler.
1675 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1676 only prepend for the Resent-* case. Fixes regression introduced in
1677 Exim 4.70 by NM/22 for Bugzilla 607.
1679 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1680 increasing numbers of administrators be bitten by this.
1682 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1684 PP/03 Bugzilla 994: added openssl_options main configuration option.
1686 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1688 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1690 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1692 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1693 without a peer certificate, leading to a segfault because of an
1694 assumption that peers always have certificates. Be a little more
1695 paranoid. Problem reported by Martin Tscholak.
1697 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1698 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1699 NB: ClamAV planning to remove STREAM in "middle of 2010".
1700 CL also introduces -bmalware, various -d+acl logging additions and
1701 more caution in buffer sizes.
1703 PP/09 Implemented reverse_ip expansion operator.
1705 PP/10 Bugzilla 937: provide a "debug" ACL control.
1707 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1709 PP/12 Bugzilla 973: Implement --version.
1711 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1713 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1715 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1717 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1718 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1719 ignore trailing whitespace.
1721 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1723 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1726 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1727 Notification from Dr Andrew Aitchison.
1729 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1730 ExtendedDetectionInfo response format.
1731 Notification from John Horne.
1733 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1736 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1737 XSL and documented dependency on system catalogs, with examples of how
1740 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1743 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1744 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1745 configuration file which is writeable by the Exim user or group.
1747 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1748 of configuration files to cover files specified with the -C option if
1749 they are going to be used with root privileges, not just the default
1752 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1753 option (effectively making it always true).
1755 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1756 files to be used while preserving root privileges.
1758 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1759 that rogue child processes cannot use them.
1761 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1762 run-time user, instead of root.
1764 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1765 Exim run-time user without dropping privileges.
1767 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1768 result string, instead of calling string_vformat() twice with the same
1771 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1772 for other users. Others should always drop root privileges if they use
1773 -C on the command line, even for a whitelisted configure file.
1775 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1777 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1783 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1784 $data_path, and $header_path variables; fixed documentation bugs and
1787 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1788 exipick to access non-standard spools, including the "frozen" queue
1791 NM/01 Bugzilla 965: Support mysql stored procedures.
1792 Patch from Alain Williams
1794 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1796 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1797 Patch from Andreas Metzler
1799 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1800 Patch from Kirill Miazine
1802 NM/05 Bugzilla 671: Added umask to procmail example.
1804 JJ/03 installed exipick 20100323.0, fixing doc bug
1806 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1807 directory. Notification and patch from Dan Rosenberg.
1809 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1811 TK/02 Improve log output when DKIM signing operation fails.
1813 MH/01 Treat the transport option dkim_domain as a colon separated
1814 list, not as a single string, and sign the message with each element,
1815 omitting multiple occurences of the same signer.
1817 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1818 Bugzilla 985, 986. Patch by Simon Arlott
1820 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1821 Patch by Simon Arlott
1823 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1824 MBX locking. Notification from Dan Rosenberg.
1830 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1832 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1834 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1836 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1838 NM/04 Bugzilla 915: Flag broken perl installation during build.
1844 TK/01 Added patch by Johannes Berg that expands the main option
1845 "spamd_address" if it starts with a dollar sign.
1847 TK/02 Write list of recipients to X-Envelope-Sender header when building
1848 the mbox-format spool file for content scanning (suggested by Jakob
1851 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1852 (http://www.dcc-servers.net/) support via dccifd. Activated by
1853 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1855 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1856 by Mark Daniel Reidel <mr@df.eu>.
1858 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1859 When building exim an external PCRE library is now needed -
1860 PCRE is a system library on the majority of modern systems.
1861 See entry on PCRE_LIBS in EDITME file.
1863 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1864 conversation. Added nologin parameter to request.
1865 Patch contributed by Kirill Miazine.
1867 TF/01 Do not log submission mode rewrites if they do not change the address.
1869 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1871 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1872 log files in place. Contributed by Roberto Lima.
1874 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1876 TF/03 Bugzilla 615: When checking the local_parts router precondition
1877 after a local_part_suffix or local_part_prefix option, Exim now
1878 does not use the address's named list lookup cache, since this
1879 contains cached lookups for the whole local part.
1881 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1882 Robert Millan. Documentation is in experimental-spec.txt.
1884 TF/04 Bugzilla 668: Fix parallel build (make -j).
1886 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1888 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1889 Patch provided by Jan Srzednicki.
1891 TF/05 Leading white space used to be stripped from $spam_report which
1892 wrecked the formatting. Now it is preserved.
1894 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1895 that they are available at delivery time.
1897 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1899 TF/08 TLS error reporting now respects the incoming_interface and
1900 incoming_port log selectors.
1902 TF/09 Produce a more useful error message if an SMTP transport's hosts
1903 setting expands to an empty string.
1905 NM/06 Bugzilla 744: EXPN did not work under TLS.
1906 Patch provided by Phil Pennock.
1908 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1909 Patch provided by Richard Godbee.
1911 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1912 acl_smtp_notquit, added index entry.
1914 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1915 Patch provided by Eugene Bujak.
1917 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1918 accept(). Patch provided by Maxim Dounin.
1920 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1921 Patch provided by Phil Pennock.
1923 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1925 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1926 Patch provided by Brad "anomie" Jorsch.
1928 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1929 Patch provided by Dean Brooks.
1931 TK/05 Add native DKIM support (does not depend on external libraries).
1933 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1934 Patch provided by Graeme Fowler.
1936 NM/16 Bugzilla 851: Documentation example syntax fix.
1938 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1940 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1943 NM/19 Bugzilla 745: TLS version reporting.
1944 Patch provided by Phil Pennock.
1946 NM/20 Bugzilla 167: bool: condition support.
1947 Patch provided by Phil Pennock.
1949 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1950 clients. Patch provided by Phil Pennock.
1952 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1953 Patch provided by Brad "anomie" Jorsch.
1955 NM/23 Bugzilla 687: Fix misparses in eximstats.
1956 Patch provided by Heiko Schlittermann.
1958 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1959 Patch provided by Heiko Schlittermann.
1961 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1962 plus update to original patch.
1964 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1966 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1967 Patch provided by David Brownlee.
1969 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1971 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1973 NM/30 Bugzilla 888: TLS documentation bugfixes.
1975 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1977 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1978 Unlike the original bugzilla I have changed all shell scripts in src tree.
1980 NM/33 Bugzilla 898: Transport filter timeout fix.
1981 Patch by Todd Rinaldo.
1983 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1984 Patch by Serge Demonchaux.
1986 NM/35 Bugzilla 39: Base64 decode bug fixes.
1987 Patch by Jakob Hirsch.
1989 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1991 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1993 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1995 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2001 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2002 ALT-N's libdkim that I have put here:
2003 http://duncanthrax.net/exim-experimental/
2005 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2006 were called 'true' and it seems that C99 defines that as a reserved
2007 keyword to be used with 'bool' variable types. That means you could
2008 not include C99-style headers which use bools without triggering
2009 build errors in sieve.c.
2011 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2012 as mailq or other aliases. Changed the --help handling significantly
2013 to do whats expected. exim_usage() emits usage/help information.
2015 SC/01 Added the -bylocaldomain option to eximstats.
2017 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2019 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2021 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2027 PH/01 Another patch from the Sieve maintainer.
2029 PH/02 When an IPv6 address is converted to a string for single-key lookup
2030 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2031 dots are used instead of colons so that keys in lsearch files need not
2032 contain colons. This was done some time before quoting was made available
2033 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2034 (notated using the quote facility) so as to distinguish them from IPv4
2035 keys. This meant that lookups for IP addresses in host lists did not work
2036 for iplsearch lookups.
2038 This has been fixed by arranging for IPv6 addresses to be expressed with
2039 colons if the lookup type is iplsearch. This is not incompatible, because
2040 previously such lookups could never work.
2042 The situation is now rather anomalous, since one *can* have colons in
2043 ordinary lsearch keys. However, making the change in all cases is
2044 incompatible and would probably break a number of configurations.
2046 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2049 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2050 conversion specification without a maximum field width, thereby enabling
2051 a rogue spamd server to cause a buffer overflow. While nobody in their
2052 right mind would setup Exim to query an untrusted spamd server, an
2053 attacker that gains access to a server running spamd could potentially
2054 exploit this vulnerability to run arbitrary code as the Exim user.
2056 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2057 $primary_hostname instead of what libspf2 thinks the hosts name is.
2059 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2060 a directory entry by the name of the lookup key. Previously, if a
2061 symlink pointed to a non-existing file or a file in a directory that
2062 Exim lacked permissions to read, a lookup for a key matching that
2063 symlink would fail. Now it is enough that a matching directory entry
2064 exists, symlink or not. (Bugzilla 503.)
2066 PH/03 The body_linecount and body_zerocount variables are now exported in the
2069 PH/04 Added the $dnslist_matched variable.
2071 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2072 This means they are set thereafter only if the connection becomes
2075 PH/06 Added the client_condition to authenticators so that some can be skipped
2076 by clients under certain conditions.
2078 PH/07 The error message for a badly-placed control=no_multiline_responses left
2079 "_responses" off the end of the name.
2081 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2083 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2084 (without spaces) instead of just copying the configuration text.
2086 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2088 PH/11 Added $max_received_linelength.
2090 PH/12 Added +ignore_defer and +include_defer to host lists.
2092 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2093 way in which PCRE > 7.0 is built.
2095 PH/14 Implemented queue_only_load_latch.
2097 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2098 MAIL command. The effect was to mangle the value on 64-bit systems.
2100 PH/16 Another patch from the Sieve maintainer.
2102 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2104 PH/18 If a system quota error occurred while trying to create the file for
2105 a maildir delivery, the message "Mailbox is full" was not appended to the
2106 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2107 only to a quota excession during the actual writing of the file.
2109 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2110 characters?) which causes problems in log lines. The DN values are now
2111 passed through string_printing() before being added to log lines.
2113 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2114 and InterBase are left for another time.)
2116 PH/21 Added message_body_newlines option.
2118 PH/22 Guard against possible overflow in moan_check_errorcopy().
2120 PH/23 POSIX allows open() to be a macro; guard against that.
2122 PH/24 If the recipient of an error message contained an @ in the local part
2123 (suitably quoted, of course), incorrect values were put in $domain and
2124 $local_part during the evaluation of errors_copy.
2130 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2131 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2134 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2135 issue a MAIL command.
2137 PH/02 In an ACL statement such as
2139 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2141 if a client was not listed at all, or was listed with a value other than
2142 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2143 the condition was not true (as it should be), so access was not denied.
2144 The bug was that the ! inversion was incorrectly passed on to the second
2145 item. This has been fixed.
2147 PH/03 Added additional dnslists conditions == and =& which are different from
2148 = and & when the dns lookup returns more than one IP address.
2150 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2151 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2153 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2154 FSYNC, which compiles an option called disable_fsync that allows for
2155 bypassing fsync(). The documentation is heavily laced with warnings.
2157 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2159 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2160 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2161 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2162 including adding "make clean"; (3) Added -fPIC when compiling the test
2163 dynamically loaded module, to get rid of a warning.
2165 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2166 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2167 = 0s. The bug is otherwise harmless.
2169 PH/07 There was a bug in the dovecot authenticator such that the value of
2170 $auth1 could be overwritten, and so not correctly preserved, after a
2171 successful authentication. This usually meant that the value preserved by
2172 the server_setid option was incorrect.
2174 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2176 PH/09 Installed PCRE release 7.0.
2178 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2179 run for batched SMTP input. It is now run at the start of every message
2180 in the batch. While fixing this I discovered that the process information
2181 (output by running exiwhat) was not always getting set for -bs and -bS
2182 input. This is fixed, and it now also says "batched" for BSMTP.
2184 PH/11 Added control=no_pipelining.
2186 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2187 patch, slightly modified), and move the expansion of helo_data till after
2188 the connection is made in the smtp transport (so it can use these
2191 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2193 PH/14 Added log_selector = +pid.
2195 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2197 PH/16 Add ${if forany and ${if forall.
2199 PH/17 Added dsn_from option to vary the From: line in DSNs.
2201 PH/18 Flush SMTP output before performing a callout, unless control =
2202 no_callout_flush is set.
2204 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2205 was true (the default) a successful delivery failed to delete the retry
2206 item, thus causing premature timeout of the address. The bug is now
2209 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2211 PH/21 Long custom messages for fakedefer and fakereject are now split up
2212 into multiline responses in the same way that messages for "deny" and
2213 other ACL rejections are.
2215 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2216 with slight modification.
2218 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2219 draft, changing the syntax and factoring some duplicate code".
2221 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2222 for deliveries of the second and subsequent messages over the same SMTP
2225 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2226 ${reduce, with only minor "tidies".
2228 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2230 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2231 expansion side effects.
2233 PH/27 When a message times out after an over-quota error from an Exim-imposed
2234 quota, the bounce message says "mailbox is full". This message was not
2235 being given when it was a system quota that was exceeded. It now should
2238 MH/03 Made $recipients available in local_scan(). local_scan() already has
2239 better access to the recipient list through recipients_list[], but
2240 $recipients can be useful in postmaster-provided expansion strings.
2242 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2243 in the case of a MAIL command with additional options following the
2244 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2245 were accidentally chopped off.
2247 PH/29 SMTP synchronization checks are implemented when a command is read -
2248 there is a check that no more input is waiting when there shouldn't be
2249 any. However, for some commands, a delay in an ACL can mean that it is
2250 some time before the response is written. In this time, more input might
2251 arrive, invalidly. So now there are extra checks after an ACL has run for
2252 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2253 pipelining has not been advertised.
2255 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2257 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2258 returned in the SMTP response when smtp_return_error_details was false.
2259 This has been fixed.
2261 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2262 instead of the C I/O that was originally supplied, because problems were
2263 reported on Solaris.
2265 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2266 Exim which did not show up earlier: it was assuming that a call to
2267 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2268 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2269 was a macro that became an assignment, so it seemed to work. This has
2270 changed to a proper function call with a void return, hence the compile
2271 error. Exim's code has been fixed.
2273 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2276 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2278 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2280 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2281 because it turns out to be too big or there is a timeout) there is no
2282 "Completed" line in the log. When some messages of this type were
2283 selected by exigrep, they were listed as "not completed". Others were
2284 picked up by some special patterns. I have improved the selection
2285 criteria to be more general.
2287 PH/38 The host_find_failed option in the manualroute router can now be set
2288 to "ignore", to completely ignore a host whose IP address cannot be
2289 found. If all hosts are ignored, the behaviour is controlled by the new
2290 host_all_ignored option.
2292 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2293 homing or because of multiple MX records with /mx) generated more than
2294 one IP address, and the following item turned out to be the local host,
2295 all the secondary addresses of the first item were incorrectly removed
2296 from the list, along with the local host and any following hosts (which
2297 is what is supposed to happen).
2299 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2300 whoever called Exim into the -H file. In the case of the daemon it was
2301 behaving confusingly. When first started, it used values for whoever
2302 started the daemon, but after a SIGHUP it used the Exim user (because it
2303 calls itself on a restart). I have changed the code so that it now always
2306 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2307 message are rejected with the same error (e.g. no authentication or bad
2308 sender address), and a DATA command is nevertheless sent (as can happen
2309 with PIPELINING or a stupid MUA), the error message that was given to the
2310 RCPT commands is included in the rejection of the DATA command. This is
2311 intended to be helpful for MUAs that show only the final error to their
2314 PH/42 Another patch from the Sieve maintainer.
2316 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2317 Eximstats - Fixed some broken HTML links and added missing column headers
2319 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2320 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2322 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2324 PH/43 Yet another patch from the Sieve maintainer.
2326 PH/44 I found a way to check for a TCP/IP connection going away before sending
2327 the response to the final '.' that terminates a message, but only in the
2328 case where the client has not sent further data following the '.'
2329 (unfortunately, this is allowed). However, in many cases there won't be
2330 any further data because there won't be any more messages to send. A call
2331 to select() can be used: if it shows that the input is "ready", there is
2332 either input waiting, or the socket has been closed. An attempt to read
2333 the next input character can distinguish the two cases. Previously, Exim
2334 would have sent an OK response which the client would never have see.
2335 This could lead to message repetition. This fix should cure that, at
2336 least in a lot of common cases.
2338 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2339 advertised in response to EHLO.
2345 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2346 fixed by 4.65/MH/01 (is this a record?) are fixed:
2348 (i) An empty string was always treated as zero by the numeric comparison
2349 operators. This behaviour has been restored.
2351 (ii) It is documented that the numeric comparison operators always treat
2352 their arguments as decimal numbers. This was broken in that numbers
2353 starting with 0 were being interpreted as octal.
2355 While fixing these problems I realized that there was another issue that
2356 hadn't been noticed. Values of message_size_limit (both the global option
2357 and the transport option) were treated as octal if they started with 0.
2358 The documentation was vague. These values are now always treated as
2359 decimal, and I will make that clear in the documentation.
2365 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2366 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2369 MH/01 Don't check that the operands of numeric comparison operators are
2370 integers when their expansion is in "skipping" mode (fixes bug
2371 introduced by 4.64-PH/07).
2373 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2374 child addresses, Exim now panics and dies. Previously, because the count
2375 is held in a short int, deliveries were likely to be lost. As such a
2376 large number of recipients for a single message is ridiculous
2377 (performance will be very, very poor), I have chosen to impose a limit
2378 rather than extend the field.
2384 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2385 leftover -K file (the existence of which was triggered by #402).
2386 While we were at it, introduced process PID as part of the -K
2387 filename. This should rule out race conditions when creating
2390 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2391 processing considerably. Previous code took too long for large mails,
2392 triggering a timeout which in turn triggers #401.
2394 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2395 in the DK code in transports.c. sendfile() is not really portable,
2396 hence the _LINUX specificness.
2398 TF/01 In the add_headers option to the mail command in an Exim filter,
2399 there was a bug that Exim would claim a syntax error in any
2400 header after the first one which had an odd number of characters
2403 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2404 callout verification, Exim cached a "reject" for the entire domain. This
2405 is correct for most verifications, but it is not correct for a recipient
2406 verification with use_sender or use_postmaster set, because in that case
2407 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2408 case of MAIL FROM:<> rejection from other early rejections (e.g.
2409 rejection of HELO). When verifying a recipient using a non-null MAIL
2410 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2411 Whatever the result of the callout, the value of the domain cache is
2412 left unchanged (for any other kind of callout, getting as far as trying
2413 RCPT means that the domain itself is ok).
2415 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2418 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2419 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2420 session, the daemon ignores ECONNECT errors and logs others; it now
2421 ignores EPIPE as well.
2423 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2424 (quoted-printable decoding).
2426 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2427 later the small subsequent patch to fix an introduced bug.
2429 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2431 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2433 PH/08 An error is now given if message_size_limit is specified negative.
2435 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2436 to be given (somewhat) arbitrary names.
2438 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2441 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2442 miscellaneous code fixes
2444 PH/10 Added the log_reject_target ACL modifier to specify where to log
2447 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2448 hostname. This is wrong, because it relates to the incoming message (and
2449 probably the interface on which it is arriving) and not to the outgoing
2450 callout (which could be using a different interface). This has been
2451 changed to use the value of the helo_data option from the smtp transport
2452 instead - this is what is used when a message is actually being sent. If
2453 there is no remote transport (possible with a router that sets up host
2454 addresses), $smtp_active_hostname is used.
2456 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2457 tweaks were necessary in order to get it to work (see also 21 below):
2458 (a) The code assumed that strncpy() returns a negative number on buffer
2459 overflow, which isn't the case. Replaced with Exim's string_format()
2461 (b) There were several signed/unsigned issues. I just did the minimum
2462 hacking in of casts. There is scope for a larger refactoring.
2463 (c) The code used strcasecmp() which is not a standard C function.
2464 Replaced with Exim's strcmpic() function.
2465 (d) The code set only $1; it now sets $auth1 as well.
2466 (e) A simple test gave the error "authentication client didn't specify
2467 service in request". It would seem that Dovecot has changed its
2468 interface. Fortunately there's a specification; I followed it and
2469 changed what the client sends and it appears to be working now.
2471 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2474 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2475 address A is aliased to B and C, where B exists and C does not. Without
2476 -v the output is "A verified" because verification stops after a
2477 successful redirection if more than one address is generated. However,
2478 with -v the child addresses are also verified. Exim was outputting "A
2479 failed to verify" and then showing the successful verification for C,
2480 with its parentage. It now outputs "B failed to verify", showing B's
2481 parentage before showing the successful verification of C.
2483 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2484 look up a TXT record in a specific list after matching in a combined
2487 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2488 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2489 they consult the DNS. I had assumed they would set it the way they
2490 wanted; and indeed my experiments on Linux seem to show that in some
2491 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2492 To be on the safe side, however, I have now made the interface to
2493 host_find_byname() similar to host_find_bydns(), with an argument
2494 containing the DNS resolver options. The host_find_byname() function now
2495 sets these options at its start, just as host_find_bydns() does. The smtp
2496 transport options dns_qualify_single and dns_search_parents are passed to
2497 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2498 of host_find_byname() use the default settings of RES_DEFNAMES
2499 (qualify_single) but not RES_DNSRCH (search_parents).
2501 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2502 spool_read_header() do less string testing, by means of a preliminary
2503 switch on the second character of optional "-foo" lines. (This is
2504 overdue, caused by the large number of possibilities that now exist.
2505 Originally there were few.) While I was there, I also converted the
2506 str(n)cmp tests so they don't re-test the leading "-" and the first
2507 character, in the hope this might squeeze out yet more improvement.
2509 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2510 flag allowing group syntax was set by the header_syntax check but not
2511 turned off, possible causing trouble later; (2) The flag was not being
2512 set at all for the header_verify test, causing "group"-style headers to
2513 be rejected. I have now set it in this case, and also caused header_
2514 verify to ignore an empty address taken from a group. While doing this, I
2515 came across some other cases where the code for allowing group syntax
2516 while scanning a header line wasn't quite right (mostly, not resetting
2517 the flag correctly in the right place). These bugs could have caused
2518 trouble for malformed header lines. I hope it is now all correct.
2520 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2521 with the "reply" argument non-NULL. The code, however (which originally
2522 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2523 but it didn't always do it. This confused somebody who was copying the
2524 code for some other use. I have removed all the tests.
2526 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2527 feature that was used to support insecure browsers during the U.S. crypto
2528 embargo. It requires special client support, and Exim is probably the
2529 only MTA that supported it -- and would never use it because real RSA is
2530 always available. This code has been removed, because it had the bad
2531 effect of slowing Exim down by computing (never used) parameters for the
2532 RSA_EXPORT functionality.
2534 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2535 authenticator to fail if there's a tab character in the incoming data
2536 (there should never be unless someone is messing about, as it's supposed
2537 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2538 if the connection is using TLS or if the remote IP is the same as the
2539 local IP, and the "valid-client-cert option" if a client certificate has
2542 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2543 authenticators. This can be used for authorization after authentication
2544 succeeds. (In the case of plaintext, it servers for both authentication
2547 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2548 if any retry times were supplied.
2550 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2551 connection, where there is no client IP address to check. In this
2552 situation, the verify now always succeeds.
2554 PH/25 Applied John Jetmore's -Mset patch.
2556 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2558 PH/27 In a string expansion for a processed (not raw) header when multiple
2559 headers of the same name were present, leading whitespace was being
2560 removed from all of them, but trailing whitespace was being removed only
2561 from the last one. Now trailing whitespace is removed from each header
2562 before concatenation. Completely empty headers in a concatenation (as
2563 before) are ignored.
2565 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2566 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2568 PH/29 [Removed. This was a change that I later backed out, and forgot to
2569 correct the ChangeLog entry (that I had efficiently created) before
2570 committing the later change.]
2572 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2573 address errors (4xx response to RCPT) over the same connection as other
2574 messages routed to the same hosts. Such deliveries are always "forced",
2575 so retry times are not inspected. This resulted in far too many retries
2576 for the affected addresses. The effect occurred only when there were more
2577 hosts than the hosts_max_try setting in the smtp transport when it had
2578 the 4xx errors. Those hosts that it had tried were not added to the list
2579 of hosts for which the message was waiting, so if all were tried, there
2580 was no problem. Two fixes have been applied:
2582 (i) If there are any address or message errors in an SMTP delivery, none
2583 of the hosts (tried or untried) are now added to the list of hosts
2584 for which the message is waiting, so the message should not be a
2585 candidate for sending over the same connection that was used for a
2586 successful delivery of some other message. This seems entirely
2587 reasonable: after all the message is NOT "waiting for some host".
2588 This is so "obvious" that I'm not sure why it wasn't done
2589 previously. Hope I haven't missed anything, but it can't do any
2590 harm, as the worst effect is to miss an optimization.
2592 (ii) If, despite (i), such a delivery is accidentally attempted, the
2593 routing retry time is respected, so at least it doesn't keep
2594 hammering the server.
2596 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2597 in ${readsocket because some servers need this prod.
2599 PH/32 Added some extra debug output when updating a wait-xxx database.
2601 PH/33 The hint "could be header name not terminated by colon", which has been
2602 given for certain expansion errors for a long time, was not being given
2603 for the ${if def:h_colon_omitted{... case.
2605 PH/34 The spec says: "With one important exception, whenever a domain list is
2606 being scanned, $domain contains the subject domain." There was at least
2607 one case where this was not true.
2609 PH/35 The error "getsockname() failed: connection reset by peer" was being
2610 written to the panic log as well as the main log, but it isn't really
2611 panic-worthy as it just means the connection died rather early on. I have
2612 removed the panic log writing for the ECONNRESET error when getsockname()
2615 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2616 runs only) independently of the message's sender address. This meant
2617 that, if the 4xx error was in fact related to the sender, a different
2618 message to the same recipient with a different sender could confuse
2619 things. In particular, this can happen when sending to a greylisting
2620 server, but other circumstances could also provoke similar problems.
2621 I have changed the default so that the retry time for these errors is now
2622 based a combination of the sender and recipient addresses. This change
2623 can be overridden by setting address_retry_include_sender=false in the
2626 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2627 remote server are returned as part of bounce messages. This was not
2628 happening for LMTP over a pipe (the lmtp transport), but now it is the
2629 same for both kinds of LMTP.
2631 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2632 in header lines that were in fact CNAMEs. This is no longer the case.
2634 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2635 and queue runs started by the daemon processed all messages. This has
2636 been fixed so that -R and -S can now usefully be given with -q<time>.
2638 PH/40 Import PCRE release 6.7 (fixes some bugs).
2640 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2642 PH/42 Give an error if -q is specified more than once.
2644 PH/43 Renamed the variables $interface_address and $interface_port as
2645 $received_ip_address and $received_port, to make it clear that these
2646 values apply to message reception, and not to the outgoing interface when
2647 a message is delivered. (The old names remain recognized, of course.)
2649 PH/44 There was no timeout on the connect() call when using a Unix domain
2650 socket in the ${readsocket expansion. There now is.
2652 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2653 be meaningful with "accept".
2655 SC/01 Eximstats V1.43
2656 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2658 SC/02 Eximstats V1.44
2659 Use a glob alias rather than an array ref in the generated
2660 parser. This improves both readability and performance.
2662 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2663 Collect SpamAssassin and rejection statistics.
2664 Don't display local sender or destination tables unless
2665 there is data to show.
2666 Added average volumes into the top table text output.
2668 SC/04 Eximstats V1.46
2669 Collect data on the number of addresses (recipients)
2670 as well as the number of messages.
2672 SC/05 Eximstats V1.47
2673 Added 'Message too big' to the list of mail rejection
2674 reasons (thanks to Marco Gaiarin).
2676 SC/06 Eximstats V1.48
2677 Mainlog lines which have GMT offsets and are too short to
2678 have a flag are now skipped.
2680 SC/07 Eximstats V1.49 (Alain Williams)
2681 Added the -emptyok flag.
2683 SC/08 Eximstats V1.50
2684 Fixes for obtaining the IP address from reject messages.
2686 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2687 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2688 whitespace changes from 4.64-PH/27
2690 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2693 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2694 are found, allow negative numbers in numeric criteria)
2696 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2698 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2699 to match changes made in 4.64-PH/43
2701 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2703 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2704 "domain" is actually an IP address, and force a failure. This locks out
2705 those revolvers/nameservers that support "A-for-A" lookups, in
2706 contravention of the specifications.
2708 PH/48 When a host name was looked up from an IP address, and the subsequent
2709 forward lookup of the name timed out, the host name was left in
2710 $sender_host_name, contrary to the specification.
2712 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2713 restricted to single-key lookups, Exim was not diagnosing an error if
2714 * or *@ was used with a query-style lookup.
2716 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2718 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2719 long ago, but noone interested enough thought of it. Let's just say that
2720 the "1.1" means that there are some new functions that weren't there at
2721 some point in the past.
2723 PH/51 Error processing for expansion failure of helo_data from an smtp
2724 transport during callout processing was broken.
2726 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2727 tested/used via the -bh/-bhc/-bs options.
2729 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2730 bug, fixed in subsequent PCRE releases).
2732 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2733 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2735 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2741 SC/01 Use a glob alias rather than an array ref in eximstats generated
2742 parser. This improves both readability and performance.
2744 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2745 Don't display local sender or destination tables in eximstats unless
2746 there is data to show.
2747 Added average volumes into the eximstats top table text output.
2749 SC/03 Collect data on the number of addresses (recipients) as well
2750 as the number of messages in eximstats.
2752 TF/01 Correct an error in the documentation for the redirect router. Exim
2753 does (usually) call initgroups() when daemonizing.
2755 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2756 with consistent privilege compared to when running as a daemon.
2758 TF/03 Note in the spec that $authenticated_id is not set for local
2759 submissions from trusted users.
2761 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2762 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2764 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2765 by adding some example configuration directives to the default
2766 configuration file. A little bit of work is required to uncomment the
2767 directives and define how usernames and passwords are checked, but
2768 there is now a framework to start from.
2770 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2771 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2772 without this. I don't know how relevant this is to other LDAP libraries.
2774 PH/02 Add the verb name to the "unknown ACL verb" error.
2776 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2778 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2780 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2781 directory not expanded when it should be if an expanded home directory
2782 was set for the address (which is overridden by the transport).
2784 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2787 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2788 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2789 because it is too late at that time, and has no effect.
2791 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2792 security issue with \' (bugzilla #107). I could not use the
2793 PQescapeStringConn() function, because it needs a PGconn value as one of
2796 PH/08 When testing addresses using -bt, indicate those final addresses that
2797 are duplicates that would not cause an additional delivery. At least one
2798 person was confused, thinking that -bt output corresponded to deliveries.
2799 (Suppressing duplicates isn't a good idea as you lose the information
2800 about possibly different redirections that led to the duplicates.)
2802 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2803 systems where poll() doesn't work, in particular OS X.
2805 PH/10 Added more information to debugging output for retry time not reached.
2807 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2808 operations in malware.c.
2810 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2813 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2814 syslog_duplication set false, log lines that would normally be written
2815 both the the main log and to the reject log were not written to syslog at
2818 PH/14 In the default configuration, change the use of "message" in ACL warn
2819 statements to "add_header".
2821 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2822 not followed by a command (e.g. "seen endif").
2824 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2825 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2828 PH/17 Added extra conditions to the default value of delay_warning_condition
2832 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2833 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2834 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2837 The Auto-Submitted: and various List- headers are standardised, whereas I
2838 don't think Precedence: ever was.
2840 PH/18 Refactored debugging code in route_finduser() to show more information,
2841 in particular, the error code if getpwnam() issues one.
2843 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2844 This is apparently needed in addition to the PH/07 change above to avoid
2845 any possible encoding problems.
2847 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2848 but not after initializing Perl.
2850 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2851 output them only if debugging. By default they are written stderr,
2852 apparently, which is not desirable.
2854 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2857 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2860 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2862 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2863 authenticated or an ident call has been made. Suppress the default
2864 values for $authenticated_id and $authenticated_sender (but permit -oMai
2865 and -oMas) when testing with -bh.
2867 PH/24 Re-jigged the order of the tests in the default configuration so that the
2868 tests for valid domains and recipients precede the DNS black list and CSA
2869 tests, on the grounds that those ones are more expensive.
2871 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2872 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2873 This bug exists in every version of Exim that I still have, right back to
2876 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2877 However, an attempt to turn on case-sensitivity in a regex key by
2878 including (?-i) didn't work because the subject string was already
2879 lowercased, and the effects were non-intuitive. It turns out that a
2880 one-line patch can be used to allow (?-i) to work as expected.
2886 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2887 other effects) broke the use of negated acl sub-conditions.
2889 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2892 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2893 "Deny" causes Exim to reject the incoming connection with a 554 error.
2894 Unfortunately, if there is a major crisis, such as a disk failure,
2895 tcp-wrappers gives "deny", whereas what one would like would be some
2896 kind of temporary error. A kludge has been added to help with this.
2897 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2898 554 error is used if errno is still zero or contains ENOENT (which occurs
2899 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2902 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2904 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2905 errors. Otherwise a message that provokes a temporary error (when other
2906 messages do not) can cause a whole host to time out.
2908 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2909 addresses were routed directly to files or pipes from a redirect router.
2910 File deliveries just didn't batch; pipe deliveries might have suffered
2913 PH/06 A failure to get a lock for a hints database would erroneously always say
2914 "Failed to get write lock", even when it was really a read lock.
2916 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2917 of 0600. This has been changed to use the value of the lockfile_mode
2918 option (which defaults to 0600).
2920 PH/08 Applied small patch from the Sieve maintainer.
2922 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2923 folder from quota calculations, a direct delivery into this folder messed
2924 up the contents of the maildirsize file. This was because the regex was
2925 used only to exclude .Trash (or whatever) when the size of the mailbox
2926 was calculated. There was no check that a delivery was happening into an
2927 excluded directory. This bug has been fixed by ignoring all quota
2928 processing for deliveries into excluded directories.
2930 PH/10 Added the maildirfolder_create_regex option to appendfile.
2936 PH/01 The code for finding all the local interface addresses on a FreeBSD
2937 system running IPv6 was broken. This may well have applied to all BSD
2938 systems, as well as to others that have similar system calls. The broken
2939 code found IPv4 interfaces correctly, but gave incorrect values for the
2940 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2941 that it would not match correctly against @[] and not recognize the IPv6
2944 PH/02 The ipliteral router was not recognizing addresses of the form user@
2945 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2947 PH/03 Added disable_ipv6.
2949 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2950 decoded form, because it is most often used to construct To: headers
2951 lines in autoreplies, and the decoded form may well be syntactically
2952 invalid. However, $reply_address has leading white space removed, and all
2953 newlines turned into spaces so that the autoreply transport does not
2956 PH/05 If group was specified without a user on a router, and no group or user
2957 was specified on a transport, the group from the router was ignored.
2959 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2960 for visible compile-time settings that can be used to change these
2961 numbers, for those that want even more. Backwards compatibility with old
2962 spool files has been maintained. However, going back to a previous Exim
2963 release will lost any variables that are in spool files.
2965 PH/07 Two small changes when running in the test harness: increase delay when
2966 passing a TCP/IP connection to a new process, in case the original
2967 process has to generate a bounce, and remove special handling of
2968 127.0.0.2 (sic), which is no longer necessary.
2970 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2971 be the same on different OS.
2973 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2976 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2977 whether --show-vars was specified or not
2979 JJ/02 exipick: Added support for new ACL variable spool format introduced
2982 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2983 syntactically invalid From: or Reply-to: line, and a filter used this to
2984 generate an autoreply, and therefore failed to obtain an address for the
2985 autoreply, Exim could try to deliver to a non-existent relative file
2986 name, causing unrelated and misleading errors. What now happens is that
2987 it logs this as a hard delivery error, but does not attempt to create a
2990 PH/11 The exinext utility has a -C option for testing purposes, but although
2991 the given file was scanned by exinext itself; it wasn't being passed on
2992 when Exim was called.
2994 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2995 an end-of-file indication when reading a command response.
2997 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2998 compiled. In many other places in Exim, IPv6 addresses are always
2999 recognized, so I have changed this. It also means that IPv4 domain
3000 literals of the form [IPV4:n.n.n.n] are now always recognized.
3002 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3003 used if the router is not running as root, for example, when verifying at
3004 ACL time, or when using -bh. The debugging output from this situation was
3005 non-existent - all you got was a failure to exec. I have made two
3008 (a) Failures to set uid/gid, the current directory, or a process leader
3009 in a subprocess such as that created by queryprogram now generate
3010 suitable debugging output when -d is set.
3012 (b) The queryprogram router detects when it is not running as root,
3013 outputs suitable debugging information if -d is set, and then runs
3014 the subprocess without attempting to change uid/gid.
3016 PH/15 Minor change to Makefile for building test_host (undocumented testing
3019 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3020 additional section of a DNS packet that returns MX or SRV records.
3021 Instead, it always explicitly searches for A/AAAA records. This avoids
3022 major problems that occur when a DNS server includes only records of one
3023 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3024 fixed another bug: if SRV records were looked up and the corresponding
3025 address records were *not* found in the additional section, the port
3026 values from the SRV records were lost.
3028 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3029 using the correct key (the original address) when searching the retry
3030 rules in order to find which one to use for generating the retry hint.
3032 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3033 adding the default one. Similarly, if it contains a Reply-To: header, the
3034 errors_reply_to option, if set, is not used.
3036 PH/19 When calculating a retry time, Exim used to measure the "time since
3037 failure" by looking at the "first failed" field in the retry record. Now
3038 it does not use this if it is later than than the arrival time of the
3039 message. Instead it uses the arrival time. This makes for better
3040 behaviour in cases where some deliveries succeed, thus re-setting the
3041 "first failed" field. An example is a quota failure for a huge message
3042 when small messages continue to be delivered. Without this change, the
3043 "time since failure" will always be short, possible causing more frequent
3044 delivery attempts for the huge message than are intended.
3045 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3047 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3048 $1, $2, $3) because the numerical variables can be reset during some
3049 expansion items (e.g. "match"), thereby losing the authentication data.
3051 PH/21 Make -bV show the size of off_t variables so that the test suite can
3052 decide whether to run tests for quotas > 2G.
3054 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3055 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3056 filecount value is greater than 2G or if a quota value is greater than 2G
3057 on a system where the size of off_t is not greater than 4, a panic error
3060 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3061 never match. The debug and -bh output now contains an explicit error
3062 message indicating a malformed IPv4 address or mask.
3064 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3065 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3066 PH/23 above applies.
3068 PH/25 Do not write to syslog when running in the test harness. The only
3069 occasion when this arises is a failure to open the main or panic logs
3070 (for which there is an explicit test).
3072 PH/26 Added the /no_tell option to "control=freeze".
3074 PH/27 If a host name lookup failed very early in a connection, for example, if
3075 the IP address matched host_lookup and the reverse lookup yielded a name
3076 that did not have a forward lookup, an error message of the form "no IP
3077 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3078 could be logged. Now it outputs the IP address instead of "NULL".
3080 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3081 allows the sender and the authenticated sender to be set when
3082 submitting a message from within Exim. Since child_open_exim() is
3083 documented for local_scan(), the new function should be too.
3085 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3086 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3087 results in an empty string is now treated as unset.
3089 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3091 PH/31 Added sender_verify_fail logging option.
3093 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3094 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3095 tidied the source and removed it altogether.
3097 PH/33 When a queue run was abandoned because the load average was too high, a
3098 log line was always written; now it is written only if the queue_run log
3099 selector is set. In addition, the log line for abandonment now contains
3100 information about the queue run such as the pid. This is always present
3101 in "start" and "stop" lines but was omitted from the "abandon" line.
3103 PH/34 Omit spaces between a header name and the colon in the error message that
3104 is given when verify = headers_syntax fails (if there are lots of them,
3105 the message gets confusing).
3107 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3108 names, as there are now some PTR records that contain slashes. This check
3109 is only to protect against broken name servers that fall over on strange
3110 characters, so the fact that it applies to all lookups doesn't matter.
3112 PH/36 Now that the new test suite is complete, we can remove some of the
3113 special code in Exim that was needed for the old test suite. For example,
3114 sorting DNS records because real resolvers return them in an arbitrary
3115 order. The new test suite's fake resolver always returns records in the
3118 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3119 bounces) except when queue_only is set, to avoid logging races between
3120 the different processes.
3122 PH/38 Panic-die if .include specifies a non-absolute path.
3124 PH/39 A tweak to the "H" retry rule from its user.
3126 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3127 a label. They prevented compilation on older perls.
3129 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3130 a warning to be raised on newish perls.
3132 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3133 on queue. Changes to match documented behaviour of showing count of
3134 messages matching specified criteria.
3136 PH/40 Changed the default ident timeout from 30s to 5s.
3138 PH/41 Added support for the use of login_cap features, on those BSD systems
3139 that have them, for controlling the resources used by pipe deliveries.
3141 PH/42 The content-scanning code uses fopen() to create files in which to put
3142 message data. Previously it was not paying any attention to the mode of
3143 the files. Exim runs with umask(0) because the rest of the code creates
3144 files with open(), and sets the required mode explicitly. Thus, these
3145 files were ending up world-writeable. This was not a big issue, because,
3146 being within the spool directory, they were not world-accessible. I have
3147 created a function called modefopen, which takes an additional mode
3148 argument. It sets umask(777), creates the file, chmods it to the required
3149 mode, then resets the umask. All the relevant calls to fopen() in the
3150 content scanning code have been changed to use this function.
3152 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3153 to 24 hours. This avoids potential overflow problems when processing G
3154 and H retry rules. I suspect nobody ever tinkers with this value.
3156 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3158 PH/45 When the plaintext authenticator is running as a client, the server's
3159 challenges are checked to ensure they are valid base64 strings. By
3160 default, the authentication attempt is cancelled if an invalid string is
3161 received. Setting client_ignore_invalid_base64 true ignores these errors.
3162 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3163 they are received. Thus, the responses can be made to depend on the
3164 challenges. If an invalid string is ignored, an empty string is placed in
3167 PH/46 Messages that are created by the autoreply transport now contains a
3168 References: header, in accordance with RFCs 2822 and 3834.
3170 PH/47 Added authenticated_sender_force to the smtp transport.
3172 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3174 PH/49 Installed latest patch from the Sieve maintainer.
3176 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3177 was also set, the appendfile transport was unnecessarily scanning a
3178 directory of message files (e.g. for maildir delivery) to find the count
3179 of files (along with the size), even though it did not need this
3180 information. It now does the scan only if it needs to find either the
3181 size of the count of files.
3183 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3185 PH/52 Two bugs concerned with error handling when the smtp transport is
3188 (i) Exim was not creating retry information for temporary errors given
3189 for individual recipients after the DATA command when the smtp transport
3190 was used in LMTP mode. This meant that they could be retried too
3191 frequently, and not timed out correctly.
3193 (ii) Exim was setting the flag that allows error details to be returned
3194 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3195 recipients that were returned after the DATA command.
3197 PH/53 This is related to PH/52, but is more general: for any failing address,
3198 when detailed error information was permitted to be returned to the
3199 sender, but the error was temporary, then after the final timeout, only
3200 "retry timeout exceeded" was returned. Now it returns the full error as
3201 well as "retry timeout exceeded".
3203 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3204 do this, and (what is worse) MTAs that accept it.
3206 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3207 will now be deprecated.
3209 PH/56 New os.c-cygwin from the Cygwin maintainer.
3211 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3212 formats (previously only available in exim formats via -bpr, -bpru,
3213 and -bpra. Now also available in native and exiqgrep formats)
3215 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3216 with very large, slow to parse queues
3218 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3220 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3222 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3223 responses to authentication challenges, though it was showing the
3224 challenges; (ii) I've removed the CR characters from the debug output for
3227 PH/58 Allow for the insertion of a newline as well as a space when a string
3228 is turned into more than one encoded-word during RFC 2047 encoding. The
3229 Sieve code now uses this.
3231 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3232 data_4xx, lost_connection, tls_required.
3234 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3235 message was being sent as an SMTP response.
3237 PH/61 Add -l and -k options to exicyclog.
3239 PH/62 When verifying, if an address was redirected to one new address, so that
3240 verification continued, and the new address failed or deferred after
3241 having set something in $address_data, the value of $address_data was not
3242 passed back to the ACL. This was different to the case when no
3243 redirection occurred. The value is now passed back in both cases.
3245 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3246 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3247 use login_cap.h, so on its own it isn't the distinguishing feature. The
3248 new name refers directly to the setclassresources() function.
3250 PH/65 Added configuration files for NetBSD3.
3252 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3254 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3255 is preferred over IPv4.
3257 PH/68 The bounce_return_message and bounce_return_body options were not being
3258 honoured for bounces generated during the reception of non-SMTP messages.
3259 In particular, this applied to messages rejected by the ACL. This bug has
3260 been fixed. However, if bounce_return_message is true and bounce_return_
3261 body is false, the headers that are returned for a non-SMTP message
3262 include only those that have been read before the error was detected.
3263 (In the case of an ACL rejection, they have all been read.)
3265 PH/69 The HTML version of the specification is now built in a directory called
3266 spec_html instead of spec.html, because the latter looks like a path with
3267 a MIME-type, and this confuses some software.
3269 PH/70 Catch two compiler warnings in sieve.c.
3271 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3272 function verify_get_ident() calls ip_connect() to connect a socket, but
3273 if the "connect()" function timed out, ip_connect() used to close the
3274 socket. However, verify_get_ident() also closes the socket later, and in
3275 between Exim writes to the log, which may get opened at this point. When
3276 the socket was closed in ip_connect(), the log could get the same file
3277 descriptor number as the socket. This naturally causes chaos. The fix is
3278 not to close the socket in ip_connect(); the socket should be closed by
3279 the function that creates it. There was only one place in the code where
3280 this was missing, in the iplookup router, which I don't think anybody now
3281 uses, but I've fixed it anyway.
3283 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3284 well as to direct DNS lookups. Otherwise the handling of names in host
3285 lists is inconsistent and therefore confusing.
3291 PH/01 Two changes to the default runtime configuration:
3293 (1) Move the checks for relay_from_hosts and authenticated clients from
3294 after to before the (commented out) DNS black list checks.
3296 (2) Add control=submission to the relay_from_hosts and authenticated
3297 clients checks, on the grounds that messages accepted by these
3298 statements are most likely to be submissions.
3300 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3302 (1) Generate an error if the third argument for the ${prvs expansion is
3305 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3308 (3) Reset the variables that are obtained from the first argument of
3309 ${prvscheck and used in the second argument before leaving the code,
3310 because their memory is reclaimed, so using them afterwards may do
3313 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3314 one (it's much easier than Tom thought :-).
3316 (5) Because of (4), we can now allow for the use of $prvscheck_result
3317 inside the third argument.
3319 PH/03 For some reason, the default setting of PATH when running a command from
3320 a pipe transport was just "/usr/bin". I have changed it to
3323 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3324 anything to be listed in the output from -bV.
3326 PH/05 When a filter generated an autoreply, the entire To: header line was
3327 quoted in the delivery log line, like this:
3329 => >A.N.Other <ano@some.domain> <original@ddress> ...
3331 This has been changed so that it extracts the operative address. There
3332 may be more than one such address. If so, they are comma-separated, like
3335 => >ano@some.domain,ona@other.domain <original@ddress> ...
3337 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3338 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3339 not being looked up in the rDNS to get a host name, Exim was showing the
3340 IP address twice in Received: lines, even though the IP addresses were
3341 identical. For example:
3343 Received: from [1.2.3.4] (helo=[1.2.3.4])
3345 However, if the real host name was known, it was omitting the HELO data
3346 if it matched the actual IP address. This has been tidied up so that it
3347 doesn't show the same IP address twice.
3349 PH/07 When both +timestamp and +memory debugging was on, the value given by
3350 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3351 called by the debug printing, thereby overwriting the timestamp buffer.
3352 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3354 PH/08 When the original message was included in an autoreply transport, it
3355 always said "this is a copy of the message, including all the headers",
3356 even if body_only or headers_only was set. It now gives an appropriate
3359 PH/09 Applied a patch from the Sieve maintainer which:
3361 o fixes some comments
3362 o adds the (disabled) notify extension core
3363 o adds some debug output for the result of if/elsif tests
3364 o points to the current vacation draft in the documentation
3365 and documents the missing references header update
3369 o fixes a bug in processing the envelope test (when testing
3370 multiple envelope elements, the last element determined the
3373 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3374 Electronic Mail") by including:
3376 Auto-submitted: auto-generated
3378 in the messages that it generates (bounce messages and others, such as
3379 warnings). In the case of bounce messages for non-SMTP messages, there was
3380 also a typo: it was using "Auto_submitted" (underscore instead of
3381 hyphen). Since every message generated by Exim is necessarily in response
3382 to another message, thes have all been changed to:
3384 Auto-Submitted: auto-replied
3386 in accordance with these statements in the RFC:
3388 The auto-replied keyword:
3390 - SHOULD be used on messages sent in direct response to another
3391 message by an automatic process,
3393 - MUST NOT be used on manually-generated messages,
3395 - MAY be used on Delivery Status Notifications (DSNs) and Message
3396 Disposition Notifications (MDNs),
3398 - MUST NOT be used on messages generated by automatic or periodic
3399 processes, except for messages which are automatic responses to
3402 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3403 to the default Received: header definition.
3405 PH/12 Added log selector acl_warn_skipped (default on).
3407 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3408 variables because (a) they are in the wrong storage pool and (b) even if
3409 they were copied, it wouldn't work properly because of the caching.
3411 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3412 checking when decoding. Apparently there are clients that generate
3413 overlong encoded strings. Why am I not surprised?
3415 PH/15 If the first argument of "${if match_address" was not empty, but did not
3416 contain an "@" character, Exim crashed. Now it writes a panic log message
3417 and treats the condition as false.
3419 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3421 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3422 extension "envelope-auth". The code is finished and in agreement with
3423 other implementations, but there is no documentation so far and in fact,
3424 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3425 not changing the active code.
3427 Print executed "if" and "elsif" statements when debugging is used. This
3428 helps a great deal to understand what a filter does.
3430 Document more things not specified clearly in RFC3028. I had all this
3431 sorted out, when out of a sudden new issues came to my mind. Oops."
3433 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3436 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3437 canonical form (as documented). However, after a host name lookup from
3438 the IP address, check_host() was doing a simple string comparison with
3439 addresses acquired from the DNS when checking that the found name did
3440 have the original IP as one of its addresses. Since any found IPv6
3441 addresses are likely to be in abbreviated form, the comparison could
3442 fail. Luckily, there already exists a function for doing the comparison
3443 by converting both addresses to binary, so now that is used instead of
3444 the text comparison.
3446 PH/20 There was another similar case to PH/19, when a complete host name was
3447 given in a host list; looking up its IP address could give an abbreviated
3448 form, whereas the current host's name might or might not be abbreviated.
3449 The same fix has been applied.
3455 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3456 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3459 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3460 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3462 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3464 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3465 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3466 Also, for "Auto-Submitted", treat anything other than "no" as
3467 non-personal, in accordance with RFC 3834. (Previously it treated
3468 anything starting "auto-" as non-personal.)
3470 TF/01 The control=submission/name=... option had a problem with syntax
3471 errors if the name included a slash character. The /name= option
3472 now slurps the rest of the string, so it can include any characters
3473 but it must come last in the list of options (after /sender_retain
3476 PH/05 Some modifications to the interface to the fake nameserver for the new
3484 TK/01 Added the "success_on_redirect" address verification option. See
3485 NewStuff for rationale and an example.
3487 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3489 PH/02 Patch to exigrep to allow it to work on syslog lines.
3491 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3492 fread() to skip over the body file's header line, because in Cygwin the
3493 header line is locked and is inaccessible.
3495 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3496 co-exist for some time) to make it clear that it is the Exim ID that is
3497 referenced, not the Message-ID: header line.
3499 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3500 string_format() function, because snprintf() does not exist on all
3503 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3504 ${stat: expansion item.
3506 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3507 protocol synchronization error", to keep the pedants happy.
3509 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3510 well as for IRIX systems, when gcc is being used. See the host.c source
3513 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3515 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3518 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3519 transport and to the smtp transport in LMTP mode.
3521 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3523 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3524 run a filter in a subprocess. This could lead to confusion in subsequent
3525 lookups in the parent process. There should also be a search_tidyup() at
3526 the end of the subprocess.
3528 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3529 only if the host matched helo_try_verify_hosts, which caused the
3530 verification to occur when the EHLO/HELO command was issued. The ACL just
3531 tested the remembered result. Now, if a previous verification attempt has
3532 not happened, "verify = helo" does it there and then.
3534 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3536 TK/03 Fix log output including CR from clamd.
3538 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3539 exist provoked a memory error which could cause a segfault.
3541 PH/15 Installed PCRE 6.2
3543 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3545 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3546 of the problem. Specifically, suggested +O2 rather than +O1 for the
3549 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3551 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3552 record, the port was not correctly logged when the outgoing_port log
3553 selector was set (it logged the transort's default port).
3555 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3556 fallback_hosts, and "hosts" in the smtp transport.
3558 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3559 host errors such as "Connection refused".
3561 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3562 authentication with radiusclient 0.4.9:
3564 - Error returned from rc_read_config was caught wrongly
3565 - Username/password not passed on to radius server due to wrong length.
3567 The presumption is that some radiusclient API changes for 4.51/PH/17
3568 were not taken care of correctly. The code is still untested by me (my
3569 Linux distribution still has 0.3.2 of radiusclient), but it was
3570 contributed by a Radius user.
3572 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3573 expanding the "port" option of the smtp transport.
3575 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3576 while reading a MIME header. Thanks to Tom Hughes for a patch.
3578 PH/24 Include config.h inside local_scan.h so that configuration settings are
3581 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3582 that in an ACL for RCPT (for example), you can examine exactly what was
3585 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3586 commands, but it was not correctly comparing the address with the actual
3587 client host address. Thus, it would show the EHLO address in Received:
3588 header lines when this was not necessary.
3590 PH/27 Added the % operator to ${eval:}.
3592 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3593 it should be ignoring failures (because with -C, for example, it has lost
3594 privilege). It wasn't ignoring creation failures other than "already
3597 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3600 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3601 because an input line was too long, either on its own, or by virtue of
3602 too many continuations, the temporary file was not being removed, and the
3603 return code was incorrect.
3605 PH/31 Missing "BOOL" in function definition in filtertest.c.
3607 PH/32 Applied Sieve patches from the maintainer.
3609 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3611 PH/33 Added "verify = not_blind".
3613 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3614 Local/Makefile (with some defaults set). These are used in built scripts
3615 such as exicyclog, but they have never been used in the exim_install
3616 script (though there are many overriding facilities there). I have
3617 arranged that the exim_install script now takes note of these two
3620 PH/35 Installed configuration files for Dragonfly.
3622 PH/36 When a locally submitted message by a trusted user did not contain a
3623 From: header, and the sender address was obtained from -f or from an SMTP
3624 MAIL command, and the trusted user did not use -F to supply a sender
3625 name, $originator_name was incorrectly used when constructing a From:
3626 header. Furthermore, $originator_name was used for submission mode
3627 messages from external hosts without From: headers in a similar way,
3628 which is clearly wrong.
3630 PH/37 Added control=suppress_local_fixups.
3632 PH/38 When log_selector = +received_sender was set, and the addition of the
3633 sender made the log line's construction buffer exactly full, or one byte
3634 less than full, an overflow happened when the terminating "\n" was
3637 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3638 when the result of a list match is failure because a DNS lookup failed.
3640 PH/40 RM_COMMAND is now used in the building process.
3642 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3643 the "build-* directories that it finds.
3645 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3646 address in a domain literal was a prefix of an interface address.
3648 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3649 when verifying a sender address, unless rewrite_headers is false.
3651 PH/44 Wrote a long comment about why errors_to addresses are verified as
3652 recipients, not senders.
3654 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3655 the ratelimit ACL was added.
3657 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3659 PH/46 Added extra information about PostgreSQL errors to the error string.
3661 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3662 avoiding the need to install special zones in a real server. This is
3663 backwards compatible; if it can't find the fake resolver, it drops back.
3664 Thus, both old and new test suites can be run.
3666 TF/02 Added util/ratelimit.pl
3668 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3669 clock is set back in time.
3671 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3672 Candler <B.Candler@pobox.com>.
3674 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3675 OK for addresses that are the result of redirections.
3677 PH/48 A number of further additions for the benefit of the new test suite,
3678 including a fake gethostbyname() that interfaces to the fake DNS resolver
3681 TF/06 The fix for widen_domains has also been applied to qualify_single and
3682 search_parents which are the other dnslookup options that can cause
3685 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3688 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3690 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3691 service" if it encountered an MX record with an empty target hostname.
3692 The message is now "an MX or SRV record indicated no SMTP service".
3694 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3695 if the DNS of the sending site is misconfigured. This is quite a
3696 common situation. This change restores the behaviour of treating a
3697 helo verification defer as a failure.
3699 PH/51 If self=fail was set on a router, the bounce message did not include the
3700 actual error message.
3706 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3708 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3709 command itself ended in error, the underlying message about the transport
3710 filter timeout was being overwritten with the pipe command error. Now the
3711 underlying error message should be appended to the second error message.
3713 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3715 PH/02 Reduce the amount of output that "make" produces by default. Full output
3716 can still be requested.
3718 PH/03 The warning log line about a condition test deferring for a "warn" verb
3719 was being output only once per connection, rather than after each
3720 occurrence (because it was using the same function as for successful
3721 "warn" verbs). This seems wrong, so I have changed it.
3723 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3724 it should not have, which might have caused a crash in the right
3725 circumstances, but probably never did.
3727 PH/04 Installed a modified version of Tony Finch's patch to make submission
3728 mode fix the return path as well as the Sender: header line, and to
3729 add a /name= option so that you can make the user's friendly name appear
3732 TF/03 Added the control = fakedefer ACL modifier.
3734 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3735 Mark Lowes for thorough testing.
3737 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3739 TK/03 Merged latest SRS patch from Miles Wilton.
3741 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3742 with the definition in sysexits.h (which is #included earlier).
3743 Fortunately, Exim does not actually use EX_OK. The code used to try to
3744 preserve the sysexits.h value, by assuming that macro definitions were
3745 scanned for macro replacements. I have been disabused of this notion,
3746 so now the code just undefines EX_OK before #including unistd.h.
3748 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3749 in the smtp transport. When a block could not be written in a single
3750 write() function, the timeout was being re-applied to each part-write.
3751 This seems wrong - if the receiver was accepting one byte at a time it
3752 would take for ever. The timeout is now adjusted when this happens. It
3753 doesn't have to be particularly precise.
3755 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3756 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3758 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3759 without a domain if the check to <postmaster@domain> fails.
3761 SC/01 Eximstats: added -xls and the ability to specify output files
3762 (patch written by Frank Heydlauf).
3764 SC/02 Eximstats: use FileHandles for outputting results.
3766 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3768 SC/04 Eximstats: fixed display of large numbers with -nvr option
3770 SC/05 Eximstats: fixed merging of reports with empty tables.
3772 SC/06 Eximstats: added the -include_original_destination flag
3774 SC/07 Eximstats: removed tabs and trailing whitespace.
3776 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3778 TK/06 MBOX spool code: Add real "From " MBOX separator line
3779 so the .eml file is really in mbox format (even though
3780 most programs do not really care). Patch from Alex Miller.
3782 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3783 The latter is generated from $received_to and is only set if the
3784 message has one envelope recipient. SA can use these headers,
3785 obviously out-of-the-box. Patch from Alex Miller.
3787 PH/08 The ${def test on a variable was returning false if the variable's
3788 value was "0", contrary to what the specification has always said!
3789 The result should be true unless the variable is empty.
3791 PH/09 The syntax error of a character other than { following "${if
3792 def:variable_name" (after optional whitespace) was not being diagnosed.
3793 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3794 accidental colon was present, for example, could give incorrect results.
3796 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3797 result is used (not including appendfile, where other changes are about
3800 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3801 This involved changing a lot of size variables from int to off_t. It
3802 should work with maildirs and everything.
3804 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3805 spamd dying while we are connected to it.
3807 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3810 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3811 function for BDB 4.3.
3813 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3815 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3816 the sources for PCRE have been re-arranged and more files are now
3819 PH/15 The code I had for printing potentially long long variables in PH/11
3820 above was not the best (it lost precision). The length of off_t variables
3821 is now inspected at build time, and an appropriate printing format (%ld
3822 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3823 to be "long long int" or "long int". This is needed for the internal
3824 formatting function string_vformat().
3826 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3827 the configuration file to be ":syslog", then the script "guesses" where
3828 the logs files are, rather than using the compiled in default. In our
3829 case the guess is not the same as the compiled default, so the script
3830 suddenly stopped working when I started to use syslog. The patch checks
3831 to see if log_file_path is "". If so, it attempts to read it from exim
3832 with no configuration file to get the compiled in version, before it
3833 falls back to the previous guessing code."
3835 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3836 implementing BATV in an Exim configuration. See NewStuff for the gory
3839 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3840 Makefile that are specific to HP-UX.
3842 PH/18 If the "use_postmaster" option was set for a recipient callout together
3843 with the "random" option, the postmaster address was used as the MAIL
3844 FROM address for the random test, but not for the subsequent recipient
3845 test. It is now used for both.
3847 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3848 patch removes a few documentation additions to RFC 3028, because the
3849 latest draft now contains them. It adds the new en;ascii-case comparator
3850 and a new error check for 8bit text in MIME parts. Comparator and
3851 require names are now matched exactly. I enabled the subaddress
3852 extension, but it is not well tested yet (read: it works for me)."
3854 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3855 rework some of the code of TK/09 above to avoid the hardwired use of
3856 "%lld" and "long long". Replaced the call to snprintf() with a call to
3859 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3860 records point to non-existent hosts", "retry timeout exceeded", and
3861 "retry time not reached for any host after a long failure period".
3863 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3864 experimental DomainKeys support:
3866 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3867 (2) On an error such as an illegally used "control", the wrong name for
3868 the control was given.
3870 These problems did NOT occur unless DomainKeys support was compiled.
3872 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3874 PH/24 Added ${if match_ip condition.
3876 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3877 databases so that it will be absolutely obvious if a crash occurs in the
3878 DB library. This is a regular occurrence (often caused by mis-matched
3881 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3882 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3883 picked up on a user's system that detects such things. There doesn't seem
3884 to be a gcc warning option for this - only an attribute that has to be
3885 put on the function's prototype. It seems that in Fedora Core 4 they have
3886 set this on a number of new functions. No doubt there will be more in due
3889 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3890 specify a transport. However, if an address that was verified by such a
3891 router was the subject of a callout, Exim crashed because it tried to
3892 read the rcpt_include_affixes from the non-existent transport. Now it
3893 just assumes that the setting of that option is false. This bug was
3894 introduced by 4.51/PH/31.
3896 PH/28 Changed -d+all to exclude +memory, because that information is very
3897 rarely of interest, but it makes the output a lot bigger. People tend to
3898 do -d+all out of habit.
3900 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3901 code in os-type was giving problems when libc.so lives in lib64, like on
3904 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3905 aren't the modern standard, and it seems that some systems' include files
3906 don't always have them. Exim was already checking for some of the newer
3907 ones like T_AAAA, and defining it itself. I've added checks for all the
3908 record types that Exim uses.
3910 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3911 not automatically generating a new one, as it is supposed to. This
3912 prevented TLS from working. If the file did exist, but contained invalid
3913 data, a new version was generated, as expected. It was only the case of a
3914 non-existent file that was broken.
3916 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3917 with a change in libdomainkeys > 0.64.
3919 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3920 from DNS. If the selector record carries the flag, it now has
3921 precedence over the domain-wide flag.
3923 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3925 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3926 the use of a port name that isn't defined in /etc/services) occurred, the
3927 message was deferred as in a normal delivery, and thus remained on the
3928 spool, instead of being failed because of the mua_wrapper setting. This
3929 is now fixed, and I tidied up some of the mua_wrapper messages at the
3932 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3933 the messages in a hash of arrays rather than using individual hashes.
3934 This is a bit cleaner and results in dramatic memory savings, albeit
3935 at a slight CPU cost.
3937 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3938 as requested by Marc Sherman.
3940 SC/10 Eximstats: added histograms for user specified patterns as requested
3943 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3945 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3946 fopen() in the content-scanning modules that did not already have it.
3952 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3953 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3955 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3957 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3959 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3960 to test Sieve filters that use "vacation".
3962 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3963 that changes the way the GnuTLS parameters are stored in the cache file.
3964 The new format can be generated externally. For backward compatibility,
3965 if the data in the cache doesn't make sense, Exim assumes it has read an
3966 old-format file, and it generates new data and writes a new file. This
3967 means that you can't go back to an older release without removing the
3970 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3971 work if there are any delivery delays because "one_time" forces the
3972 parent to be marked "delivered", so its unseen clone is never tried
3973 again. For this reason, Exim now forbids the simultaneous setting of
3976 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3977 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3978 entry if you want to know the details. The fix, however, neglected to
3979 consider the case when local delivery batching is involved. The test for
3980 "previously delivered" was not happening when checking to see if an
3981 address could be batched with a previous (undelivered) one; under
3982 certain circumstances this could lead to multiple deliveries to the same
3985 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3986 in its include files, and this causes problems building Exim.
3988 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3989 header_syntax) but Exim was just ignoring anything given after a slash.
3990 In particular, this caused confusion with an attempt to use "verify =
3991 reverse_host_lookup/defer_ok". An error is now given when options are
3992 supplied for verify items that do not have them. (Maybe reverse_host_
3993 lookup should have a defer_ok option, but that's a different point.)
3995 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3996 defined by RFC 821) to 2048, because there were problems with some AUTH
3997 commands, and RFC 1869 says the size should be increased for extended
3998 SMTP commands that take arguments.
4000 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4003 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4004 "unknown" error; now it says that the functionality isn't in the binary.
4006 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4007 an address' error message when a string expansion fails (syntax or
4008 whatever). Otherwise the password may appear in the log. Following change
4009 PH/42 below, there is no longer a chance of it appearing in a bounce
4012 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4014 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4015 of its addresses was ever tried. (Bugzilla bug #2.)
4017 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4018 the result incorrectly in the debug output. (It correctly added a newline
4019 to what was transported.)
4021 TF/01 Added $received_time.
4023 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4024 commented out examples of how to interface to a virus scanner and to
4025 SpamAssassin. Also added commented examples of av_scanner and
4026 spamd_address settings.
4028 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4029 and controls are allowed in which ACLs. There were a couple of minor
4030 errors. Some of the entries in the conditions table (which is a table of
4031 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4032 negation of where the condition IS allowed.
4034 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4036 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4037 header file does not have a version number, so I've had to invent a new
4038 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4039 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4040 radiusclient), but it was contributed by a Radius user.
4042 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4043 files or directories, for OpenSSL.
4045 PH/19 When an Exim process that is running as root has to create an Exim log
4046 file, it does so in a subprocess that runs as exim:exim so as to get the
4047 ownership right at creation (otherwise, other Exim processes might see
4048 the file with the wrong ownership). There was no test for failure of this
4049 fork() call, which would lead to the process getting stuck as it waited
4050 for a non-existent subprocess. Forks do occasionally fail when resources
4051 run out. I reviewed all the other calls to fork(); they all seem to check
4054 PH/20 When checking for unexpected SMTP input at connect time (before writing
4055 the banner), Exim was not dealing correctly with a non-positive return
4056 from the read() function. If the client had disconnected by this time,
4057 the result was a log entry for a synchronization error with an empty
4058 string after "input=" when read() returned zero. If read() returned -1
4059 (an event I could not check), uninitialized data bytes were printed.
4060 There were reports of junk text (parts of files, etc) appearing after
4063 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4065 PH/22 Added support for macro redefinition, and (re)definition in between
4066 driver and ACL definitions.
4068 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4069 forgetting to use the resulting value; it was using the unexpanded value.
4071 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4072 hadn't been configured. The fix is from Juergen Kreileder, who
4073 understands it better than I do:
4075 "Here's what I see happening with three configured cyrus_sasl
4076 authenticators configured (plain, login, cram-md5):
4078 On startup auth_cyrus_sasl_init() gets called for each of these.
4079 This means three calls to sasl_listmech() without a specified mech_list.
4080 => SASL tests which mechs of all available mechs actually work
4081 => three warnings about OTP not working
4082 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4084 With the patch, sasl_listmech() also gets called three times. But now
4085 SASL's mech_list option is set to the server_mech specified in the the
4086 authenticator. Or in other words, the answer from sasl_listmech()
4087 gets limited to just the mech you're testing for (which is different
4089 => the return list contains just 'plain' or 'login', 'cram-md5' or
4090 nothing depending on the value of ob->server_mech.
4092 I've just tested the patch: Authentication still works fine,
4093 unavailable mechs specified in the exim configuration are still
4094 caught, and the auth.log warnings about OTP are gone."
4096 PH/25 When debugging is enabled, the contents of the command line are added
4097 to the debugging output, even when log_selector=+arguments is not
4100 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4101 answer is "GNU", and only if the return is "GNU/something" is the answer
4104 PH/27 $acl_verify_message is now set immediately after the failure of a
4105 verification in an ACL, and so is available in subsequent modifiers. In
4106 particular, the message can be preserved by coding like this:
4108 warn !verify = sender
4109 set acl_m0 = $acl_verify_message
4111 Previously, $acl_verify_message was set only while expanding "message"
4112 and "log_message" when a very denied access.
4114 PH/28 Modified OS/os.c-Linux with
4116 -#ifndef OS_LOAD_AVERAGE
4117 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4119 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4120 nomenclature these days.)
4122 PH/29 Installed patch from the Sieve maintainer that adds the options
4123 sieve_useraddress and sieve_subaddress to the redirect router.
4125 PH/30 In these circumstances:
4126 . Two addresses routed to the same list of hosts;
4127 . First host does not offer TLS;
4128 . First host accepts first address;
4129 . First host gives temporary error to second address;
4130 . Second host offers TLS and a TLS session is established;
4131 . Second host accepts second address.
4132 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4133 and peerdn, if requested) that were in fact used only for the second
4136 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4137 attention to any local part prefix or suffix that was matched by the
4138 router that accepted the address. It now behaves in the same way as it
4139 does for delivery: the affixes are removed from the local part unless
4140 rcpt_include_affixes is set on the transport.
4142 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4143 timeout during the DATA phase of an incoming message.
4145 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4146 applied a patch sanctioned by the Sieve maintainer.
4148 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4149 the uid or gid is negative. A case of a negative gid caused this to be
4150 noticed. The fix allows for either to be negative.
4152 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4153 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4156 PH/36 The change PH/12 above was broken. Fixed it.
4158 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4159 the grounds that routing the same address twice would always produce the
4160 same answer. This might have been true once, but it is certainly no
4161 longer true now. Routing a child address may depend on the previous
4162 routing that produced that child. Some complicated redirection strategies
4163 went wrong when messages had multiple recipients, and made Exim's
4164 behaviour dependent on the order in which the addresses were given.
4166 I have moved the duplicate checking until after the routing is complete.
4167 Exim scans the addresses that are assigned to local and remote
4168 transports, and removes any duplicates. This means that more work will be
4169 done, as duplicates will always all be routed, but duplicates are
4170 presumably rare, so I don't expect this is of any significance.
4172 For deliveries to pipes, files, and autoreplies, the duplicate checking
4173 still happens during the routing process, since they are not going to be
4176 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4177 It corrects a timeout issue with spamd. This is Ian's comment: "The
4178 background is that sometimes spamd either never reads data from a
4179 connection it has accepted, or it never writes response data. The exiscan
4180 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4181 blindly assumes that writes won't block so it may never time out."
4183 PH/39 Allow G after quota size as well as K and M.
4185 PH/40 The value set for $authenticated_id in an authenticator may not contain
4186 binary zeroes or newlines because the value is written to log lines and
4187 to spool files. There was no check on this. Now the value is run through
4188 the string_printing() function so that such characters are converted to
4189 printable escape sequences.
4191 PH/41 $message_linecount is a new variable that contains the total number of
4192 lines in the message. Compare $body_linecount, which is the count for the
4195 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4196 bounce and delay warning messages, except in certain special cases, which
4199 (a) An SMTP error message from a remote host;
4200 (b) A message specified in a :fail: redirection;
4201 (c) A message specified in a "fail" command in a system filter;
4202 (d) A message specified in a FAIL return from the queryprogram router;
4203 (e) A message specified by the cannot_route_message router option.
4205 In these cases only, Exim does include the error details in bounce and
4206 warning messages. There are also a few cases where bland messages such
4207 as "unrouteable address" or "local delivery error" are given.
4209 PH/43 $value is now also set for the "else" part of a ${run expansion.
4211 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4212 being worked on, but at least Exim now implements the latest version to
4215 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4216 process to complete was treated as a delivery failure, a timeout while
4217 writing the message to the pipe was logged, but erroneously treated as a
4218 successful delivery. Such timeouts include transport filter timeouts. For
4219 consistency with the overall process timeout, these timeouts are now
4220 treated as errors, giving rise to delivery failures by default. However,
4221 there is now a new Boolean option for the pipe transport called
4222 timeout_defer, which, if set TRUE, converts the failures into defers for
4223 both kinds of timeout. A transport filter timeout is now identified in
4226 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4227 systems where "make" and "gmake" are different, calling "gmake" at top
4228 level broke things. I've arranged for the value of $(MAKE) to be passed
4229 from the Makefile to this script so that it can call the same version of
4233 A note about Exim versions 4.44 and 4.50
4234 ----------------------------------------
4236 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4237 changes of various kinds. As a consequence, a big documentation update was
4238 needed. This delayed the release for rather longer than seemed good, especially
4239 in the light of a couple of (minor) security issues. Therefore, the changes
4240 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4241 release. So 4.44 and 4.50 are in effect two different branches that both start
4244 I have left the 4.50 change log unchanged; it contains all the changes since
4245 4.43. The change log for 4.44 is below; many of its items are identical to
4246 those for 4.50. This seems to be the most sensible way to preserve the
4247 historical information.
4253 1. Minor wording change to the doc/README.SIEVE file.
4255 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4256 computation of the current number of files was incorrect.
4258 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4259 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4262 4. Give more explanation in the error message when the command for a transport
4263 filter fails to execute.
4265 5. There are several places where Exim runs a non-Exim command in a
4266 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4267 was being done only for the command run by the queryprogram router. It is
4268 now done for all such subprocesses. The other cases are: ${run, transport
4269 filters, and the commands run by the lmtp and pipe transports.
4271 6. Added CONFIGURE_GROUP build-time option.
4273 7. Some older OS have a limit of 256 on the maximum number of file
4274 descriptors. Exim was using setrlimit() to set 1000 as a large value
4275 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4276 systems. I've change it so that if it can't get 1000, it tries for 256.
4278 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4279 was an oversight, and furthermore, ever since the addition of extra
4280 controls (e.g. 4.43/32), the checks on when to allow different forms of
4281 "control" were broken. There should now be diagnostics for all cases when a
4282 control that does not make sense is encountered.
4284 9. Added the /retain_sender option to "control=submission".
4286 10. $recipients is now available in the predata ACL (oversight).
4288 11. Tidy the search cache before the fork to do a delivery from a message
4289 received from the command line. Otherwise the child will trigger a lookup
4290 failure and thereby defer the delivery if it tries to use (for example) a
4291 cached ldap connection that the parent has called unbind on.
4293 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4294 of $address_data from the recipient verification was clobbered by the
4295 sender verification.
4297 13. The value of address_data from a sender verification is now available in
4298 $sender_address_data in subsequent conditions in the ACL statement.
4300 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4302 15. Added a new option "connect=<time>" to callout options, to set a different
4305 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4306 was its contents. (It was OK if the option was not defined at all.)
4308 17. A "Completed" log line is now written for messages that are removed from
4309 the spool by the -Mrm option.
4311 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4312 information about exactly what failed.
4314 19. Added -dd to debug only the daemon process.
4316 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4317 handles timeouts, both on the server side and network timeouts. Renamed the
4318 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4320 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4321 It is now set to "smtps".
4323 22. $host_address is now set to the target address during the checking of
4324 ignore_target_hosts.
4326 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4327 being passed; this would have caused $sender_host_name to have been used if
4328 matching the list had actually called for a host name (not very likely,
4329 since this list is usually IP addresses). A host name is now passed as
4332 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4333 code that specifies a non-restarting handler (typically sigaction() in
4334 modern systems) in an attempt to fix a rare and obscure crash bug.
4336 25. Narrowed the window for a race in the daemon that could cause it to ignore
4337 SIGCHLD signals. This is not a major problem, because they are used only to
4338 wake it up if nothing else does.
4340 26. A malformed maildirsize file could cause Exim to calculate negative values
4341 for the mailbox size or file count. Odd effects could occur as a result.
4342 The maildirsize information is now recalculated if the size or filecount
4345 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4346 support for a long time. Removed HAVE_SYS_VFS_H.
4348 28. Installed the latest version of exipick from John Jetmore.
4350 29. In an address list, if the pattern was not a regular expression, an empty
4351 subject address (from a bounce message) matched only if the pattern was an
4352 empty string. Non-empty patterns were not even tested. This was the wrong
4353 because it is perfectly reasonable to use an empty address as part of a
4354 database query. An empty address is now tested by patterns that are
4355 lookups. However, all the other forms of pattern expect the subject to
4356 contain a local part and a domain, and therefore, for them, an empty
4357 address still always fails if the pattern is not itself empty.
4359 30. Exim went into a mad DNS loop when attempting to do a callout where the
4360 host was specified on an smtp transport, and looking it up yielded more
4361 than one IP address.
4363 31. Re-factored the code for checking spool and log partition space into a
4364 function that finds that data and another that does the check. The former
4365 is then used to implement four new variables: $spool_space, $log_space,
4366 $spool_inodes, and $log_inodes.
4368 32. The RFC2047 encoding function was originally intended for short strings
4369 such as real names; it was not keeping to the 75-character limit for
4370 encoded words that the RFC imposes. It now respects the limit, and
4371 generates multiple encoded words if necessary. To be on the safe side, I
4372 have increased the buffer size for the ${rfc2047: expansion operator from
4375 33. It is now permitted to omit both strings after an "if" condition; if the
4376 condition is true, the result is "true". As before, when the second string
4377 is omitted, a false condition yields an empty string. This makes it less
4378 cumbersome to write custom ACL and router conditions.
4380 34. Failure to deliver a bounce message always caused it to be frozen, even if
4381 there was an errors_to setting on the router. The errors_to setting is now
4384 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4385 canonical form (fully expanded) before being placed in
4386 $sender_host_address.
4388 36. The table in the code that translates DNS record types into text (T_A to
4389 "A" for instance) was missing entries for NS and CNAME. It is just possible
4390 that this could have caused confusion if both these types were looked up
4391 for the same domain, because the text type is used as part of Exim's
4392 per-process caching. But the chance of anyone hitting this buglet seems
4395 37. The dnsdb lookup has been extended in a number of ways.
4397 (1) There is a new type, "zns", which walks up the domain tree until it
4398 finds some nameserver records. It should be used with care.
4400 (2) There is a new type, "mxh", which is like "mx" except that it returns
4401 just the host names, not the priorities.
4403 (3) It is now possible to give a list of domains (or IP addresses) to be
4404 looked up. The behaviour when one of the lookups defers can be
4405 controlled by a keyword.
4407 (4) It is now possible to specify the separator character for use when
4408 multiple records are returned.
4410 38. The dnslists ACL condition has been extended: it is now possible to supply
4411 a list of IP addresses and/or domains to be looked up in a particular DNS
4414 39. Added log_selector=+queue_time_overall.
4416 40. When running the queue in the test harness, wait just a tad after forking a
4417 delivery process, to get repeatability of debugging output.
4419 41. Include certificate and key file names in error message when GnuTLS fails
4420 to set them up, because the GnuTLS error message doesn't include the name
4421 of the failing file when there is a problem reading it.
4423 42. Allow both -bf and -bF in the same test run.
4425 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4427 44. The "Exiscan patch" is now merged into the mainline Exim source.
4429 45. Sometimes the final signoff response after QUIT could fail to get
4430 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4431 < 0 before doing a fflush(). This bug looks as though it goes back to the
4432 introduction of TLS in release 3.20, but "sometimes" must have been rare
4433 because the tests only now provoked it.
4435 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4436 (this can affect the format of dates).
4438 47. exim_tidydb, when checking for the continued existence of a message for
4439 which it has found a message-specific retry record, was not finding
4440 messages that were in split spool directories. Consequently, it was
4441 deleting retry records that should have stayed in existence.
4443 48. Steve fixed some bugs in eximstats.
4445 49. The SPA authentication driver was not abandoning authentication and moving
4446 on to the next authenticator when an expansion was forced to fail,
4447 contradicting the general specification for all authenticators. Instead it
4448 was generating a temporary error. It now behaves as specified.
4450 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4451 (the order specifies the preference for clients). The order is now AES256,
4452 AES128, 3DES, ARCFOUR128.
4454 51. Small patch to Sieve code - explicitly set From: when generating an
4457 52. Exim crashed if a remote delivery caused a very long error message to be
4458 recorded - for instance if somebody sent an entire SpamAssassin report back
4459 as a large number of 550 error lines. This bug was coincidentally fixed by
4460 increasing the size of one of Exim's internal buffers (big_buffer) that
4461 happened as part of the Exiscan merge. However, to be on the safe side, I
4462 have made the code more robust (and fixed the comments that describe what
4465 53. Now that there can be additional text after "Completed" in log lines (if
4466 the queue_time_overall log selector is set), a one-byte patch to exigrep
4467 was needed to allow it to recognize "Completed" as not the last thing in
4470 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4471 patch that reportedly fixes this has been added. I am not expert enough to
4472 create a test for it. This is what the patch creator wrote:
4474 "I found a little strange behaviour of ldap code when working with
4475 Windows 2003 AD Domain, where users was placed in more than one
4476 Organization Units. When I tried to give exim partial DN, the exit code
4477 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4478 But simultaneously result of request was absolutely normal ldap result,
4479 so I produce this patch..."
4481 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4482 REFERENCE, so I have modified the code to exclude the patch when that macro
4485 55. Some experimental protocols are using DNS PTR records for new purposes. The
4486 keys for these records are domain names, not reversed IP addresses. The
4487 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4488 leaves it alone. Component reversal etc. now happens only for IP addresses.
4491 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4493 57. Double the size of the debug message buffer (to 2048) so that more of very
4494 long debug lines gets shown.
4496 58. The exicyclog utility now does better if the number of log files to keep
4497 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4499 59. Two changes related to the smtp_active_hostname option:
4501 (1) $smtp_active_hostname is now available as a variable.
4502 (2) The default for smtp_banner uses $smtp_active_hostname instead
4503 of $primary_hostname.
4505 60. The host_aton() function is supposed to be passed a string that is known
4506 to be a valid IP address. However, in the case of IPv6 addresses, it was
4507 not checking this. This is a hostage to fortune. Exim now panics and dies
4508 if the condition is not met. A case was found where this could be provoked
4509 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4510 components; fortuitously, this particular loophole had already been fixed
4511 by change 4.50/55 above.
4513 If there are any other similar loopholes, the new check in host_aton()
4514 itself should stop them being exploited. The report I received stated that
4515 data on the command line could provoke the exploit when Exim was running as
4516 exim, but did not say which command line option was involved. All I could
4517 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4518 running as the user.
4521 61. There was a buffer overflow vulnerability in the SPA authentication code
4522 (which came originally from the Samba project). I have added a test to the
4523 spa_base64_to_bits() function which I hope fixes it.
4526 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4527 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4529 63. The daemon start-up calls getloadavg() while still root for those OS that
4530 need the first call to be done as root, but it missed one case: when
4531 deliver_queue_load_max is set with deliver_drop_privilege. This is
4532 necessary for the benefit of the queue runner, because there is no re-exec
4533 when deliver_drop_privilege is set.
4535 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4536 This has been fixed.
4538 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4539 was in use, was not putting the data itself into the right store pool;
4540 consequently, it could be overwritten for a subsequent message in the same
4541 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4544 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4546 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4547 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4548 the function were treating the return as a boolean value, which happened to
4549 work because 0=false and not-0=true, but is not correct code.
4551 68. The host_aton() function was not handling scoped IPv6 addresses (those
4552 with, for example, "%eth0" on the end) correctly.
4554 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4555 negated items (that is, ~something) in unsigned ints. Some compilers
4556 apparently mutter when there is no cast.
4558 70. If an address verification called from an ACL failed, and did not produce a
4559 user-specific message (i.e. there was only a "system" message), nothing was
4560 put in $acl_verify_message. In this situation, it now puts the system
4563 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4564 session; change 4.31/43 added the unwanted input to the log line - except
4565 that it did not do this in the start of session case. It now does.
4567 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4568 This is wrong and can cause the other end to generate a synchronization
4569 error if it is another Exim or anything else that does the synchronization
4570 check. A QUIT command is no longer sent after a timeout.
4572 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4573 during host lookups.
4575 74. The defer_ok option of callout verification was not working if it was used
4576 when verifying addresses in header lines, that is, for this case:
4578 verify = header_sender/callout=defer_ok
4580 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4581 those file descriptors could be used for SMTP connections. If anything
4582 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4583 could be sent to the SMTP client, causing chaos. The daemon now opens
4584 stdin, stdout, and stderr to /dev/null when it puts itself into the
4587 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4588 log by default. The user can override this with suitable Perl magic.
4590 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4591 the log message when discard triggers, was not working for the DATA ACL or
4592 for the non-SMTP ACL.
4594 78. Error message wording change in sieve.c.
4596 79. If smtp_accept_max_per_host was set, the number of connections could be
4597 restricted to fewer than expected, because the daemon was trying to set up
4598 a new connection before checking whether the processes handling previous
4599 connections had finished. The check for completed processes is now done
4600 earlier. On busy systems, this bug wouldn't be noticed because something
4601 else would have woken the daemon, and it would have reaped the completed
4604 80. If a message was submitted locally by a user whose login name contained one
4605 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4606 It caused a spool format error. I have fixed the spool reading code. A
4607 related problem was that the "from" clause in the Received: line became
4608 illegal because of the space(s). It is now covered by ${quote_local_part.
4610 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4613 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4615 83. Patch from Sieve maintainer to fix unterminated string problem in
4616 "vacation" handling.
4618 84. Some minor changes to the Linux configuration files to help with other
4619 OS variants using glibc.
4621 85. One more patch for Sieve to update vacation handling to latest spec.
4624 ----------------------------------------------------
4625 See the note above about the 4.44 and 4.50 releases.
4626 ----------------------------------------------------
4632 1. Change 4.43/35 introduced a bug that caused file counts to be
4633 incorrectly computed when quota_filecount was set in an appendfile
4636 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4637 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4640 3. Give more explanation in the error message when the command for a transport
4641 filter fails to execute.
4643 4. There are several places where Exim runs a non-Exim command in a
4644 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4645 was being done only for the command run by the queryprogram router. It is
4646 now done for all such subprocesses. The other cases are: ${run, transport
4647 filters, and the commands run by the lmtp and pipe transports.
4649 5. Some older OS have a limit of 256 on the maximum number of file
4650 descriptors. Exim was using setrlimit() to set 1000 as a large value
4651 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4652 systems. I've change it so that if it can't get 1000, it tries for 256.
4654 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4655 was an oversight, and furthermore, ever since the addition of extra
4656 controls (e.g. 4.43/32), the checks on when to allow different forms of
4657 "control" were broken. There should now be diagnostics for all cases when a
4658 control that does not make sense is encountered.
4660 7. $recipients is now available in the predata ACL (oversight).
4662 8. Tidy the search cache before the fork to do a delivery from a message
4663 received from the command line. Otherwise the child will trigger a lookup
4664 failure and thereby defer the delivery if it tries to use (for example) a
4665 cached ldap connection that the parent has called unbind on.
4667 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4668 of $address_data from the recipient verification was clobbered by the
4669 sender verification.
4671 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4672 was its contents. (It was OK if the option was not defined at all.)
4674 11. A "Completed" log line is now written for messages that are removed from
4675 the spool by the -Mrm option.
4677 12. $host_address is now set to the target address during the checking of
4678 ignore_target_hosts.
4680 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4681 being passed; this would have caused $sender_host_name to have been used if
4682 matching the list had actually called for a host name (not very likely,
4683 since this list is usually IP addresses). A host name is now passed as
4686 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4687 code that specifies a non-restarting handler (typically sigaction() in
4688 modern systems) in an attempt to fix a rare and obscure crash bug.
4690 15. Narrowed the window for a race in the daemon that could cause it to ignore
4691 SIGCHLD signals. This is not a major problem, because they are used only to
4692 wake it up if nothing else does.
4694 16. A malformed maildirsize file could cause Exim to calculate negative values
4695 for the mailbox size or file count. Odd effects could occur as a result.
4696 The maildirsize information is now recalculated if the size or filecount
4699 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4700 support for a long time. Removed HAVE_SYS_VFS_H.
4702 18. Updated exipick to current release from John Jetmore.
4704 19. Allow an empty sender to be matched against a lookup in an address list.
4705 Previously the only cases considered were a regular expression, or an
4708 20. Exim went into a mad DNS lookup loop when doing a callout where the
4709 host was specified on the transport, if the DNS lookup yielded more than
4712 21. The RFC2047 encoding function was originally intended for short strings
4713 such as real names; it was not keeping to the 75-character limit for
4714 encoded words that the RFC imposes. It now respects the limit, and
4715 generates multiple encoded words if necessary. To be on the safe side, I
4716 have increased the buffer size for the ${rfc2047: expansion operator from
4719 22. Failure to deliver a bounce message always caused it to be frozen, even if
4720 there was an errors_to setting on the router. The errors_to setting is now
4723 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4724 canonical form (fully expanded) before being placed in
4725 $sender_host_address.
4727 24. Updated eximstats to version 1.33
4729 25. Include certificate and key file names in error message when GnuTLS fails
4730 to set them up, because the GnuTLS error message doesn't include the name
4731 of the failing file when there is a problem reading it.
4733 26. Expand error message when OpenSSL has problems setting up cert/key files.
4736 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4737 (this can affect the format of dates).
4739 28. exim_tidydb, when checking for the continued existence of a message for
4740 which it has found a message-specific retry record, was not finding
4741 messages that were in split spool directories. Consequently, it was
4742 deleting retry records that should have stayed in existence.
4744 29. eximstats updated to version 1.35
4745 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4746 1.35 - bugfix such that pie charts by volume are generated correctly
4748 30. The SPA authentication driver was not abandoning authentication and moving
4749 on to the next authenticator when an expansion was forced to fail,
4750 contradicting the general specification for all authenticators. Instead it
4751 was generating a temporary error. It now behaves as specified.
4753 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4754 (the order specifies the preference for clients). The order is now AES256,
4755 AES128, 3DES, ARCFOUR128.
4757 31. Small patch to Sieve code - explicitly set From: when generating an
4760 32. Exim crashed if a remote delivery caused a very long error message to be
4761 recorded - for instance if somebody sent an entire SpamAssassin report back
4762 as a large number of 550 error lines. This bug was coincidentally fixed by
4763 increasing the size of one of Exim's internal buffers (big_buffer) that
4764 happened as part of the Exiscan merge. However, to be on the safe side, I
4765 have made the code more robust (and fixed the comments that describe what
4768 33. Some experimental protocols are using DNS PTR records for new purposes. The
4769 keys for these records are domain names, not reversed IP addresses. The
4770 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4771 leaves it alone. Component reversal etc. now happens only for IP addresses.
4774 34. The host_aton() function is supposed to be passed a string that is known
4775 to be a valid IP address. However, in the case of IPv6 addresses, it was
4776 not checking this. This is a hostage to fortune. Exim now panics and dies
4777 if the condition is not met. A case was found where this could be provoked
4778 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4779 components; fortuitously, this particular loophole had already been fixed
4780 by change 4.50/55 or 4.44/33 above.
4782 If there are any other similar loopholes, the new check in host_aton()
4783 itself should stop them being exploited. The report I received stated that
4784 data on the command line could provoke the exploit when Exim was running as
4785 exim, but did not say which command line option was involved. All I could
4786 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4787 running as the user.
4790 35. There was a buffer overflow vulnerability in the SPA authentication code
4791 (which came originally from the Samba project). I have added a test to the
4792 spa_base64_to_bits() function which I hope fixes it.
4795 36. The daemon start-up calls getloadavg() while still root for those OS that
4796 need the first call to be done as root, but it missed one case: when
4797 deliver_queue_load_max is set with deliver_drop_privilege. This is
4798 necessary for the benefit of the queue runner, because there is no re-exec
4799 when deliver_drop_privilege is set.
4801 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4802 was in use, was not putting the data itself into the right store pool;
4803 consequently, it could be overwritten for a subsequent message in the same
4804 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4807 38. Sometimes the final signoff response after QUIT could fail to get
4808 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4809 < 0 before doing a fflush(). This bug looks as though it goes back to the
4810 introduction of TLS in release 3.20, but "sometimes" must have been rare
4811 because the tests only now provoked it.
4817 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4818 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4819 Now it returns an integer. A number of calls were still expecting a T/F
4820 return. Fortuitously, in all cases, the tests worked in OK situations,
4821 which is the norm. However, things would have gone wrong on any write
4822 failures on the smtp file descriptor. This function is used when sending
4823 messages over SMTP and also when doing verify callouts.
4825 2. When Exim is called to do synchronous delivery of a locally submitted
4826 message (the -odf or -odi options), it no longer closes stderr before doing
4829 3. Implemented the mua_wrapper option.
4831 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4833 5. Implemented the functions header_remove(), header_testname(),
4834 header_add_at_position(), and receive_remove_recipient(), and exported them
4837 6. If an ACL "warn" statement specified the addition of headers, Exim already
4838 inserted X-ACL-Warn: at the start if there was no header name. However, it
4839 was not making this test for the second and subsequent header lines if
4840 there were newlines in the string. This meant that an invalid header could
4841 be inserted if Exim was badly configured.
4843 7. Allow an ACL "warn" statement to add header lines at the start or after all
4844 the Received: headers, as well as at the end.
4846 8. Added the rcpt_4xx retry error code.
4848 9. Added postmaster_mailfrom=xxx to callout verification option.
4850 10. Added mailfrom=xxxx to the callout verification option, for verify=
4853 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4854 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4856 12. Inserted some casts to stop certain compilers warning when using pointer
4857 differences as field lengths or precisions in printf-type calls (mostly
4858 affecting debugging statements).
4860 13. Added optional readline() support for -be (dynamically loaded).
4862 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4863 same clock tick as a message's arrival, so that its received time was the
4864 same as the "first fail" time on the retry record, and that message
4865 remained on the queue past the ultimate address timeout, every queue runner
4866 would try a delivery (because it was past the ultimate address timeout) but
4867 after another failure, the ultimate address timeout, which should have then
4868 bounced the address, did not kick in. This was a "< instead of <=" error;
4869 in most cases the first failure would have been in the next clock tick
4870 after the received time, and all would be well.
4872 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4873 being recognized when the domain list was tested by the match_domain
4874 condition in an expansion string.
4876 16. Added the ${str2b64: operator.
4878 17. Exim was always calling setrlimit() to set a large limit for the number of
4879 processes, without checking whether the existing limit was already
4880 adequate. (It did check for the limit on file descriptors.) Furthermore,
4881 errors from getrlimit() and setrlimit() were being ignored. Now they are
4882 logged to the main and panic logs, but Exim does carry on, to try to do its
4883 job under whatever limits there are.
4885 18. Imported PCRE 5.0.
4887 19. Trivial typo in log message " temporarily refused connection" (the leading
4890 20. If the log selector return_path_on_delivery was set and an address was
4891 redirected to /dev/null, the delivery process crashed because it assumed
4892 that a return path would always be set for a "successful" delivery. In this
4893 case, the whole delivery is bypassed as an optimization, and therefore no
4896 21. Internal re-arrangement: the function for sending a challenge and reading
4897 a response while authentication was assuming a zero-terminated challenge
4898 string. It's now changed to take a pointer and a length, to allow for
4899 binary data in such strings.
4901 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4903 23. Exim was not respecting finduser_retries when seeking the login of the
4904 uid under which it was called; it was always trying 10 times. (The default
4905 setting of finduser_retries is zero.) Also, it was sleeping after the final
4906 failure, which is pointless.
4908 24. Implemented tls_on_connect_ports.
4910 25. Implemented acl_smtp_predata.
4912 26. If the domain in control=submission is set empty, Exim assumes that the
4913 authenticated id is a complete email address when it generates From: or
4914 Sender: header lines.
4916 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4917 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4918 chown and chgrp in /bin and hostname in /usr/bin.
4920 28. Exim was keeping the "process log" file open after each use, just as it
4921 does for the main log. This opens the possibility of it remaining open for
4922 long periods when the USR1 signal hits a daemon. Occasional processlog
4923 errors were reported, that could have been caused by this. Anyway, it seems
4924 much more sensible not to leave this file open at all, so that is what now
4927 29. The long-running daemon process does not normally write to the log once it
4928 has entered its main loop, and it closes the log before doing so. This is
4929 so that log files can straightforwardly be renamed and moved. However,
4930 there are a couple of unusual error situations where the daemon does write
4931 log entries, and I had neglected to close the log afterwards.
4933 30. The text of an SMTP error response that was received during a remote
4934 delivery was being truncated at 512 bytes. This is too short for some of
4935 the long messages that one sometimes sees. I've increased the limit to
4938 31. It is now possible to make retry rules that apply only when a message has a
4939 specific sender, in particular, an empty sender.
4941 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4942 it possible to be selective about when SMTP synchronization is enforced.
4944 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4946 32. Implemented hosts_connection_nolog.
4948 33. Added an ACL for QUIT.
4950 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4953 35. Added mailbox_size and mailbox_filecount to appendfile.
4955 36. Added control = no_multiline_responses to ACLs.
4957 37. There was a bug in the logic of the code that waits for the clock to tick
4958 in the case where the clock went backwards by a substantial amount such
4959 that the microsecond fraction of "now" was more than the microsecond
4960 fraction of "then" (but the whole seconds number was less).
4962 38. Added support for the libradius Radius client library this is found on
4963 FreeBSD (previously only the radiusclient library was supported).
4969 1. When certain lookups returned multiple values in the form name=value, the
4970 quoting of the values was not always being done properly. Specifically:
4971 (a) If the value started with a double quote, but contained no whitespace,
4973 (b) If the value contained whitespace other than a space character (i.e.
4974 tabs or newlines or carriage returns) it was not quoted.
4975 This fix has been applied to the mysql and pgsql lookups by writing a
4976 separate quoting function and calling it from the lookup code. The fix
4977 should probably also be applied to nisplus, ibase and oracle lookups, but
4978 since I cannot test any of those, I have not disturbed their existing code.
4980 2. A hit in the callout cache for a specific address caused a log line with no
4981 reason for rejecting RCPT. Now it says "Previous (cached) callout
4982 verification failure".
4984 3. There was an off-by-one bug in the queryprogram router. An over-long
4985 return line was truncated at 256 instead of 255 characters, thereby
4986 overflowing its buffer with the terminating zero. As well as fixing this, I
4987 have increased the buffer size to 1024 (and made a note to document this).
4989 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4990 when Exim is waiting for an SMTP response from a remote server, Exim
4991 restarts its select() call on the socket, thereby resetting its timeout.
4992 This is not a problem when such interrupts are rare. Somebody set up a cron
4993 job to run exiwhat every 2 minutes, which is less than the normal select()
4994 timeout (5 or 10 minutes). This meant that the select() timeout never
4995 kicked in because it was always reset. I have fixed this by comparing the
4996 time when an interrupt arrives with the time at the start of the first call
4997 to select(). If more time than the timeout has elapsed, the interrupt is
4998 treated as a timeout.
5000 5. Some internal re-factoring in preparation for the addition of Sieve
5001 extensions (by MH). In particular, the "personal" test is moved to a
5002 separate function, and given an option for scanning Cc: and Bcc: (which is
5003 not set for Exim filters).
5005 6. When Exim created an email address using the login of the caller as the
5006 local part (e.g. when creating a From: or Sender: header line), it was not
5007 quoting the local part when it contained special characters such as @.
5009 7. Installed new OpenBSD configuration files.
5011 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5012 try to make them clearer.
5014 9. Callout options, other than the timeout value, were being ignored when
5015 verifying sender addresses in header lines. For example, when using
5017 verify = header_sender/callout=no_cache
5019 the cache was (incorrectly) being used.
5021 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5022 only the Cygwin environment.
5024 11. When return_path_on_delivery was set as a log selector, if different remote
5025 addresses in the same message used different return paths and parallel
5026 remote delivery occurred, the wrong values would sometimes be logged.
5027 (Whenever a remote delivery process finished, the return path value from
5028 the most recently started remote delivery process was logged.)
5030 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5031 lines when AUTH and/or TLS are in use. This is the "received protocol"
5032 field. Exim used to use "asmtp" for authenticated SMTP, without any
5033 indication (in the protocol name) for TLS use. Now it follows the RFC and
5034 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5035 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5036 names appear in log lines as well as in Received: header lines.
5038 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5039 extensions, and comparison tests, and to fix some bugs.
5041 14. Changes to the "personal" filter test:
5043 (1) The test was buggy in that it was just doing the equivalent of
5044 "contains" tests on header lines. For example, if a user's address was
5045 anne@some.where, the "personal" test would incorrectly be true for
5047 To: susanne@some.where
5049 This test is now done by extracting each address from the header in turn,
5050 and checking the entire address. Other tests that are part of "personal"
5051 are now done using regular expressions (for example, to check local parts
5052 of addresses in From: header lines).
5054 (2) The list of non-personal local parts in From: addresses has been
5055 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5056 taken from the Sieve specification recommendations.
5058 (3) If the message contains any header line starting with "List-" it is
5059 treated as non-personal.
5061 (4) The test for "circular" in the Subject: header line has been removed
5062 because it now seems ill-conceived.
5064 15. Minor typos in src/EDITME comments corrected.
5066 16. Installed latest exipick from John Jetmore.
5068 17. If headers_add on a router specified a text string that was too long for
5069 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5070 of string_sprintf() is now avoided.
5072 18. $message_body_size was not set (it was always zero) when running the DATA
5073 ACL and the local_scan() function.
5075 19. For the "mail" command in an Exim filter, no default was being set for
5076 the once_repeat time, causing a random time value to be used if "once" was
5077 specified. (If the value happened to be <= 0, no repeat happened.) The
5078 default is now 0s, meaning "never repeat". The "vacation" command was OK
5079 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5080 (I found it when inspecting the code).
5082 20. There is now an overall timeout for performing a callout verification. It
5083 defaults to 4 times the callout timeout, which applies to individual SMTP
5084 commands during the callout. The overall timeout applies when there is more
5085 than one host that can be tried. The timeout is checked before trying the
5086 next host. This prevents very long delays if there are a large number of
5087 hosts and all are timing out (e.g. when the network connections are timing
5088 out). The value of the overall timeout can be changed by specifying an
5089 additional sub-option for "callout", called "maxwait". For example:
5091 verify = sender/callout=5s,maxwait=20s
5093 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5094 to the end before writing, but this should make it even safer).
5096 22. Exim was forgetting that it had advertised PIPELINING for the second and
5097 subsequent messages on an SMTP connection. It was also not resetting its
5098 memory on STARTTLS and an internal HELO.
5100 23. When Exim logs an SMTP synchronization error within a session, it now
5101 records whether PIPELINING has been advertised or not.
5103 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5104 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5105 rather than long int.
5107 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5109 26. Added the never_mail option to autoreply.
5115 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5116 crash if the getsockname() call failed; this can happen if a connection is
5117 closed very soon after it is established. The problem was simply in the
5118 order in which certain operations were done, causing Exim to try to write
5119 to the SMTP stream before it had set up the file descriptor. The bug has
5120 been fixed by making things happen in the correct order.
5126 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5127 before the connection was closed, thus losing the rejection response.
5129 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5130 some early Solaris releases, but causes trouble in current releases where
5131 socklen_t is defined.
5133 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5136 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5139 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5140 that releasing the top of it at the end releases what was used for sub-
5141 expansions (unless the block got too big). However, discard this block if
5142 the first thing is a variable or header, so that we can use its block when
5143 it is dynamic (useful for very large $message_headers, for example).
5145 6. Lookups now cache *every* query, not just the most recent. A new, separate
5146 store pool is used for this. It can be recovered when all lookup caches are
5147 flushed. Lookups now release memory at the end of their result strings.
5148 This has involved some general refactoring of the lookup sources.
5150 7. Some code has been added to the store_xxx() functions to reduce the amount
5151 of flapping under certain conditions.
5153 8. log_incoming_interface used to affect only the <= reception log lines. Now
5154 it causes the local interface and port to be added to several more SMTP log
5155 lines, for example "SMTP connection from", and rejection lines.
5157 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5159 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5161 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5162 could be overwritten at the end of the current message (or the start of a
5163 new message if it was set in a HELO ACL). The value is now preserved for
5164 the duration of the SMTP connection.
5166 12. If a transport had a headers_rewrite setting, and a matching header line
5167 contained an unqualified address, that address was qualified, even if it
5168 did not match any rewriting rules. The underlying bug was that the values
5169 of the flags that permit the existence of unqualified sender and recipient
5170 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5171 non-local messages, and by -bnq for local messages) were not being
5172 preserved with the message after it was received.
5174 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5175 "next input=" as part of the text comprising the host identity instead of
5176 the correct text. The code was using the same buffer for two different
5177 strings. However, depending on which order the printing function evaluated
5178 its arguments, the bug did not always show up. Under Linux, for example, my
5179 test suite worked just fine.
5181 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5182 doesn't work with some older versions of Perl. It has been changed to "my",
5183 which in any case is probably the better facility to use.
5185 15. A really picky compiler found some instances of statements for creating
5186 error messages that either had too many or two few arguments for the format
5189 16. The size of the buffer for calls to the DNS resolver has been increased
5190 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5191 for addresses that have a lot of PTR records. This alleviates a problem; it
5192 does not fully solve it.
5194 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5195 buffer now truncates the list and logs the incident, which is the same
5196 action as happens when Exim is looking up a host name and its aliases.
5197 Previously in this situation something unpredictable would happen;
5198 sometimes it was "internal error: store_reset failed".
5200 18. If a server dropped the connection unexpectedly when an Exim client was
5201 using GnuTLS and trying to read a response, the client delivery process
5202 crashed while trying to generate an error log message.
5204 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5205 string, for example:
5207 warn message = H1: something\nH2: something
5209 the text was added as a single header line from Exim's point of view
5210 though it ended up OK in the delivered message. However, searching for the
5211 second and subsequent header lines using $h_h2: did not work. This has been
5212 fixed. Similarly, if a system filter added multiple headers in this way,
5213 the routers could not see them.
5215 20. Expanded the error message when iplsearch is called with an invalid key to
5216 suggest using net-iplsearch in a host list.
5218 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5219 ACLs are no longer actually imposed (and a message to that effect is
5222 22. If a "gecos" field in a passwd entry contained escaped characters, in
5223 particular, if it contained a \" sequence, Exim got it wrong when building
5224 a From: or a Sender: header from that name. A second bug also caused
5225 incorrect handling when an unquoted " was present following a character
5226 that needed quoting.
5228 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5229 was not being matched caselessly.
5231 24. Arranged for all hyphens in the exim.8 source to be escaped with
5234 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5235 properly. Recipient callout cache records were still being keyed to include
5236 the sender, even when use_sender was set false. This led to far more
5237 callouts that were necessary. The sender is no longer included in the key
5238 when use_sender is false.
5240 26. Added "control = submission" modifier to ACLs.
5242 27. Added the ${base62d: operator to decode base 62 numbers.
5244 28. dnsdb lookups can now access SRV records.
5246 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5247 the configuration file.
5249 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5250 (-v) mode. This makes the output for a verbose queue run more intelligible.
5252 31. Added a use_postmaster feature to recipient callouts.
5254 32. Added the $body_zerocount variable, containing the number of binary zero
5255 bytes in the message body.
5257 33. The time of last modification of the "new" subdirectory is now used as the
5258 "mailbox time last read" when there is a quota error for a maildir
5261 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5263 35. Added +ignore_unknown as a special item in host lists.
5265 36. Code for decoding IPv6 addresses in host lists is now included, even if
5266 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5267 address was recognized as an IP address, but was then not correctly decoded
5268 into binary, causing unexpected and incorrect effects when compared with
5275 1. Very minor rewording of debugging text in manualroute to say "list of
5276 hosts" instead of "hostlist".
5278 2. If verify=header_syntax was set, and a header line with an unqualified
5279 address (no domain) and a large number of spaces between the end of the
5280 name and the colon was received, the reception process suffered a buffer
5281 overflow, and (when I tested it) crashed. This was caused by some obsolete
5282 code that should have been removed. The fix is to remove it!
5284 3. When running in the test harness, delay a bit after writing a bounce
5285 message to get a bit more predictability in the log output.
5287 4. Added a call to search_tidyup() just before forking a reception process. In
5288 theory, someone could use a lookup in the expansion of smtp_accept_max_
5289 per_host which, without the tidyup, could leave open a database connection.
5291 5. Added the variables $recipient_data and $sender_data which get set from a
5292 lookup success in an ACL "recipients" or "senders" condition, or a router
5293 "senders" option, similar to $domain_data and $local_part_data.
5295 6. Moved the writing of debug_print from before to after the "senders" test
5298 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5299 problems for message scanning, either using a data ACL, or using
5300 local_scan() because the Received: header was not generated till after they
5301 were called (in order to set the time as the time of reception completion).
5302 I have revised the way this works. The header is now generated after the
5303 body is received, but before the ACL or local_scan() are called. After they
5304 are run, the timestamp in the header is updated.
5310 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5311 before starting a queue runner without re-exec. This happened only when
5312 deliver_drop_privilege was set or when the Exim user was set to root. The
5313 effect of the bug was that timeouts during subsequent deliveries caused
5314 crashes instead of being properly handled. The handler is now left at its
5315 default (and expected) setting.
5317 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5318 message, again when deliver_drop_privilege is set or Exim is run as root.
5319 The bug described in (1) was not present in this case, but the tidying up
5320 of the other signals was missing. I have made the two cases consistent.
5322 3. The ignore_target_hosts setting on a manualroute router was being ignored
5323 for hosts that were looked up using the /MX notation.
5325 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5328 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5329 operated on the sender address. After changing the $sender_address to <>
5330 for the sender address verify, Exim was re-instated it as the original
5331 (before rewriting) address, but remembering that it had rewritten it, so it
5332 wasn't rewriting it again. This bug also had the effect of breaking the
5333 sender address verification caching when the sender address was rewritten.
5335 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5336 This has been changed so that if the ip literal address matches
5337 ignore_target_hosts, the router declines.
5339 7. Added expansion conditions match_domain, match_address, and match_local_
5340 part (NOT match_host).
5342 8. The placeholder for the Received: header didn't have a length field set.
5344 9. Added code to Exim itself and to exim_lock to test for a specific race
5345 condition that could lead to file corruption when using MBX delivery. The
5346 issue is with the lockfile that is created in /tmp. If this file is removed
5347 after a process has opened it but before that process has acquired a lock,
5348 there is the potential for a second process to recreate the file and also
5349 acquire a lock. This could lead to two Exim processes writing to the file
5350 at the same time. The added code performs the same test as UW imapd; it
5351 checks after acquiring the lock that its file descriptor still refers to
5352 the same named file.
5354 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5355 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5358 11. Added the smtp_active_hostname option. If used, this will typically be made
5359 to depend on the incoming interface address. Because $interface_address is
5360 not set up until the daemon has forked a reception process, error responses
5361 that can happen earlier (such as "too many connections") no longer contain
5364 12. If an expansion in a condition on a "warn" statement fails because a lookup
5365 defers, the "warn" statement is abandoned, and the next ACL statement is
5366 processed. Previously this caused the whole ACL to be aborted.
5368 13. Added the iplsearch lookup type.
5370 14. Added ident_timeout as a log selector.
5372 15. Added tls_certificate_verified as a log selector.
5374 16. Added a global option tls_require_ciphers (compare the smtp transport
5375 option of the same name). This controls incoming TLS connections.
5377 17. I finally figured out how to make tls_require_ciphers do a similar thing
5378 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5379 before starting the TLS session.
5381 18. Tabs are now shown as \t in -bP output.
5383 19. If the log selector return_path_on_delivery was set, Exim crashed when
5384 bouncing a message because it had too many Received: header lines.
5386 20. If two routers both had headers_remove settings, and the first one included
5387 a superfluous trailing colon, the final name in the first list and the
5388 first name in the second list were incorrectly joined into one item (with a
5389 colon in the middle).
5395 1. Added -C and -D options to the exinext utility, mainly to make it easier
5396 to include in the automated testing, but these could be helpful when
5397 multiple configurations are in use.
5399 2. The exinext utility was not formatting the output nicely when there was
5400 an alternate port involved in the retry record key, nor when there was a
5401 message id as well (for retries that were specific to a specific message
5402 and a specific host). It was also confused by IPv6 addresses, because of
5403 the additional colons they contain. I have fixed the IPv4 problem, and
5404 patched it up to do a reasonable job for IPv6.
5406 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5407 delivery, the log line now contains "pipelined" if PIPELINING was used.
5409 4. An SMTP transport process used to panic and die if the bind() call to set
5410 an explicit outgoing interface failed. This has been changed; it is now
5411 treated in the same way as a connect() failure.
5413 5. A reference to $sender_host_name in the part of a conditional expansion
5414 that was being skipped was still causing a DNS lookup. This no longer
5417 6. The def: expansion condition was not recognizing references to header lines
5418 that used bh_ and bheader_.
5420 7. Added the _cache feature to named lists.
5422 8. The code for checking quota_filecount in the appendfile transport was
5423 allowing one more file than it should have been.
5425 9. For compatibility with Sendmail, the command line option
5433 and sets the incoming protocol and host name (for trusted callers). The
5434 host name and its colon can be omitted when only the protocol is to be set.
5435 Note the Exim already has two private options, -pd and -ps, that refer to
5436 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5437 "s", but I don't think that's a major issue.
5439 10. A number of refactoring changes to the code, none of which should affect
5442 (a) The number of logging options was getting close to filling up the
5443 32-bit word that was used as a bit map. I have split them into two classes:
5444 those that are passed in the argument to log_write(), and those that are
5445 only ever tested independently outside of that function. These are now in
5446 separate 32-bit words, so there is plenty of room for expansion again.
5447 There is no change in the user interface or the logging behaviour.
5449 (b) When building, for example, log lines, the code previously used a
5450 macro that called string_cat() twice, in order to add two strings. This is
5451 not really sufficiently general. Furthermore, there was one instance where
5452 it was actually wrong because one of the argument was used twice, and in
5453 one call a function was used. (As it happened, calling the function twice
5454 did not affect the overall behaviour.) The macro has been replaced by a
5455 function that can join an arbitrary number of extra strings onto a growing
5458 (c) The code for expansion conditions now uses a table and a binary chop
5459 instead of a serial search (which was left over from when there were very
5460 few conditions). Also, it now recognizes conditions like "pam" even when
5461 the relevant support is not compiled in: a suitably worded error message is
5462 given if an attempt is made to use such a condition.
5464 11. Added ${time_interval:xxxxx}.
5466 12. A bug was causing one of the ddress fields not to be passed back correctly
5467 from remote delivery subprocesses. The field in question was not being
5468 subsequently used, so this caused to problems in practice.
5470 13. Added new log selectors queue_time and deliver_time.
5472 14. Might have fixed a bug in maildirsizefile handling that threw up
5473 "unexpected character" debug warnings, and recalculated the data
5474 unnecessarily. In any case, I expanded the warning message to give more
5477 15. Added the message "Restricted characters in address" to the statements in
5478 the default ACL that block characters like @ and % in local parts.
5480 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5481 Three changes have been made:
5483 (a) There was a serious bug; a negative response to MAIL caused the whole
5484 recipient domain to be cached as invalid, thereby blocking all messages
5485 to all local parts at the same domain, from all senders. This bug has
5486 been fixed. The domain is no longer cached after a negative response to
5487 MAIL if the sender used is not empty.
5489 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5492 (c) A new callout option, "use_sender" has been added for people who want
5493 the modified behaviour.
5499 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5502 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5503 indeed breaks things for older releases.
5505 3. Added additional logging to the case where there is a problem reading data
5506 from a filter that is running in a subprocess using a pipe, in order to
5507 try to track down a specific problem.
5509 4. Testing facility fudge: when running in the test harness and attempting
5510 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5511 getting "No route to host". Convert this to a timeout.
5513 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5516 6. Some OS don't have socklen_t but use size_t instead. This affects the
5517 fifth argument of getsockopt() amongst other things. This is now
5518 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5519 can be set for individual OS. I have set it for SunOS5, OSF1, and
5520 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5521 some earlier ones do not.
5523 7. Change 4.30/15 was not doing the test caselessly.
5525 8. The standard form for an IPv6 address literal was being rejected by address
5526 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5527 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5528 this, as well as the form without the "IPv6" on the front (but only when
5529 address literals are enabled, of course).
5531 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5533 10. Exim crashed if a message with an empty sender address specified by -f
5534 encountered a router with an errors_to setting. This could be provoked only
5535 by a command such as
5539 where an empty string was supplied; "<>" did not hit this bug.
5541 11. Installed PCRE release 4.5.
5543 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5544 remained set. It is now erased.
5546 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5547 times from message ids (which are base 36 rather than the normal 62).
5549 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5550 were being counted as actual protocol errors, and logged if the log
5551 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5552 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5553 and DATA following a set of rejected RCPTs do not count as protocol errors.
5554 In other words, Exim assumes they were pipelined, though this may not
5555 actually be the case. Of course, in all cases the client gets an
5556 appropriate error code.
5558 15. If a lookup fails in an ACL condition, a message about the failure may
5559 be available; it is used if testing the ACL cannot continue, because most
5560 such messages specify what the cause of the deferral is. However, some
5561 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5562 that caused an old message to be retained and used if a later statement
5563 caused a defer, replacing the real cause of the deferral.
5565 16. If an IP address had so many PTR records that the DNS lookup buffer
5566 was not large enough to hold them, Exim could crash while trying to process
5567 the truncated data. It now detects and logs this case.
5569 17. Further to 4.21/58, another change has been made: if (and only if) the
5570 first line of a message (the first header line) ends with CRLF, a bare LF
5571 in a subsequent header line has a space inserted after it, so as not to
5572 terminate the header.
5574 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5575 unnecessarily, used atoi() instead of strtol(), and didn't check the
5576 termination when getting file sizes from file names by regex.
5578 19. Completely re-implemented the support for maildirsize files, in the light
5579 of a number of problems with the previous contributed implementation
5580 (4.30/29). In particular:
5582 . If the quota is zero, the maildirsize file is maintained, but no quota is
5585 . If the maildir directory does not exist, it is created before any attempt
5586 to write a maildirsize file.
5588 . The quota value in the file is just a cache; if the quota is changed in
5589 the transport, the new value overrides.
5591 . A regular expression is available for excluding directories from the
5594 20. The autoreply transport checks the characters in options that define the
5595 message's headers; it allows continued headers, but it was checking with
5596 isspace() after an embedded newline instead of explicitly looking for a
5599 21. If all the "regular" hosts to which an address was routed had passed their
5600 expiry times, and had not reached their retry times, the address was
5601 bounced, even if fallback hosts were defined. Now Exim should go on to try
5604 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5605 equivalent code in the SMTP transport. Some hosts send humungous responses
5606 to HELO/EHLO, more than 1024 it seems.
5608 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5609 gives compiler warnings in some environments. I've now done it "properly",
5612 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5613 (because of problems with the built-in one) was declared to return uschar *
5614 instead of char *, causing compiler failure.
5616 25. Fixed a file descriptor leak when processing alias/forward files.
5618 26. Fixed a minor format string issue in dbfn.c.
5620 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5622 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5623 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5626 29. When a sender address is verified, it is cached, to save repeating the test
5627 when there is more than one recipient in a message. However, when the
5628 verification involves a callout, it is possible for different callout
5629 options to be set for different recipients. It is too complicated to keep
5630 track of this in the cache, so now Exim always runs a verification when a
5631 callout is required, relying on the callout cache for the optimization.
5632 The overhead is duplication of the address routing, but this should not be
5635 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5636 to be verified with callout=postmaster, and the main callout worked but the
5637 postmaster check failed, the verification correctly failed. However, if a
5638 subsequent RCPT command asked for sender verification *without* the
5639 postmaster check, incorrect caching caused this verification also to fail,
5642 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5643 it was not caching the DNS options (qualify_single, search_parents) that
5644 were used when the lookup failed. A subsequent lookup with different
5645 options therefore always gave the same answer, though there were cases
5646 where it should not have. (Example: a "domains = !$mx_any" option on a
5647 dnslookup router: the "domains" option is always processed without any
5648 widening, but the router might have qualify_single set.) Now Exim uses the
5649 cached value only when the same options are set.
5651 32. Added John Jetmore's "exipick" utility to the distribution.
5653 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5654 than a timeout (e.g. a certificate is required, and is not provided), an
5655 Exim server now closes the connection immediately. Previously it waited for
5656 the client to close - but if the client is SSL, it seems that they each
5657 wait for each other, leading to a delay before one of them times out.
5659 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5660 maintained 0.8.x compatibility because I don't think many are using it, and
5661 it is clearly obsolete.
5663 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5666 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5667 client certificate was expired. A simple patch fixes this, though I don't
5668 understand the full logic of why the verify callback is called multiple
5671 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5672 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5673 which causes problems with some clients (such as the Certicom SSL Plus
5674 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5675 disables the coutermeasure allowing Eudora to connect."
5677 38. Exim was not checking that a write() to a log file succeeded. This could
5678 lead to Bad Things if a log got too big, in particular if it hit a file
5679 size limit. Exim now panics and dies if it cannot write to a log file, just
5680 as it does if it cannot open a log file.
5682 39. Modified OS/Makefile-Linux so that it now contains
5684 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5686 The two -D definitions ensure that Exim is compiled with large file
5687 support, which makes it possible to handle log files that are bigger than
5690 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5691 instance) a domain was checked against a named list that involved a lookup,
5692 causing $domain_data to be set, then another domain was checked against the
5693 same list, then the first domain was re-checked, the value of $domain_data
5694 after the final check could be wrong. In particular, if the second check
5695 failed, it could be set empty. This bug probably also applied to
5698 41. The strip_trailing_dot option was not being applied to the address given
5699 with the -f command-line option.
5701 42. The code for reading a message's header from the spool was incrementing
5702 $received_count, but never initializing it. This meant that the value was
5703 incorrect (doubled) while delivering a message in the same process in which
5704 it was received. In the most common configuration of Exim, this never
5705 happens - a fresh exec is done - but it can happen when
5706 deliver_drop_privilege is set.
5708 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5709 it now includes up to 150 characters of the unexpected data in the log
5712 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5713 and building data strings. The size of both of these buffers was 10 000
5714 bytes - far larger than anybody would *ever* want, thought I. Needless to
5715 say, somebody hit the limit. I have increased the maximum line length to
5716 20 000 and the maximum data length of concatenated lines to 100 000. I have
5717 also fixed two bugs, because there was no checking on these buffers. Tsk,
5718 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5719 buffer is too small.
5721 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5722 lsearch lookups. Now it does.
5724 46. When parsing a route_list item in a manualroute router, a fixed-length
5725 buffer was used for the list of hosts. I made this 1024 bytes long,
5726 thinking that nobody would ever have a list of hosts that long. Wrong.
5727 Somebody had a whole pile of complicated expansion conditions, and the
5728 string was silently truncated, leading to an expansion error. It turns out
5729 that it is easier to change to an unlimited length (owing to other changes
5730 that have happened since this code was originally written) than to build
5731 structure for giving a limitation error. The length of the item that
5732 expands into the list of hosts is now unlimited.
5734 47. The lsearch lookup could not handle data where the length of text line was
5735 more than 4095 characters. Such lines were truncated, leading to shortened
5736 data being returned. It should now handle lines of any length.
5738 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5739 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5742 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5743 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5744 no longer happens in comment lines. A list of replacements is now placed
5745 at the head of all of the source files, except those whose only change is
5746 to replace PERL_COMMAND in the very first #! line.
5748 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5749 messages on the queue, with a bottom-up merge sort, using code contributed
5750 by Michael Haardt. This should make operations like -bp somewhat faster on
5751 large queues. It won't affect queue runners, except when queue_run_in_order
5754 51. Installed eximstats 1.31 in the distribution.
5756 52. Added support for SRV lookups to the dnslookup router.
5758 53. If an ACL referred to $message_body or $message_body_end, the value was not
5759 reset for any messages that followed in the same SMTP session.
5761 54. The store-handling optimization for building very long strings was not
5762 differentiating between the different store pools. I don't think this
5763 actually made any difference in practice, but I've tidied it.
5765 55. While running the routers to verify a sender address, $sender_address
5766 was still set to the sender address. This is wrong, because when routing to
5767 send a bounce to the sender, it would be empty. Therefore, I have changed
5768 it so that, while verifying a sender address, $sender_address is set to <>.
5769 (There is no change to what happens when verifying a recipient address.)
5771 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5772 target A or AAAA records (if not already returned) without resetting the
5773 qualify_single or search_parents options of the DNS resolver. These are
5774 inappropriate in this case because the targets of MX and SRV records must
5775 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5776 target that, when qualified, matched something in the local domain. These
5777 two options are now turned off when doing these lookups.
5779 57. It seems that at least some releases of Reiserfs (which does not have the
5780 concept of a fixed number of inodes) returns zero and not -1 for the
5781 number of available inodes. This interacted badly with check_spool_inodes,
5782 which assumed that -1 was the "no such thing" setting. What I have done is
5783 to check that the total number of inodes is greater than zero before doing
5784 the test of how many are available.
5786 58. When a "warn" ACL statement has a log_message modifier, the message is
5787 remembered, and not repeated. This is to avoid a lot of repetition when a
5788 message has many recipients that cause the same warning to be written.
5789 However, Exim was preserving the list of already written lines for an
5790 entire SMTP session, which doesn't seem right. The memory is now reset if a
5791 new message is started.
5793 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5794 debugging output unless log_rewrite was also set.
5796 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5797 of (int)(handle) when we know that handle contains (void *)(-1).
5799 61. The Exim daemon panic-logs an error return when it closes the incoming
5800 connection. However "connection reset by peer" seems to be common, and
5801 isn't really an error worthy of noting specially, so that particular error
5804 62. When Exim is trying to find all the local interfaces, it used to panic and
5805 die if the ioctl to get the interface flags failed. However, it seems that
5806 on at least one OS (Solaris 9) it is possible to have an interface that is
5807 included in the list of interfaces, but for which you get a failure error
5808 for this call. This happens when the interface is not "plumbed" into a
5809 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5810 failure of the "get flags" call assumes that the interface is down.
5812 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5813 makes life easier for people who are doing arithmetic on fields extracted
5814 from dates, where you often get leading zeros that should not be
5815 interpreted as octal.
5817 64. Added qualify_domain to the redirect router, to override the global
5820 65. If a pathologically long header line contained very many addresses (the
5821 report of this problem mentioned 10 000) and each of them was rewritten,
5822 Exim could use up a very large amount of memory. (It kept on making new
5823 copies of the header line as it rewrote, and never released the old ones.)
5824 At the expense of a bit more processing, the header rewriting function has
5825 been changed so that it no longer eats memory in this way.
5827 66. The generation of the Received: header has been moved from the time that a
5828 message starts to be received, to the time that it finishes. The timestamp
5829 in the Received: header should now be very close to that of the <= log
5830 line. There are two side-effects of this change:
5832 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5833 logged header lines no longer include the local Received: line, because
5834 it has not yet been created. The same applies to a copy of the message
5835 that is returned to a non-SMTP sender when a message is rejected.
5837 (b) When a filter file is tested using -bf, no additional Received: header
5838 is added to the test message. After some thought, I decided that this
5841 This change does not affect the value of $received_for. It is still set
5842 after address rewriting, but before local_scan() is called.
5844 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5846 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5847 gave an unhelpful panic error message, and a defer error. I have managed to
5848 change this behaviour so that it now rejects any supplied certificate,
5849 which seems right, as the list of acceptable certificates is empty.
5851 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5852 gave an unhelpful defer error. I have not managed to make this reject any
5853 supplied certificates, but the error message it gives is "no certificate
5854 supplied", which is not helpful.
5856 70. exigrep's output now also includes lines that are not associated with any
5857 message, but which match the given pattern. Implemented by a patch from
5858 Martin Sluka, which also tidied up the Perl a bit.
5860 71. Recipient callout verification, like sender verification, was using <> in
5861 the MAIL FROM command. This isn't really the right thing, since the actual
5862 sender may affect whether the remote host accepts the recipient or not. I
5863 have changed it to use the actual sender in the callout; this means that
5864 the cache record is now keyed on a recipient/sender pair, not just the
5865 recipient address. There doesn't seem to be a real danger of callout loops,
5866 since a callout by the remote host to check the sender would use <>.
5867 [SEE ABOVE: changed after hitting problems.]
5869 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5870 temporary errors. However, in the case of such a code being given after
5871 the end of a data transmission (i.e. after ".") Exim was failing to write
5872 a retry record for the message. (Yes, there was some broken host that was
5873 actually sending 8xx at this point.)
5875 73. An unknown lookup type in a host list could cause Exim to panic-die when
5876 the list was checked. (An example that provoked this was putting <; in the
5877 middle of a list instead of at the start.) If this happened during a DATA
5878 ACL check, a -D file could be left lying around. This kind of configuration
5879 error no longer causes Exim to die; instead it causes a defer error. The
5880 incident is still logged to the main and panic logs.
5882 74. Buglet left over from Exim 3 conversion. The message "too many messages
5883 in one connection" was written to the rejectlog but not the mainlog, except
5884 when address rewriting (yes!) was being logged.
5886 75. Added write_rejectlog option.
5888 76. When a system filter was run not as root (that is, when system_filter_user
5889 was set), the values of the $n variables were not being returned to the
5890 main process; thus, they were not subsequently available in the $sn
5893 77. Added +return_path_on_delivery log selector.
5895 78. A connection timeout was being treated differently from recipients deferred
5896 when testing hosts_max_try with a message that was older than the host's
5897 retry timeout. (The host should not be counted, thus allowing all hosts to
5898 be tried at least once before bouncing.) This may have been the cause of an
5899 occasionally reported bug whereby a message would remain on the queue
5900 longer than the retry timeout, but would be bounced if a delivery was
5901 forced. I say "may" because I never totally pinned down the problem;
5902 setting up timeout/retry tests is difficult. See also the next item.
5904 79. The ultimate address timeout was not being applied to errors that involved
5905 a combination of host plus message (for example, a timeout on a MAIL
5906 command). When an address resolved to a number of possible hosts, and they
5907 were not all tried for each delivery (e.g. because of hosts_max_try), a
5908 message could remain on the queue longer than the retry timeout.
5910 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5913 81. Fixed an obscure SMTP outgoing bug which required at least the following
5914 conditions: (a) there was another message waiting for the same server;
5915 (b) the server returned 5xx to all RCPT commands in the first message so
5916 that the message was not completed; (c) the server dropped the connection
5917 or gave a negative response to the RSET that Exim sends to abort the
5918 transaction. The observed case was a dropped connection after DATA that had
5919 been sent in pipelining mode. That is, the server had advertised PIPELINING
5920 but was not implementing it correctly. The effect of the bug was incorrect
5921 behaviour, such as trying another host, and this could lead to a crash.
5927 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5928 and daemon.c were passed as pointers to ints; they should have been
5929 pointers to socklen_t variables (which are typically unsigned ints).
5931 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5934 3. Fixed a really odd bug that affected only the testing scheme; patching a
5935 certain fixed string in the binary changed the value of another string that
5936 happened to be identical to the end of the original first string.
5938 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5939 name", it returns that address as the IP address. On some operating
5940 systems (e.g. Solaris), it also passes back the IP address string as the
5941 "host name". However, on others (e.g. Linux), it passes back an empty
5942 string. Exim wasn't checking for this, and was changing the host name to an
5943 empty string, assuming it had been canonicalized.
5945 5. Although rare, it is permitted to have more than one PTR record for a given
5946 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5947 all the names associated with an address, because they do in Solaris.
5948 However, it seems that they do not in Linux for data that comes from the
5949 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5950 I found this out when I moved to a new Linux workstation and tried to run
5951 the Exim test suite.
5953 To get round this problem I have changed the code so that it now does its
5954 own call to the DNS to look up PTR records when searching for a host name.
5955 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5956 addresses that are only in /etc/hosts are still found.
5958 This behaviour is, however, controlled by an option called host_lookup_
5959 order, which defaults to "bydns:byaddr". If people want to use the other
5960 order, or indeed, just use one or the other means of lookup, they can
5961 specify it in this variable.
5963 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5964 some operating systems, this comes back from gethostbyaddr() as an empty
5965 string, and this is what Exim used to test for. However, it seems that in
5966 other systems, "." is yielded. Exim now tests for this case too.
5968 7. The values of check_spool_space and check_log_space are now held internally
5969 as a number of kilobytes instead of an absolute number of bytes. If a
5970 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5971 kilobyte. This means that much larger values can be stored.
5973 8. Exim monitor: an attempt to get the action menu when not actually pointing
5974 at a message produces an empty menu entitled "No message selected". This
5975 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5976 no entries in it ("Shell widget menu has zero width and/or height"). So I
5977 have added a single, blank menu entry in this case.
5979 9. Added ${quote_local_part.
5981 10. MIME decoding is now applied to the contents of Subject: header lines when
5984 11. Now that a reference to $sender_host_address automatically causes a reverse
5985 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5986 host lookup before query-style lookups in lists that might use this
5987 variable. This has therefore been abolished, and the "net-" prefix is no
5988 longer necessary for query-style lookups.
5990 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5991 have been a typo for LFLAGS, so it has been changed.
5993 13. The install script calls Exim with "-C /dev/null" in order to find the
5994 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5995 to be output. However, since Exim outputs its version number before the
5996 error, it didn't break the script. It just looked ugly. I fixed this by
5997 always allowing "-C /dev/null" if the caller is root.
5999 14. Ignore overlarge ACL variable number when reading spool file - insurance
6000 against a later release with more variables having written the file.
6002 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6003 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6004 this, as well as the form without the "IPv6" on the front.
6006 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6007 OS/Makefile-Darwin file.
6009 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6010 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6012 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6013 variable. However, some OS use ints for inodes. I've added cast to long int
6014 to get rid of the compiler warning.
6016 19. I had forgotten to lock out "/../" in configuration file names when
6017 ALT_CONFIG_PREFIX was set.
6019 20. Routers used for verification do not need to specify transports. However,
6020 if such a router generated a host list, and callout was configured, Exim
6021 crashed, because it could not find a port number from the (non-existent)
6022 transport. It now assumes port 25 in this circumstance.
6024 21. Added the -t option to exigrep.
6026 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6027 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6028 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6029 them is set, LOOKUP_LSEARCH is forced.
6031 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6033 24. Added sender and host information to the "rejected by local_scan()" log
6034 line; previously there was no indication of these.
6036 25. Added .include_if_exists.
6038 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6039 Linux. It turns out that not all file systems support this. Apparently some
6040 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6041 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6042 not supported on the file descriptor, is now ignored when Exim is trying to
6043 sync a directory. This applies only to Linux.
6045 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6047 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6048 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6051 29. Added support for maildirsize files from supplied patch (modified a bit).
6053 30. The use of :fail: followed by an empty string could lead Exim to respond to
6054 sender verification failures with (e.g.):
6056 550 Verification failed for <xxx>
6057 550 Sender verify failed
6059 where the first response line was missing the '-' that indicates it is not
6060 the final line of the response.
6062 31. The loop for finding the name of the user that called Exim had a hardwired
6063 limit of 10; it now uses the value of finduser_retries, which is used for
6064 all other user lookups.
6066 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6069 33. Exim was neglecting to zero errno before one call of strtol() when
6070 expanding a string and expecting an integer value. On some systems this
6071 resulted in spurious "integer overflow" errors. Also, it was casting the
6072 result into an int without checking.
6074 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6075 did not work. The code looks as if it has *never* worked, though it appears
6076 to have been documented since at least release 1.62. I have made it work.
6078 35. The "timeout_DNS" error in retry rules, also documented since at least
6079 1.62, also never worked. As it isn't clear exactly what this means, and
6080 clearly it isn't a major issue, I have abolished the feature by treating it
6081 as "timeout", and writing a warning to the main and panic logs.
6083 36. The display of retry rules for -brt wasn't always showing the error code
6086 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6087 timeout_connect_A, timeout_connect_MX.
6089 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6090 to the empty sender.
6092 39. The daemon was not analysing the content of -oX till after it had closed
6093 stderr and disconnected from the controlling terminal. This meant that any
6094 syntax errors were only noted on the panic log, and the return code from
6095 the command was 0. By re-arranging the code a little, I've made the
6096 decoding happen first, so such errors now appear on stderr, and the return
6097 code is 1. However, the actual setting up of the sockets still happens in
6098 the disconnected process, so errors there are still only recorded on the
6101 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6102 connections (as happens on some IP stacks) was logged at start up time as
6103 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6104 it from "IPv6 and IPv4", which means that two separate sockets are being
6107 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6108 says whether AF_INET or AF_INET6 was passed as an argument.
6110 42. Exiwhat output was messed up when time zones were included in log
6113 43. Exiwhat now gives more information about the daemon's listening ports,
6114 and whether -tls-on-connect was used.
6116 44. The "port" option of the smtp transport is now expanded.
6118 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6119 silently ignored. Now an error message is written to the main and panic
6122 46. There's a new ACL modifier called "logwrite" which writes to a log file
6123 as soon as it is encountered.
6125 47. Added $local_user_uid and $local_user_gid at routing time.
6127 48. Exim crashed when trying to verify a sender address that was being
6130 49. Exim was recognizing only a space character after ".include". It now also
6131 recognizes a tab character.
6133 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6134 extracting the relevant information from the specification. The man page no
6135 longer contains scrambled data for the -d option, and I've added a section
6136 at the front about calling Exim under different names.
6138 51. Added "extra_headers" argument to the "mail" command in filter files.
6140 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6143 53. Installed eximstats 1.29.
6145 54. Added transport_filter_timeout as a generic transport option.
6147 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6148 Cc: header lines. This was required by RFC 822, but it not required by RFC
6151 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6152 incoming messages that did not have them. Now it does so only if the
6153 message originates locally, that is, if there is no associated remote host
6154 address. When Resent- header lines are present, this applies to the Resent-
6155 lines rather than the non-Resent- lines.
6157 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6158 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6160 58. Messages for configuration errors now include the name of the main
6161 configuration file - useful now that there may be more than one file in a
6162 list (.included file names were always shown).
6164 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6165 for those rare installations that do not start the daemon as root or run it
6166 setuid root. I've cut out the call to initgroups() if the daemon is not
6169 60. The Exim user and group can now be bound into the binary as text strings
6170 that are looked up at the start of Exim's processing.
6172 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6174 62. Added $mailstore_basename variable.
6176 63. Installed patch to sieve.c from Michael Haardt.
6178 64. When Exim failed to open the panic log after failing to open the main log,
6179 the original message it was trying to log was written to stderr and debug
6180 output, but if they were not available (the usual case in production), it
6181 was lost. Now it is written to syslog before the two lines that record the
6182 failures to open the logs.
6184 65. Users' Exim filters run in subprocesses under the user's uid. It is
6185 possible for a "deliver" command or an alias in a "personal" command to
6186 provoke an address rewrite. If logging of address rewriting is configured,
6187 this fails because the process is not running as root or exim. There may be
6188 a better way of dealing with this, but for the moment (because 4.30 needs
6189 to be released), I have disabled address rewrite logging when running a
6190 filter in a non-root, non-exim process.
6196 1. The buildconfig auxiliary program wasn't quoting the value set for
6197 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6198 not defined. This bug was masked in 4.22 by the effect that was fixed in
6201 2. Some messages that were rejected after a message id was allocated were
6202 shown as "incomplete" by exigrep. It no longer does this for messages that
6203 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6205 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6206 have allow_domain_literals set, the ID did not get logged in the <= line.
6207 Domain literals are now always recognized in Message-ID: header lines.
6209 4. The first argument for a ${extract expansion item is the key name or field
6210 number. Leading and trailing spaces in this item were not being ignored,
6211 causing some misleading effects.
6213 5. When deliver_drop_privilege was set, single queue runner processes started
6214 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6215 same command in the process it spins off) were not dropping privilege.
6217 6. When the daemon running as "exim" started a queue runner, it always
6218 re-executed Exim in the spun-off process. This is a waste of effort when
6219 deliver_drop_privilege is set. The new process now just calls the
6220 queue-runner function directly.
6226 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6229 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6230 ignored. Though the use of -oP was forcing the writing of a pid file, it
6231 was always written to the default place.
6233 3. If the message "no IP address found for host xxxx" is generated during
6234 incoming verification, it is now followed by identification of the incoming
6235 connection (so you can more easily find what provoked it).
6237 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6239 5. Added some features to "harden" Exim a bit more against certain attacks:
6241 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6242 be put in Local/Makefile. This is like the never_users runtime option,
6243 but it cannot be overridden. The default setting is "root".
6245 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6246 prefix string with which any file named in a -C command line option
6249 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6250 is retained for -C and -D only if the caller of Exim is root. Without
6251 it, the exim user may also use -C and -D and retain privilege.
6253 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6254 command line option is disabled.
6256 6. Macro names set by the -D option must start with an upper case letter, just
6257 like macro names defined in the configuration file.
6259 7. Added "dereference=" facility to LDAP.
6261 8. Two instances of the typo "uknown" in the source files are fixed.
6263 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6264 the Configure-Makefile script screwed up while processing it.
6266 10. Incorporated PCRE 4.4.
6268 11. The SMTP synchronization check was not operating right at the start of an
6269 SMTP session. For example, it could not catch a HELO sent before the client
6270 waited for the greeting. There is now a check for outstanding input at the
6271 point when the greeting is written. Because of the duplex, asynchronous
6272 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6273 way, but not yet received, when the check is performed.
6275 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6276 on TCP/IP sockets, because this apparently causes some broken clients to
6279 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6280 unchanged) from the Cygwin maintainer.
6282 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6283 was supported instead of testing for mbx. Effectively a typo.
6285 15. The spa authenticator server code was not checking that the input it
6286 received was valid base64.
6288 16. The debug output line for the "set" modifier in ACLs was not showing the
6289 name of the variable that was being set.
6291 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6293 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6294 lookup on its own. Something else was needed to trigger the lookup. For
6295 example, a match in host_lookup or the need for a host name in a host list.
6296 Now, if $sender_host_name is referenced and the host name has not yet been
6297 looked up, a lookup is performed. If the lookup fails, the variable remains
6298 empty, and $host_lookup_failed is set to "1".
6300 19. Added "eqi" as a case-independent comparison operator.
6302 20. The saslauthd authentication condition could segfault if neither service
6303 nor realm was specified.
6305 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6306 error message that was logged was misleading, and incoming SMTP
6307 connections were dropped. The message is now more accurate, and temporary
6308 errors are given to SMTP connections.
6310 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6311 (except RSET) with a 421 error, until QUIT is received. However, it was
6312 failing to send a response to QUIT.
6314 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6315 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6316 ACL is now run after the helo_try_verify_hosts code.
6318 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6319 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6320 case-independent, so other case variants are also recognized). Apparently
6321 some systems use these upper case variants.
6323 25. If more than two messages were waiting for the same host, and a transport
6324 filter was specified for the transport, Exim sent two messages over the
6325 same TCP/IP connection, and then failed with "socket operation on non-
6326 socket" when it tried to send the third.
6328 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6330 27. The extern definition of crypt16() in expand.c was not being excluded when
6331 the OS had its own crypt16() function.
6333 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6334 as a preferred synonym for return_size_limit, both as an option and as an
6337 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6339 30. Changed the default configuration ACL to relax the local part checking rule
6340 for addresses that are not in any local domains. For these addresses,
6341 slashes and pipe symbols are allowed within local parts, but the sequence
6342 /../ is explicitly forbidden.
6344 31. SPA server authentication was not clearing the challenge buffer before
6347 32. log_message in a "warn" ACL statement was writing to the reject log as
6348 well as to the main log, which contradicts the documentation and doesn't
6349 seem right (because no rejection is happening). So I have stopped it.
6351 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6352 However, I am unable to do any testing of this.
6354 34. Fixed an infelicity in the appendfile transport. When checking directories
6355 for a mailbox, to see if any needed to be created, it was accidentally
6356 using path names with one or more superfluous leading slashes; tracing
6357 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6359 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6360 contents are added to the log line that is written for every discarded
6361 recipient. (Previously a log_message setting was ignored.)
6363 36. The ${quote: operator now quotes the string if it is empty.
6365 37. The install script runs exim in order to find its version number. If for
6366 some reason other than non-existence or emptiness, which it checks, it
6367 could not run './exim', it was installing it with an empty version number,
6368 i.e. as "exim-". This error state is now caught, and the installation is
6371 38. An argument was missing from the function that creates an error message
6372 when Exim fails to connect to the socket for saslauthd authentication.
6373 This could cause Exim to crash, or give a corrupted message.
6375 39. Added isip, isip4, and isip6 to ${if conditions.
6377 40. The ACL variables $acl_xx are now saved with the message, and can be
6378 accessed later in routers, transports, and filters.
6380 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6381 strings in the file are not string-expanded.
6383 42. If a MAIL command specified a SIZE value that was too large to fit into an
6384 int variable, the check against message_size_limit failed. Such values are
6385 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6386 day this will have to be increased, but I don't think I want to be around
6387 when emails are that large.
6394 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6395 iconv() is not standard in FreeBSD.
6397 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6398 IPv6 enabled. The observed symptom was a segmentation fault on return from
6399 the function os_common_find_running_interfaces() in src/os.c.
6401 3. In the check_special_case() function in daemon.c I had used "errno" as an
6402 argument name, which causes warnings on some systems. This was basically a
6403 typo, since it was named "eno" in the comments!
6405 4. The code that waits for the clock to tick (at a resolution of some fraction
6406 of a second) so as to ensure message-id uniqueness was always waiting for
6407 at least one whole tick, when it could have waited for less. [This is
6408 almost certainly not relevant at current processor speeds, where it is
6409 unlikely to ever wait at all. But we try to future-proof.]
6411 5. The function that sleeps for a time interval that includes fractions of a
6412 second contained a race. It did not block SIGALRM between setting the
6413 timer, and suspending (a couple of lines later). If the interval was short
6414 and the sigsuspend() was delayed until after it had expired, the suspension
6415 never ended. On busy systems this could lead to processes getting stuck for
6418 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6419 process, before it forks any delivery processes. The open lookup caching
6420 mechanism meant that the open file or database connection was passed into
6421 the delivery process. The problem was that delivery processes always tidy
6422 up cached lookup data. This could cause a problem for the next delivery
6423 process started by the queue runner, because the external queue runner
6424 process does not know about the closure. So the next delivery process
6425 still has data in the lookup cache. In the case of a file lookup, there was
6426 no problem because closing a file descriptor in a subprocess doesn't affect
6427 the parent. However, if the lookup was caching a connection to a database,
6428 the connection was closed, and the second delivery process was likely to
6429 see errors such as "PGSQL: query failed: server closed the connection
6430 unexpectedly". The problem has been fixed by closing all cached lookups
6431 in a queue runner before running a delivery process.
6433 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6434 seem to have the "const" qualifier which it has on other OS. I've
6437 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6438 *of the same type* (client or server) with the same public name that an
6439 error should be diagnosed.
6441 9. When Exim looked up a host name for an IP address, but failed to find the
6442 original IP address when looking up the host name (a safety check), it
6443 output the message "<ip address> does not match any IP for NULL", which was
6444 confusing, to say the least. The bug was that the host name should have
6445 appeared instead of "NULL".
6447 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6448 user that is built into the binary, and the -C or -D options is used, root
6449 privilege is dropped before the configuration file is read. In addition,
6450 logging is switched to stderr instead of the normal log files. If the
6451 configuration then re-defines the Exim user, the unprivileged environment
6452 is probably not what is expected, so Exim logs a panic warning message (but
6455 However, if deliver_drop_privilege is set, the unprivileged state may well
6456 be exactly what is intended, so the warning has been cut out in that case,
6457 and Exim is allowed to try to write to its normal log files.
6463 1. smtp_return_error_details was not giving details for temporary sender
6464 or receiver verification errors.
6466 2. Diagnose a configuration error if two authenticators have the same public
6469 3. Exim used not to create the message log file for a message until the first
6470 delivery attempt. This could be confusing when incoming messages were held
6471 for policy or load reasons. The message log file is now created at the time
6472 the message is received, and an initial "Received" line is written to it.
6474 4. The automatically generated man page for command line options had a minor
6475 bug that caused no ill effects; however, a more serious problem was that
6476 the procedure for building the man page automatically didn't always
6477 operate. Consequently, release 4.20 contains an out-of-date version. This
6478 shouldn't happen again.
6480 5. When building Exim with embedded Perl support, the script that builds the
6481 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6482 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6484 6. The freeze_tell option was not being used for messages that were frozen on
6485 arrival, either by an ACL or by local_scan().
6487 7. Added the smtp_incomplete_transaction log selector.
6489 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6490 was accepting AUTH without a new EHLO.
6492 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6493 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6496 10. Logging of TCP/IP connections (when configured) now happens in the main
6497 daemon process instead of the child process, so that the TCP/IP connection
6498 count is more accurate (but it can never be perfect).
6500 11. The use of "drop" in a nested ACL was not being handled correctly in the
6501 outer ACL. Now, if condition failure induced by the nested "drop" causes
6502 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6503 or "require"), the connection is dropped.
6505 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6506 that yield "discard" can now be used with an "accept" or a "discard" verb,
6507 but an error is generated for any others (because I can't see a useful way
6508 to define what should happen).
6510 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6511 are now processed in the same way as lines in the Exim configuration file.
6512 In particular, continuation lines are supported.
6514 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6516 15. Added -ti meaning -t -i.
6518 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6519 domains, and warn by logging if others are found.
6521 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6522 returned from GIFCONF when Exim is trying to find the list of interfaces on
6523 a host. The code in os.c has been modified to copy each ifreq to an aligned
6524 structure in all cases.
6526 Also, in some cases, the returned ifreq's were being copied to a 'struct
6527 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6528 means the last couple of bytes of an IPv6 address could be chopped if the
6529 ifreq contained only a normal sockaddr (14 bytes storage).
6531 18. Named domain lists were not supported in the hosts_treat_as_local option.
6532 An entry such as +xxxx was not recognized, and was treated as a literal
6535 19. Ensure that header lines added by a DATA ACL are included in the reject log
6536 if the ACL subsequently rejects the message.
6538 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6539 MD5 (which is deprecated).
6541 21. When testing a filter file using -bf, Exim was writing a message when it
6542 took the sender from a "From " line in the message, but it was not doing so
6543 when it took $return_path from a Return-Path: header line. It now does.
6545 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6546 with a valid header line field name (a series of printing characters
6547 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6549 23. Changed "disc" in the source to "disk" to conform to the documentation and
6550 the book and for uniformity.
6552 24. Ignore Sendmail's -Ooption=value command line item.
6554 25. When execve() failed while trying to run a command in a pipe transport,
6555 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6556 could be confused with a return value of 69 from the command itself. This
6557 has been changed to 127, the value the shell returns if it is asked to run
6558 a non-existent command. The wording for the related log line suggests a
6559 non-existent command as the problem.
6561 26. If received_header_text expands to an empty string, do not add a Received:
6562 header line to the message. (Well, it adds a token one on the spool, but
6563 marks it "old" so that it doesn't get used or transmitted.)
6565 27. Installed eximstats 1.28 (addition of -nt option).
6567 28. There was no check for failure on the call to getsockname() in the daemon
6568 code. This can fail if there is a shortage of resources on the system, with
6569 ENOMEM, for example. A temporary error is now given on failure.
6571 29. Contrary to the C standard, it seems that in some environments, the
6572 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6573 program. Exim now does this explicitly; it affects the formatting of
6574 timestamps using strftime().
6576 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6577 complaints. I've now initialized all the variables, to avoid this.
6579 32. Header lines added by a system filter were not being "seen" during
6580 transport-time rewrites.
6582 33. The info_callback() function passed to OpenSSL is set up with type void
6583 (*)(SSL *, int, int), as described somewhere. However, when calling the
6584 function (actually a macro) that sets it up, the type void(*)() is
6585 expected. I've put in a cast to prevent warnings from picky compilers.
6587 34. If a DNS black list lookup found a CNAME record, but there were no A
6588 records associated with the domain it pointed at, Exim crashed.
6590 35. If a DNS black list lookup returned more than one A record, Exim ignored
6591 all but the first. It now scans all returned addresses if a particular IP
6592 value is being sought. In this situation, the contents of the
6593 $dnslist_value variable are a list of all the addresses, separated by a
6596 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6597 to accept a host name and all its aliases if the forward lookup for any of
6598 them yielded the IP address of the incoming connection. Now it accepts only
6599 those names whose forward lookup yields the correct IP address. Any other
6600 names are discarded. This closes a loophole whereby a rogue DNS
6601 administrator could create reverse DNS records to break through a
6602 wildcarded host restriction in an ACL.
6604 37. If a user filter or a system filter that ran in a subprocess used any of
6605 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6606 the wrong values were passed to the pipe command ($thisaddress had the
6607 value of $0, $0 had the value of $1, etc). This bug was introduced by
6608 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6610 38. Improved the line breaking for long SMTP error messages from ACLs.
6611 Previously, if there was no break point between 40 and 75 characters, Exim
6612 left the rest of the message alone. Two changes have been made: (a) I've
6613 reduced the minimum length to 35 characters; (b) if it can't find a break
6614 point between 35 and 75 characters, it looks ahead and uses the first one
6615 that it finds. This may give the occasional overlong line, but at least the
6616 remaining text gets split now.
6618 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6619 file descriptors might be low, and that setting 1000 would always raise it.
6620 It turns out that in some environments, the limit is already over 1000 and
6621 that lowering it causes trouble. So now Exim takes care not to decrease it.
6623 40. When delivering a message, the value of $return_path is set to $sender_
6624 address at the start of routing (routers may change the value). By an
6625 oversight, this default was not being set up when an address was tested by
6626 -bt or -bv, which affected the outcome if any router or filter referred to
6629 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6630 log only when "message" or "log_message" are set. However, if one of the
6631 conditions was an address verification, or a call to a nested ACL, the
6632 messages generated by the underlying test were being passed through. This
6633 no longer happens. The underlying message is available in $acl_verify_
6634 message for both "message" and "log_message" expansions, so it can be
6635 passed through if needed.
6637 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6638 new expansion $bh_ to give the encoded byte string without charset
6639 translation. Translation happens only if iconv() is available; HAVE_ICONV
6640 indicates this at build time. HEADERS_CHARSET gives the charset to
6641 translate to; headers_charset can change it in the configuration, and
6642 "headers charset" can change it in an individual filter file.
6644 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6645 that creates RFC 2047 encoded "words" labels them as that charset instead
6646 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6647 expansion operator; (ii) when Exim creates a From: line for a local
6648 message; (iii) when a header line is rewritten to include a "phrase" part.
6650 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6651 buggy, causing it to skip the first lines of messages whose message ID
6652 ended in 'D'. This would not have bitten before Exim release 4.14, because
6653 message IDs were unlikely to end in 'D' before then. The effect was to have
6654 incorrect size information for certain domains.
6656 45. #include "config.h" was missing at the start of the crypt16.c module. This
6657 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6660 46. If there was a timeout during a "random" callout check, Exim treated it as
6661 a failure of the random address, and carried on sending RSET and the real
6662 address. If the delay was just some slowness somewhere, the response to the
6663 original RCPT would be taken as a response to RSET and so on, causing
6664 mayhem of various kinds.
6666 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6667 when I implemented it. It didn't allow for the fact that some option values
6668 may legitimately be negative (e.g. size_addition), and it didn't even do
6669 the right test for positive values.
6671 48. Domain names in DNS records are case-independent. Exim always looks them up
6672 in lower case. Some resolvers return domain names in exactly the case they
6673 appear in the zone file, that is, they may contain uppercase letters. Not
6674 all resolvers do this - some return always lower case. Exim was treating a
6675 change of case by a resolver as a change of domain, similar to a widening
6676 of a domain abbreviation. This triggered its re-routing code and so it was
6677 trying to route what was effectively the same domain again. This normally
6678 caused routing to fail (because the router wouldn't handle the domain
6679 twice). Now Exim checks for this case specially, and just changes the
6680 casing of the domain that it ultimately uses when it transmits the message
6683 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6686 50. If a filter generated a file delivery with a non-absolute name (possible if
6687 no home directory exists for the router), the forbid_file option was not
6690 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6691 the existing equality matching.
6693 52. Exim was using ints instead of ino_t variables in some places where it was
6694 dealing with inode numbers.
6696 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6697 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6698 called TMPDIR, and if it finds it is different, it changes its value.
6700 54. The smtp_printf() function is now made available to local_scan() so
6701 additional output lines can be written before returning. There is also an
6702 smtp_fflush() function to enable the detection of a dropped connection.
6703 The variables smtp_input and smtp_batched_input are exported to
6706 55. Changed the default runtime configuration: the message "Unknown user"
6707 has been removed from the ACL, and instead placed on the localuser router,
6708 using the cannot_route_message feature. This means that any verification
6709 failures that generate their own messages won't get overridden. Similarly,
6710 the "Unrouteable address" message that was in the ACL for unverifiable
6711 relay addresses has also been removed.
6713 56. Added hosts_avoid_esmtp to the smtp transport.
6715 57. The exicyclog script was not checking for the esoteric option
6716 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6717 will work only if exicyclog is run under the appropriate euid.
6719 58. Following a discussion on the list, the rules by which Exim recognises line
6720 endings on incoming messages have been changed. The -dropcr and drop_cr
6721 options are now no-ops, retained only for backwards compatibility. The
6722 following line terminators are recognized: LF CRLF CR. However, special
6723 processing applies to CR:
6725 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6726 nor a local message in the state where . is a terminator.
6728 (ii) If a bare CR is encountered in a header line, an extra space is added
6729 after the line terminator so as not to end the header. The reasoning
6730 behind this is that bare CRs in header lines are most likely either
6731 to be mistakes, or people trying to play silly games.
6733 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6734 was being incorrectly given as 18 bytes larger than it should have been.
6735 This is a VOB (very old bug).
6737 60. This may never have affected anything current, but just in case it has:
6738 When the local host is found other than at the start of a list of hosts,
6739 the local host, those with the same MX, and any that follow, are discarded.
6740 When the list in question was part of a longer list of hosts, the following
6741 hosts (not currently being processed) were also being discarded. This no
6742 longer happens. I'm not sure if this situation could ever has previously
6745 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6748 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6749 header. This is something that is recommended in a new Internet Draft, and
6750 is something that is documented as being done by Sendmail. There are two
6751 possible values. For messages generated by the autoreply transport, Exim
6754 Auto-Submitted: auto-replied
6756 whereas for all other generated messages (e.g. bounces) it adds
6758 Auto-Submitted: auto-generated
6760 63. The "personal" condition in filters now includes a test for the
6761 Auto-Submitted: header. If it contains the string "auto-" the message it
6762 not considered personal.
6764 64. Added rcpt_include_affixes as a generic transport option.
6766 65. Added queue_only_override (default true).
6768 66. Added the syslog_duplication option.
6770 67. If what should have been the first header line of a message consisted of
6771 a space followed by a colon, Exim was mis-interpreting it as a header line.
6772 It isn't of course - it is syntactically invalid and should therefore be
6773 treated as the start of the message body. The misbehaviour could have
6774 caused a number of strange effects, including loss of data in subsequent
6775 header lines, and spool format errors.
6777 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6778 client host had authenticated. This control can now be exercised by an ACL
6779 for more flexibility.
6781 69. By default, callouts do not happen when testing with -bh. There is now a
6782 variant, -bhc, which does actually run the callout code, including
6783 consulting and updating the callout cache.
6785 70. Added support for saslauthd authentication, courtesy of Alexander
6788 71. If statvfs() failed on the spool or log directories while checking their
6789 size for availability, Exim confusingly gave the error "space shortage".
6790 Furthermore, in debugging mode it crashed with a floating point exception.
6791 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6792 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6793 serious problem, Exim now writes to the main and panic logs when this
6794 happens, with details of the failure. It then refuses to accept the
6795 incoming message, giving the message "spool directory problem" or "log
6796 directory problem" with a 421 code for SMTP messages.
6798 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6799 0, 1, and 2 exist, because some OS complain for execs without them (see
6800 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6801 descriptors. However, the code omitted to check that the open succeeded,
6802 causing mysterious errors if for some reason the permissions on /dev/null
6803 got screwed. Now Exim writes a message to the main and panic logs, and
6804 bombs out if it can't open /dev/null.
6806 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6807 interact so that it is all more flexible. It is supposed to remain
6808 backwards compatible. Also added extra_local_interfaces.
6810 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6811 to bomb out with an assertion failure - to the client this appears as a
6812 connection drop. This problem occurs in the part of the code that was taken
6813 from the Samba project. Fortunately, the assertion is in a very simple
6814 function, so I have fixed this by reproducing the function inline in the
6815 one place where it is called, and arranging for authentication to fail
6816 instead of killing the process with assert().
6818 75. The SPA client code was not working when the server requested OEM rather
6819 than Unicode encoding.
6821 76. Added code to make require_files with a specific uid setting more usable in
6822 the case where statting the file as root fails - usually a non-root-mounted
6823 NFS file system. When this happens and the failure is EACCES, Exim now
6824 forks a subprocess and does the per-uid checking as the relevant uid.
6826 77. Added process_log_path.
6828 78. If log_file_path was not explicitly set, a setting of check_log_space or
6829 check_log_inodes was ignored.
6831 79. If a space check for the spool or log partitions fails, the incident is now
6832 logged. Of course, in the latter case the data may get lost...
6834 80. Added the %p formatting code to string_format() so that it can be used to
6835 print addresses in debug_print(). Adjusted all the address printing in the
6836 debugging in store.c to use %p rather than %d.
6838 81. There was a concern that a line of code in smtp_in.c could overflow a
6839 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6840 initially expressed, the concern was not well-founded, because trailing
6841 spaces are removed early. However, if the trailing spaces were followed by
6842 a NULL, they did not get removed, so the overflow was possible. Two fixes
6845 (a) I re-wrote the offending code in a cleaner fashion.
6846 (b) If an incoming SMTP command contains a NULL character, it is rejected
6849 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6850 runs initgroups(), so that if the Exim user is in any additional groups,
6851 they will be used during message reception.
6857 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob