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
101 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
102 cmdline element to get the documented default one. Previously it was
105 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
106 are unsuable for non-message-specific reasons - eg. connection timeout,
107 banner-time rejection.
109 JH/16 Fix logging of delivery remote port, when specified by router, under
112 PP/06 Repair manualroute's ability to take options in any order, even if one
113 is the name of a transport.
116 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
118 JH/17 Change the list-building routines interface to use the expanding-string
119 triplet model, for better allocation and copying behaviour.
125 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
126 than -2003 did; needs libidn2 in addition to libidn.
128 JH/02 The path option on a pipe transport is now expanded before use.
130 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
131 Patch provided by "Björn", documentation fix added too.
133 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
134 missing a wire-to-host endian conversion.
136 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
137 close after a BDAT command line could be taken as a following command,
138 giving a synch failure. Fix by only checking for synch immediately
139 before acknowledging the chunk.
141 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
142 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
143 macro. Patches provided by Josh Soref.
145 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
146 Previously we did not; the RFC seems ambiguous and VRFY is not listed
147 by IANA as a service extension. However, John Klensin suggests that we
150 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
151 the dkim code may be unix-mode line endings rather than smtp wire-format
152 CRLF, so prepend a CR to any bare LF.
154 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
155 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
157 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
158 the first were themselves being wrongly included in the feed into dkim
159 processing; with most chunk sizes in use this resulted in an incorrect
160 body hash calculated value.
162 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
163 DKIM signature block, for verification. Although advised against by
164 standards it is specifically not ruled illegal.
166 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
168 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
169 missing a body hash (the bh= tag).
171 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
172 It seems that HAProxy sends the Proxy Protocol information in clear and
173 only then does a TLS startup, so do the same.
175 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
176 TCP connections (such as for Spamd) unless the daemon successfully set
177 Fast Open mode on its listening sockets. This fixes breakage seen on
178 too-old kernels or those not configured for Fast Open, at the cost of
179 requiring both directions being enabled for TFO, and TFO never being used
180 by non-daemon-related Exim processes.
182 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
183 endings, at least on the first header line. Try to canonify any that get
184 past that check, despite the cost.
186 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
187 now limited to an arbitrary five deep, while parsing addresses with the
188 strip_excess_angle_brackets option enabled.
190 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
191 instead leave the unprompted TLS handshake in socket buffer for the
192 TLS library to consume.
194 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
196 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
198 JH/16 Drop variables when they go out of scope. Memory management drops a whole
199 region in one operation, for speed, and this leaves assigned pointers
200 dangling. Add checks run only under the testsuite which checks all
201 variables at a store-reset and panics on a dangling pointer; add code
202 explicitly nulling out all the variables discovered. Fixes one known
203 bug: a transport crash, where a dangling pointer for $sending_ip_address
204 originally assigned in a verify callout, is re-used.
206 PP/06 Drop '.' from @INC in various Perl scripts.
208 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
210 PP/08 Reduce a number of compilation warnings under clang; building with
211 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
212 should be warning-free.
214 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
216 HS/01 Fix portability problems introduced by PP/08 for platforms where
217 realloc(NULL) is not equivalent to malloc() [SunOS et al].
219 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
220 chunk. This allows us to accept broken chunked messages. We need a more
221 general solution here.
223 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
224 already-broken messages in the queue.
226 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
228 JH/19 Fix reference counting bug in routing-generated-address tracking.
234 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
235 supports it and a size is available (ie. the sending peer gave us one).
237 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
238 years of being deprecated). The replacements are the ACLs
239 acl_smtp_mime and acl_not_smtp_mime.
241 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
242 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
243 or even an in-clear connection were permitted. Now, if the host lookup
244 was dnssec and dane was requested then the host is only used if the
245 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
246 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
247 if one fails this test.
248 This means that a poorly-configured remote DNS will make it incommunicado;
249 but it protects against a DNS-interception attack on it.
251 JH/04 Bug 1810: make continued-use of an open smtp transport connection
252 non-noisy when a race steals the message being considered.
254 JH/05 If main configuration option tls_certificate is unset, generate a
255 self-signed certificate for inbound TLS connections.
257 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
258 in rewrites and routers.
260 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
261 and logged a warning sing 4.83; now they are a configuration file error.
263 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
264 (lacking @domain). Apply the same qualification processing as RCPT.
266 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
268 JH/10 Support ${sha256:} applied to a string (as well as the previous
271 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
272 a cutthrough deliver is pending, as we always want to make a connection.
273 This also avoids re-routing the message when later placing the cutthrough
274 connection after a verify cache hit.
275 Do not update it with the verify result either.
277 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
278 when routing results in more than one destination address.
280 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
281 signing (which inhibits the cutthrough capability). Previously only
282 the presence of an option was tested; now an expansion evaluating as
283 empty is permissible (obviously it should depend only on data available
284 when the cutthrough connection is made).
286 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
287 the relevant preceding SMTP command did not note the pipelining mode.
289 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
290 Previously they were not counted.
292 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
293 as one having no matching records. Previously we deferred the message
294 that needed the lookup.
296 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
297 distinguished as "(=".
299 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
300 for missing MX records. Previously it only worked for missing A records.
302 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
304 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
305 after the data-go-ahead and data-ack. Patch from Jason Betts.
307 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
308 even for a "none" policy. Patch from Tony Meyer.
310 JH/22 Fix continued use of a connection for further deliveries. If a port was
311 specified by a router, it must also match for the delivery to be
314 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
315 When one of the recipients of a message was unsuitable for the connection
316 (has no matching addresses), we lost track of needing to mark it
317 deferred. As a result mail would be lost.
319 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
321 JH/25 Decoding ACL controls is now done using a binary search; the source code
322 takes up less space and should be simpler to maintain. Merge the ACL
323 condition decode tables also, with similar effect.
325 JH/26 Fix problem with one_time used on a redirect router which returned the
326 parent address unchanged. A retry would see the parent address marked as
327 delivered, so not attempt the (identical) child. As a result mail would
330 JH/27 Fix a possible security hole, wherein a process operating with the Exim
331 UID can gain a root shell. Credit to http://www.halfdog.net/ for
332 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
335 JH/28 Enable {spool,log} filesystem space and inode checks as default.
336 Main config options check_{log,spool}_{inodes,space} are now
337 100 inodes, 10MB unless set otherwise in the configuration.
339 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
340 Previously it only applied to the main-section connection policy
343 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
345 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
346 by me. Added RFC7919 DH primes as an alternative.
348 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
349 are not in the system include path.
351 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
352 GnuTLS, when a session startup failed (eg because the client disconnected)
353 Exim did stdio operations after fclose. This was exposed by a recent
354 change which nulled out the file handle after the fclose.
356 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
357 signed directly by the cert-signing cert, rather than an intermediate
358 OCSP-signing cert. This is the model used by LetsEncrypt.
360 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
362 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
363 an incoming connection.
365 HS/02 Bug 1802: Do not half-close the connection after sending a request
368 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
369 fallback to "prime256v1".
371 JH/34 SECURITY: Use proper copy of DATA command in error message.
372 Could leak key material. Remotely exploitable. CVE-2016-9963.
378 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
379 and 3.4.4 - once the server is enabled to respond to an OCSP request
380 it does even when not requested, resulting in a stapling non-aware
381 client dropping the TLS connection.
383 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
384 support variable-length bit vectors. No functional change.
386 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
387 The I= interface field on outgoing lines is now after the H= remote
388 host field, same as incoming lines. There is a separate
389 outgoing_interface log selector which allows you to disable the
392 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
393 If not running log_selector +smtp_connection the mainlog would be held
394 open indefinitely after a "too many connections" event, including to a
395 deleted file after a log rotate. Leave the per net connection logging
396 leaving it open for efficiency as that will be quickly detected by the
397 check on the next write.
399 HS/01 Bug 1671: Fix post transport crash.
400 Processing the wait-<transport> messages could crash the delivery
401 process if the message IDs didn't exist for some reason. When
402 using 'split_spool_directory=yes' the construction of the spool
403 file name failed already, exposing the same netto behaviour.
405 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
406 mime_regex ACL conditions.
408 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
409 to DSN fail messages (bounces): remote IP, remote greeting, remote response
410 to HELO, local diagnostic string.
412 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
413 log line to debug. Even when configured with a tls authenticator many
414 client connections are expected to not authenticate in this way, so
415 an authenticate fail is not an error.
417 HS/02 Add the Exim version string to the process info. This way exiwhat
418 gives some more detail about the running daemon.
420 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
421 matter for fast-change records such as DNSBLs.
423 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
424 retry record, even if constant. There may be multiple transports with
425 different interface settings and the retry behaviour needs to be kept
428 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
430 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
432 JH/10 Bug 840: fix log_defer_output option of pipe transport
434 JH/11 Bug 830: use same host for all RCPTS of a message, even under
435 hosts_randomize. This matters a lot when combined with mua_wrapper.
437 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
438 ${quote_pgsql:<string>} operator.
440 JH/13 Bug 1708: avoid misaligned access in cached lookup.
442 JH/14 Change header file name for freeradius-client. Relevant if compiling
443 with Radius support; from the Gentoo tree and checked under Fedora.
445 JH/15 Bug 1712: Introduce $prdr_requested flag variable
447 JH/16 Bug 1714: Permit an empty string as expansion result for transport
448 option transport_filter, meaning no filtering.
450 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
452 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
453 defaults to "*" (all hosts). The variable is now available when not built
454 with TLS, default unset, mainly to enable keeping the testsuite sane.
455 If a server certificate is not supplied (via tls_certificate) an error is
456 logged, and clients will find TLS connections fail on startup. Presumably
457 they will retry in-clear.
458 Packagers of Exim are strongly encouraged to create a server certificate
459 at installation time.
461 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
462 with the $config_file variable.
464 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
465 in transport context, after the attempt, and per-recipient. The latter type
466 is per host attempted. The event data is the error message, and the errno
467 information encodes the lookup type (A vs. MX) used for the (first) host,
468 and the trailing two digits of the smtp 4xx response.
470 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
471 to write to mainlog (or rejectlog, paniclog) in the window between file
472 creation and permissions/ownership being changed. Particularly affects
473 installations where exicyclog is run as root, rather than exim user;
474 result is that the running daemon panics and dies.
476 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
478 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
479 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
480 "pri" and "weight". Note that the previous implicit priority given by the
481 list order is no longer honoured.
483 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
486 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
487 by defining SUPPORT_SOCKS.
489 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
490 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
491 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
492 variables are renamed to proxy_{local,external}_{address,port}.
494 JH/27 Move Internationalisation support from Experimental to mainline, enabled
495 for a build by defining SUPPORT_I18N
497 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
498 of the query string, and make ${quote_redis:} do that quoting.
500 JH/29 Move Events support from Experimental to mainline, enabled by default
501 and removable for a build by defining DISABLE_EVENT.
503 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
505 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
506 cached by the daemon.
508 JH/32 Move Redis support from Experimental to mainline, enabled for a build
509 by defining LOOKUP_REDIS. The libhiredis library is required.
511 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
512 keys are given for lookup.
514 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
515 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
516 only supported when built with TLS support. The PolarSSL SHA routines
517 are still used when the TLS library is too old for convenient support.
519 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
520 openssl_options), for security. OpenSSL forces this from version 1.1.0
521 server-side so match that on older versions.
523 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
524 allocation for $value could be released as the expansion processing
525 concluded, but leaving the global pointer active for it.
527 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
528 and to use the domains and local_parts ACL conditions.
530 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
531 incorrectly not doubled on cutthrough transmission, hence seen as a
532 body-termination at the receiving system - resulting in truncated mails.
533 Commonly the sender saw a TCP-level error, and retransmitted the message
534 via the normal store-and-forward channel. This could result in duplicates
535 received - but deduplicating mailstores were liable to retain only the
536 initial truncated version.
538 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
540 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
542 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
543 we're in there, support oversigning also; bug 1309.
545 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
547 HS/04 Add support for keep_environment and add_environment options.
549 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
550 either intentional arithmetic overflow during PRNG, or testing config-
553 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
554 delivery resulted in actual delivery. Cancel cutthrough before DATA
557 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
558 rejects a recipient: pass the reject to the originator.
560 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
561 Many were false-positives and ignorable, but it's worth fixing the
564 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
565 for the new environment-manipulation done at startup. Move the routines
566 from being local to tls.c to being global via the os.c file.
568 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
569 an extract embedded as result-arg for a map, the first arg for extract
570 is unavailable so we cannot tell if this is a numbered or keyed
571 extraction. Accept either.
577 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
580 JH/02 The smtp transport option "multi_domain" is now expanded.
582 JH/03 The smtp transport now requests PRDR by default, if the server offers
585 JH/04 Certificate name checking on server certificates, when exim is a client,
586 is now done by default. The transport option tls_verify_cert_hostnames
587 can be used to disable this per-host. The build option
588 EXPERIMENTAL_CERTNAMES is withdrawn.
590 JH/05 The value of the tls_verify_certificates smtp transport and main options
591 default to the word "system" to access the system default CA bundle.
592 For GnuTLS, only version 3.0.20 or later.
594 JH/06 Verification of the server certificate for a TLS connection is now tried
595 (but not required) by default. The verification status is now logged by
596 default, for both outbound TLS and client-certificate supplying inbound
599 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
602 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
603 Status Notification (bounce) messages are now MIME format per RFC 3464.
604 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
605 under the control of the dsn_advertise_hosts option, and routers may
606 have a dsn_lasthop option.
608 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
609 default, modifiable by a malware= option. The list separator for
610 the options can now be changed in the usual way. Bug 68.
612 JH/10 The smtp_receive_timeout main option is now expanded before use.
614 JH/11 The incoming_interface log option now also enables logging of the
615 local interface on delivery outgoing connections.
617 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
618 if the interface and destination host and port all match.
620 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
623 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
624 Patch from Andrew Lewis.
626 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
627 now supports optional time-restrictions, weighting, and priority
628 modifiers per server. Patch originally by <rommer@active.by>.
630 JH/16 The spamd_address main option now supports a mixed list of local
631 and remote servers. Remote servers can be IPv6 addresses, and
632 specify a port-range.
634 JH/17 Bug 68: The spamd_address main option now supports an optional
635 timeout value per server.
637 JH/18 Bug 1581: Router and transport options headers_add/remove can
638 now have the list separator specified.
640 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
643 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
646 JH/21 Support for the A6 type of dns record is withdrawn.
648 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
649 rather than the verbs used.
651 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
652 from 255 to 1024 chars.
654 JH/24 Verification callouts now attempt to use TLS by default.
656 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
657 are generic router options now. The defaults didn't change.
659 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
660 Original patch from Alexander Shikoff, worked over by JH.
662 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
663 files if ZCAT_COMMAND is not executable.
665 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
667 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
669 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
670 Normally benign, it bites when the pair was led to by a CNAME;
671 modern usage is to not canonicalize the domain to a CNAME target
672 (and we were inconsistent anyway for A-only vs AAAA+A).
674 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
676 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
677 when evaluating $sender_host_dnssec.
679 JH/31 Check the HELO verification lookup for DNSSEC, adding new
680 $sender_helo_dnssec variable.
682 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
684 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
686 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
688 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
689 documented as working, but never had. Support all but $spam_report.
691 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
692 added for tls authenticator.
694 HS/03 Add perl_taintmode main config option
700 TL/01 When running the test suite, the README says that variables such as
701 no_msglog_check are global and can be placed anywhere in a specific
702 test's script, however it was observed that placement needed to be near
703 the beginning for it to behave that way. Changed the runtest perl
704 script to read through the entire script once to detect and set these
705 variables, reset to the beginning of the script, and then run through
706 the script parsing/test process like normal.
708 TL/02 The BSD's have an arc4random API. One of the functions to induce
709 adding randomness was arc4random_stir(), but it has been removed in
710 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
711 function when detected.
713 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
714 cause callback expansion.
716 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
717 syntax errors in an expansion can be treated as a string instead of
718 logging or causing an error, due to the internal use of bool_lax
719 instead of bool when processing it.
721 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
722 server certificates when making smtp deliveries.
724 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
726 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
728 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
730 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
731 Merged patch from Sebastian Wiedenroth.
733 JH/05 Fix results-pipe from transport process. Several recipients, combined
734 with certificate use, exposed issues where response data items split
735 over buffer boundaries were not parsed properly. This eventually
736 resulted in duplicates being sent. This issue only became common enough
737 to notice due to the introduction of connection certificate information,
738 the item size being so much larger. Found and fixed by Wolfgang Breyha.
740 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
741 size buffer was used, resulting in syntax errors when an expansion
744 JH/07 Add support for directories of certificates when compiled with a GnuTLS
745 version 3.3.6 or later.
747 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
748 is EXPERIMENTAL_EVENT, the main-configuration and transport options
749 both become "event_action", the variables become $event_name, $event_data
750 and $event_defer_errno. There is a new variable $verify_mode, usable in
751 routers, transports and related events. The tls:cert event is now also
752 raised for inbound connections, if the main configuration event_action
755 TL/06 In test suite, disable OCSP for old versions of openssl which contained
756 early OCSP support, but no stapling (appears to be less than 1.0.0).
758 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
759 server certificate names available under the smtp transport option
760 "tls_verify_cert_hostname" now do not permit multi-component wildcard
763 JH/10 Time-related extraction expansions from certificates now use the main
764 option "timezone" setting for output formatting, and are consistent
765 between OpenSSL and GnuTLS compilations. Bug 1541.
767 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
768 encoded parameter in the incoming message. Bug 1558.
770 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
771 include certificate info, eximon was claiming there were spoolfile
774 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
776 JH/14 Log delivery-related information more consistently, using the sequence
777 "H=<name> [<ip>]" wherever possible.
779 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
780 are problematic for Debian distribution, omit them from the release
783 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
785 JH/16 Fix string representation of time values on 64bit time_t architectures.
788 JH/17 Fix a null-indirection in certextract expansions when a nondefault
789 output list separator was used.
794 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
795 checkers that were complaining about end of non-void function with no
798 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
799 This was a regression introduced in 4.83 by another bugfix.
801 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
803 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
804 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
810 TF/01 Correctly close the server side of TLS when forking for delivery.
812 When a message was received over SMTP with TLS, Exim failed to clear up
813 the incoming connection properly after forking off the child process to
814 deliver the message. In some situations the subsequent outgoing
815 delivery connection happened to have the same fd number as the incoming
816 connection previously had. Exim would try to use TLS and fail, logging
817 a "Bad file descriptor" error.
819 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
820 utilities have not been installed.
822 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
823 temporary space as the ACL may create new global variables.
825 TL/01 LDAP support uses per connection or global context settings, depending
826 upon the detected version of the libraries at build time.
828 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
829 to extract and use the src ip:port in logging and expansions as if it
830 were a direct connection from the outside internet. PPv2 support was
831 updated based on HAProxy spec change in May 2014.
833 JH/02 Add ${listextract {number}{list}{success}{fail}}.
835 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
836 Properly escape header and check for NULL return.
838 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
841 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
843 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
844 characters in header names, implemented as a verify condition.
845 Contributed by Michael Fischer v. Mollard.
847 TL/05 Rename SPF condition results err_perm and err_temp to standardized
848 results permerror and temperror. Previous values are deprecated but
849 still accepted. In a future release, err_perm and err_temp will be
850 completely removed, which will be a backward incompatibility if the
851 ACL tests for either of these two old results. Patch contributed by
852 user bes-internal on the mailing list.
854 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
856 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
857 selectors, in both main and reject logs.
859 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
862 JH/07 Add malware type "sock" for talking to simple daemon.
864 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
866 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
867 routers/transports under cutthrough routing.
869 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
870 numbers. Touch up "bool" conditional to keep the same definition.
872 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
874 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
876 JH/12 Expand items in router/transport headers_add or headers_remove lists
877 individually rather than the list as a whole. Bug 1452.
879 Required for reasonable handling of multiple headers_ options when
880 they may be empty; requires that headers_remove items with embedded
881 colons must have them doubled (or the list-separator changed).
883 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
884 view the policy declared in the DMARC record. Currently, $dmarc_status
885 is a combined value of both the record presence and the result of the
888 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
890 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
891 dnslookup router and the smtp transport (applying to the forward
894 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
895 of ldap servers used for a specific lookup. Patch provided by Heiko
898 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
899 New variable $lookup_dnssec_authenticated for observability.
901 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
902 Patch submitted by Lars Timman.
904 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
906 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
907 Requires trusted mode and valid format message id, aborts otherwise.
908 Patch contributed by Heiko Schlichting.
910 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
911 certextract with support for various fields. Bug 1358.
913 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
914 is requested by default, modifiable by smtp transport option
917 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
918 operate on certificate variables to give certificate fingerprints
919 Also new ${sha256:cert_variable}.
921 JH/23 The PRDR feature is moved from being Experimental into the mainline.
923 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
924 Christian Aistleitner.
926 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
928 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
929 file. Patch from Wolfgang Breyha.
931 JH/25 Expand the coverage of the delivery $host and $host_address to
932 client authenticators run in verify callout. Bug 1476.
934 JH/26 Port service names are now accepted for tls_on_connect_ports, to
935 align with daemon_smtp_ports. Bug 72.
937 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
938 support and error reporting did not work properly.
940 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
941 and is readable. Patch from Andrew Colin Kissa.
943 TL/14 Enhance documentation of ${run expansion and how it parses the
944 commandline after expansion, particularly in the case when an
945 unquoted variable expansion results in an empty value.
947 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
949 PP/02 Fix internal collision of T_APL on systems which support RFC3123
950 by renaming away from it. Addresses GH issue 15, reported by
953 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
955 TL/15 SECURITY: prevent double expansion in math comparison functions
956 (can expand unsanitized data). Not remotely exploitable.
963 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
965 PP/02 Make -n do something, by making it not do something.
966 When combined with -bP, the name of an option is not output.
968 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
971 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
972 $sender_host_name and config options to manage this, and basic check
975 PP/05 DSCP support for outbound connections and control modifier for inbound.
977 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
978 (Only plugin which currently uses this is kerberos4, which nobody should
979 be using, but we should make it available and other future plugins might
980 conceivably use it, even though it would break NAT; stuff *should* be
981 using channel bindings instead).
983 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
984 name; added for Sendmail compatibility; requires admin caller.
985 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
986 just ignore it); requires trusted caller.
987 Also parse but ignore: -Ac -Am -X<logfile>
990 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
992 TL/02 Add +smtp_confirmation as a default logging option.
994 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
995 Patch by Magnus Holmgren from 2007-02-20.
997 TL/04 Bugzilla 1281 - Spec typo.
998 Bugzilla 1283 - Spec typo.
999 Bugzilla 1290 - Spec grammar fixes.
1001 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1003 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1005 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1006 mailing list by by Dmitry Isaikin.
1008 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1010 JH/02 Support "G" suffix to numbers in ${if comparisons.
1012 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1014 NM/01 Bugzilla 1197 - Spec typo
1015 Bugzilla 1196 - Spec examples corrections
1017 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1019 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1020 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1023 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1024 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1026 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1027 "acl {{name}{arg}...}", and optional args on acl condition
1030 JH/05 Permit multiple router/transport headers_add/remove lines.
1032 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1034 JH/07 Avoid using a waiting database for a single-message-only transport.
1035 Performance patch from Paul Fisher. Bugzilla 1262.
1037 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1040 JH/09 Add $headers_added variable, with content from use of ACL modifier
1041 add_header (but not yet added to the message). Bugzilla 199.
1043 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1044 Pulled from Bugzilla 817 by Wolfgang Breyha.
1046 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1048 (nb: this is the same fix as in Exim 4.80.1)
1050 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1053 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1056 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1058 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1059 advertise SMTP AUTH mechanism to us, instead of a generic
1060 protocol violation error. Also, make Exim more robust to bad
1061 data from the Dovecot auth socket.
1063 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1065 When a queue runner is handling a message, Exim first routes the
1066 recipient addresses, during which it prunes them based on the retry
1067 hints database. After that it attempts to deliver the message to
1068 any remaining recipients. It then updates the hints database using
1071 So if a recipient address works intermittently, it can get repeatedly
1072 deferred at routing time. The retry hints record remains fresh so the
1073 address never reaches the final cutoff time.
1075 This is a fairly common occurrence when a user is bumping up against
1076 their storage quota. Exim had some logic in its local delivery code
1077 to deal with this. However it did not apply to per-recipient defers
1078 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1080 This change adds a proper retry rule check during routing so that the
1081 final cutoff time is checked against the message's age. We only do
1082 this check if there is an address retry record and there is not a
1083 domain retry record; this implies that previous attempts to handle
1084 the address had the retry_use_local_parts option turned on. We use
1085 this as an approximation for the destination being like a local
1086 delivery, as in LMTP.
1088 I suspect this new check makes the old local delivery cutoff check
1089 redundant, but I have not verified this so I left the code in place.
1091 TF/02 Correct gecos expansion when From: is a prefix of the username.
1093 Test 0254 submits a message to Exim with the header
1097 When I ran the test suite under the user fanf2, Exim expanded
1098 the header to contain my full name, whereas it should have added
1099 a Resent-Sender: header. It erroneously treats any prefix of the
1100 username as equal to the username.
1102 This change corrects that bug.
1104 GF/01 DCC debug and logging tidyup
1105 Error conditions log to paniclog rather than rejectlog.
1106 Debug lines prefixed by "DCC: " to remove any ambiguity.
1108 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1110 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1111 Bug spotted by Jeremy Harris; was flawed since initial commit.
1112 Would have resulted in OCSP responses post-SNI triggering an Exim
1113 NULL dereference and crash.
1115 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1117 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1118 Bug detection, analysis and fix by Samuel Thibault.
1119 Bugzilla 1331, Debian bug #698092.
1121 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1123 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1124 Server implementation by Todd Lyons, client by JH.
1125 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1126 config variable "prdr_enable" controls whether the server
1127 advertises the facility. If the client requests PRDR a new
1128 acl_data_smtp_prdr ACL is called once for each recipient, after
1129 the body content is received and before the acl_smtp_data ACL.
1130 The client is controlled by both of: a hosts_try_prdr option
1131 on the smtp transport, and the server advertisement.
1132 Default client logging of deliveries and rejections involving
1133 PRDR are flagged with the string "PRDR".
1135 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1136 fclose(). Diagnosis by Todd Lyons.
1138 PP/17 Update configure.default to handle IPv6 localhost better.
1139 Patch by Alain Williams (plus minor tweaks).
1142 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1143 This is now consistent with GnuTLS, and is now documented: the
1144 previous undocumented portable approach to treating the option as
1145 unset was to force an expansion failure. That still works, and
1146 an empty string is now equivalent.
1148 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1149 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1150 not performing validation itself.
1152 PP/20 Added force_command boolean option to pipe transport.
1153 Patch from Nick Koston, of cPanel Inc.
1155 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1158 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1160 PP/21 Fix eximon continuous updating with timestamped log-files.
1161 Broken in a format-string cleanup in 4.80, missed when I repaired the
1162 other false fix of the same issue.
1163 Report and fix from Heiko Schlichting.
1166 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1167 Report from Prashanth Katuri.
1169 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1170 It's SecureTransport, so affects any MacOS clients which use the
1171 system-integrated TLS libraries, including email clients.
1173 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1174 using a MIME ACL for non-SMTP local injection.
1175 Report and assistance in diagnosis by Warren Baker.
1177 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1179 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1181 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1182 last id that failed so it may be referenced in subsequent ACL's.
1184 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1187 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1188 ldap library initialization, allowing self-signed CA's to be
1189 used. Also properly sets require_cert option later in code by
1190 using NULL (global ldap config) instead of ldap handle (per
1191 session). Bug diagnosis and testing by alxgomz.
1193 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1194 the src/util/ subdirectory.
1196 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1197 renamed to Transport Post Delivery Action by Jeremy Harris, as
1200 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1201 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1202 redis_servers = needs to be configured which will be used by the redis
1203 lookup. Patch from Warren Baker, of The Packet Hub.
1205 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1207 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1208 hostname or reverse DNS when processing a host list. Used suggestions
1209 from multiple comments on this bug.
1211 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1213 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1214 Missed a few lines, added it to make the runtest require no keyboard
1217 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1218 contains upper case chars. Make router use caseful_local_part.
1220 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1221 support when GnuTLS has been built with p11-kit.
1227 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1229 This, or similar/improved, will also be change PP/11 of 4.82.
1235 PP/01 Handle short writes when writing local log-files.
1236 In practice, only affects FreeBSD (8 onwards).
1237 Bugzilla 1053, with thanks to Dmitry Isaikin.
1239 NM/01 Bugzilla 949 - Documentation tweak
1241 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1244 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1246 PP/02 Implemented gsasl authenticator.
1248 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1250 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1251 `pkg-config foo` for cflags/libs.
1253 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1254 with rest of GSASL and with heimdal_gssapi.
1256 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1257 `pkg-config foo` for cflags/libs for the TLS implementation.
1259 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1260 properties get this fed in as external SSF. A number of robustness
1261 and debugging improvements to the cyrus_sasl authenticator.
1263 PP/08 cyrus_sasl server now expands the server_realm option.
1265 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1266 Patch by Jeremy Harris.
1268 PP/10 Added dbmjz lookup type.
1270 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1272 PP/12 MAIL args handles TAB as well as SP, for better interop with
1273 non-compliant senders.
1274 Analysis and variant patch by Todd Lyons.
1276 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1277 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1278 Patch from Dirk Mueller <dmueller@suse.com>
1280 PP/13 tls_peerdn now print-escaped for spool files.
1281 Observed some $tls_peerdn in wild which contained \n, which resulted
1282 in spool file corruption.
1284 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1285 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1286 or write after TLS renegotiation, which otherwise led to messages
1289 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1290 as a tracking header (ie: a signed header comes before the signature).
1291 Patch from Wolfgang Breyha.
1293 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1294 comma-sep list; embedded commas doubled.
1296 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1298 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1300 Report and patch from Dmitry Banschikov.
1302 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1303 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1304 failures. We appear to now support TLS1.1+ with Exim.
1306 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1307 lets Exim select keys and certificates based upon TLS SNI from client.
1308 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1309 before an outbound SMTP session. New log_selector, +tls_sni.
1311 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1312 NULL dereference. Report and patch from Alun Jones.
1314 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1315 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1316 Not seeing resolver debug output on NetBSD, but suspect this is a
1317 resolver implementation change.
1319 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1320 Left warnings. Added "eximon gdb" invocation mode.
1322 PP/21 Defaulting "accept_8bitmime" to true, not false.
1324 PP/22 Added -bw for inetd wait mode support.
1326 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1327 locate the relevant includes and libraries. Made this the default.
1329 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1330 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1332 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1333 This may cause build issues on older platforms.
1335 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1336 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1337 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1338 Added SNI support via GnuTLS too.
1339 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1341 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1343 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1344 Applied second patch from Janne, implementing suggestion to default
1345 multiple-strings-in-record handling to match SPF spec.
1347 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1349 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1350 read-only, out of scope).
1351 Patch from Wolfgang Breyha, report from Stuart Northfield.
1353 PP/29 Fix three issues highlighted by clang analyser static analysis.
1354 Only crash-plausible issue would require the Cambridge-specific
1355 iplookup router and a misconfiguration.
1356 Report from Marcin Mirosław.
1358 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1360 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1361 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1362 As part of this, removing so much warning spew let me fix some minor
1363 real issues in debug logging.
1365 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1366 assignment on my part. Fixed.
1368 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1369 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1370 Janne Snabb (who went above and beyond: thank you).
1372 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1373 string otherwise requires a connection and a bunch more work and it's
1374 relatively easy to get wrong. Should also expose TLS library linkage
1377 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1378 64-bit ${eval} (JH/03).
1380 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1381 GNU libc to support some of the 64-bit stuff, should not lead to
1382 conflicts. Defined before os.h is pulled in, so if a given platform
1383 needs to override this, it can.
1385 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1386 protection layer was required, which is not implemented.
1387 Bugzilla 1254, patch from Wolfgang Breyha.
1389 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1390 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1391 tls_dhparam take prime identifiers. Also unbreak combination of
1392 OpenSSL+DH_params+TLSSNI.
1394 PP/39 Disable SSLv2 by default in OpenSSL support.
1400 PP/01 Solaris build fix for Oracle's LDAP libraries.
1401 Bugzilla 1109, patch from Stephen Usher.
1403 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1405 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1408 TF/02 Fix a couple more cases where we did not log the error message
1409 when unlink() failed. See also change 4.74-TF/03.
1411 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1412 lock up or crash if it happened to be inside a call to libc when it
1413 got a SIGUSR1 from exiwhat.
1415 The SIGUSR1 handler appends the current process status to the process
1416 log which is later printed by exiwhat. It used to use the general
1417 purpose logging code to do this, but several functions it calls are
1418 not safe for signals.
1420 The new output code in the SIGUSR1 handler is specific to the process
1421 log, and simple enough that it's easy to inspect for signal safety.
1422 Removing some special cases also simplifies the general logging code.
1423 Removing the spurious timestamps from the process log simplifies
1426 TF/04 Improved ratelimit ACL condition.
1428 The /noupdate option has been deprecated in favour of /readonly which
1429 has clearer semantics. The /leaky, /strict, and /readonly update modes
1430 are mutually exclusive. The update mode is no longer included in the
1431 database key; it just determines when the database is updated. (This
1432 means that when you upgrade Exim will forget old rate measurements.)
1434 Exim now checks that the per_* options are used with an update mode that
1435 makes sense for the current ACL. For example, when Exim is processing a
1436 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1437 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1438 must specify per_mail/readonly. If you omit the update mode it defaults to
1439 /leaky where that makes sense (as before) or /readonly where required.
1441 The /noupdate option is now undocumented but still supported for
1442 backwards compatibility. It is equivalent to /readonly except that in
1443 ACLs where /readonly is required you may specify /leaky/noupdate or
1444 /strict/noupdate which are treated the same as /readonly.
1446 A useful new feature is the /count= option. This is a generalization
1447 of the per_byte option, so that you can measure the throughput of other
1448 aggregate values. For example, the per_byte option is now equivalent
1449 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1451 The per_rcpt option has been generalized using the /count= mechanism
1452 (though it's more complicated than the per_byte equivalence). When it is
1453 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1454 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1455 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1456 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1457 Note that using per_rcpt with a non-readonly update mode in more than
1458 one ACL will cause the recipients to be double-counted. (The per_mail
1459 and per_byte options don't have this problem.)
1461 The handling of very low rates has changed slightly. If the computed rate
1462 is less than the event's count (usually one) then this event is the first
1463 after a long gap. In this case the rate is set to the same as this event's
1464 count, so that the first message of a spam run is counted properly.
1466 The major new feature is a mechanism for counting the rate of unique
1467 events. The new per_addr option counts the number of different
1468 recipients that someone has sent messages to in the last time period. It
1469 behaves like per_rcpt if all the recipient addresses are different, but
1470 duplicate recipient addresses do not increase the measured rate. Like
1471 the /count= option this is a general mechanism, so the per_addr option
1472 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1473 example, measure the rate that a client uses different sender addresses
1474 with the options per_mail/unique=$sender_address. There are further
1475 details in the main documentation.
1477 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1479 TF/06 Removed a few PCRE remnants.
1481 TF/07 Automatically extract Exim's version number from tags in the git
1482 repository when doing development or release builds.
1484 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1485 Bugzilla 879. Patch from Paul Fisher.
1487 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1488 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1491 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1493 PP/05 Variable $av_failed, true if the AV scanner deferred.
1494 Bugzilla 1078. Patch from John Horne.
1496 PP/06 Stop make process more reliably on build failure.
1497 Bugzilla 1087. Patch from Heiko Schlittermann.
1499 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1500 Bugzilla 1089. Patch from Heiko Schlittermann.
1502 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1503 Bugzilla 1131. Patch from Holger Weiß.
1505 PP/09 Handle IPv6 addresses with SPF.
1506 Bugzilla 860. Patch from Wolfgang Breyha.
1508 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1510 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1513 PP/11 match_* no longer expand right-hand-side by default.
1514 New compile-time build option, EXPAND_LISTMATCH_RHS.
1515 New expansion conditions, "inlist", "inlisti".
1517 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1519 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1521 PP/14 fix log_write() format string regression from TF/03.
1522 Bugzilla 1152. Patch from Dmitry Isaikin.
1528 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1530 PP/02 Harmonised TLS library version reporting; only show if debugging.
1531 Layout now matches that introduced for other libraries in 4.74 PP/03.
1533 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1535 PP/04 New "dns_use_edns0" global option.
1537 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1540 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1541 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1543 TK/01 Updated PolarSSL code to 0.14.2.
1544 Bugzilla 1097. Patch from Andreas Metzler.
1546 PP/07 Catch divide-by-zero in ${eval:...}.
1547 Fixes bugzilla 1102.
1549 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1552 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1553 format-string attack -- SECURITY: remote arbitrary code execution.
1555 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1556 time unintentionally subject to list matching rules, letting the header
1557 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1558 arbitrary string expansion). This allowed for information disclosure.
1560 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1561 INT_MIN/-1 -- value coerced to INT_MAX.
1567 NM/01 Workaround for PCRE version dependency in version reporting
1570 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1571 This fixes portability to compilers other than gcc, notably
1572 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1574 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1575 makefiles for portability to HP-UX and POSIX correctness.
1577 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1578 Also via indented variable definition in the Makefile.
1579 (Debugging by Oliver Heesakkers).
1581 PP/02 Restore caching of spamd results with expanded spamd_address.
1582 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1584 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1585 Improves build reliability. Fix from: Frank Elsner
1587 NM/02 Fix wide character breakage in the rfc2047 coding
1588 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1590 NM/03 Allow underscore in dnslist lookups
1591 Fixes bug 1026. Patch from Graeme Fowler
1593 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1594 Code patches from Adam Ciarcinski of NetBSD.
1596 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1599 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1600 is logged, to avoid truncation. Patch from John Horne.
1602 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1603 Patch from Jakob Hirsch.
1605 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1606 SQL string expansion failure details.
1607 Patch from Andrey Oktyabrski.
1609 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1610 Patch from Simon Arlott.
1612 PP/09 New lookups functionality failed to compile on old gcc which rejects
1613 extern declarations in function scope.
1614 Patch from Oliver Fleischmann
1616 PP/10 Use sig_atomic_t for flags set from signal handlers.
1617 Check getgroups() return and improve debugging.
1618 Fixed developed for diagnosis in bug 927 (which turned out to be
1621 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1622 Patch from Mark Zealey.
1624 PP/12 Bugzilla 1056: Improved spamd server selection.
1625 Patch from Mark Zealey.
1627 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1628 Based on patch from Heiko Schlittermann.
1630 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1631 Patch from Uwe Doering, sign-off by Michael Haardt.
1633 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1634 variable declaration deep within a block. Bug and patch from
1637 PP/15 lookups-Makefile IRIX compatibility coercion.
1639 PP/16 Make DISABLE_DKIM build knob functional.
1641 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1642 Patch by Simon Arlott
1644 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1645 variable argument macros. Our copy now differs from upstream.
1651 TF/01 Failure to get a lock on a hints database can have serious
1652 consequences so log it to the panic log.
1654 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1655 controlled using the smtp_confirmation log selector.
1657 TF/03 Include the error message when we fail to unlink a spool file.
1659 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1660 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1661 for maintaining out-of-tree patches for some time.
1663 PP/01 Bugzilla 139: Documentation and portability issues.
1664 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1665 Handle per-OS dynamic-module compilation flags.
1667 PP/02 Let /dev/null have normal permissions.
1668 The 4.73 fixes were a little too stringent and complained about the
1669 permissions on /dev/null. Exempt it from some checks.
1670 Reported by Andreas M. Kirchwitz.
1672 PP/03 Report version information for many libraries, including
1673 Exim version information for dynamically loaded libraries. Created
1674 version.h, now support a version extension string for distributors
1675 who patch heavily. Dynamic module ABI change.
1677 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1678 privilege escalation vulnerability whereby the Exim run-time user
1679 can cause root to append content of the attacker's choosing to
1682 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1685 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1686 If dropping privileges for untrusted macros, we disabled normal logging
1687 on the basis that it would fail; for the Exim run-time user, this is not
1688 the case, and it resulted in successful deliveries going unlogged.
1689 Fixed. Reported by Andreas Metzler.
1695 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1696 only prepend for the Resent-* case. Fixes regression introduced in
1697 Exim 4.70 by NM/22 for Bugzilla 607.
1699 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1700 increasing numbers of administrators be bitten by this.
1702 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1704 PP/03 Bugzilla 994: added openssl_options main configuration option.
1706 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1708 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1710 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1712 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1713 without a peer certificate, leading to a segfault because of an
1714 assumption that peers always have certificates. Be a little more
1715 paranoid. Problem reported by Martin Tscholak.
1717 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1718 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1719 NB: ClamAV planning to remove STREAM in "middle of 2010".
1720 CL also introduces -bmalware, various -d+acl logging additions and
1721 more caution in buffer sizes.
1723 PP/09 Implemented reverse_ip expansion operator.
1725 PP/10 Bugzilla 937: provide a "debug" ACL control.
1727 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1729 PP/12 Bugzilla 973: Implement --version.
1731 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1733 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1735 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1737 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1738 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1739 ignore trailing whitespace.
1741 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1743 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1746 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1747 Notification from Dr Andrew Aitchison.
1749 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1750 ExtendedDetectionInfo response format.
1751 Notification from John Horne.
1753 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1756 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1757 XSL and documented dependency on system catalogs, with examples of how
1760 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1763 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1764 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1765 configuration file which is writeable by the Exim user or group.
1767 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1768 of configuration files to cover files specified with the -C option if
1769 they are going to be used with root privileges, not just the default
1772 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1773 option (effectively making it always true).
1775 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1776 files to be used while preserving root privileges.
1778 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1779 that rogue child processes cannot use them.
1781 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1782 run-time user, instead of root.
1784 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1785 Exim run-time user without dropping privileges.
1787 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1788 result string, instead of calling string_vformat() twice with the same
1791 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1792 for other users. Others should always drop root privileges if they use
1793 -C on the command line, even for a whitelisted configure file.
1795 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1797 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1803 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1804 $data_path, and $header_path variables; fixed documentation bugs and
1807 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1808 exipick to access non-standard spools, including the "frozen" queue
1811 NM/01 Bugzilla 965: Support mysql stored procedures.
1812 Patch from Alain Williams
1814 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1816 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1817 Patch from Andreas Metzler
1819 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1820 Patch from Kirill Miazine
1822 NM/05 Bugzilla 671: Added umask to procmail example.
1824 JJ/03 installed exipick 20100323.0, fixing doc bug
1826 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1827 directory. Notification and patch from Dan Rosenberg.
1829 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1831 TK/02 Improve log output when DKIM signing operation fails.
1833 MH/01 Treat the transport option dkim_domain as a colon separated
1834 list, not as a single string, and sign the message with each element,
1835 omitting multiple occurences of the same signer.
1837 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1838 Bugzilla 985, 986. Patch by Simon Arlott
1840 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1841 Patch by Simon Arlott
1843 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1844 MBX locking. Notification from Dan Rosenberg.
1850 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1852 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1854 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1856 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1858 NM/04 Bugzilla 915: Flag broken perl installation during build.
1864 TK/01 Added patch by Johannes Berg that expands the main option
1865 "spamd_address" if it starts with a dollar sign.
1867 TK/02 Write list of recipients to X-Envelope-Sender header when building
1868 the mbox-format spool file for content scanning (suggested by Jakob
1871 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1872 (http://www.dcc-servers.net/) support via dccifd. Activated by
1873 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1875 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1876 by Mark Daniel Reidel <mr@df.eu>.
1878 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1879 When building exim an external PCRE library is now needed -
1880 PCRE is a system library on the majority of modern systems.
1881 See entry on PCRE_LIBS in EDITME file.
1883 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1884 conversation. Added nologin parameter to request.
1885 Patch contributed by Kirill Miazine.
1887 TF/01 Do not log submission mode rewrites if they do not change the address.
1889 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1891 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1892 log files in place. Contributed by Roberto Lima.
1894 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1896 TF/03 Bugzilla 615: When checking the local_parts router precondition
1897 after a local_part_suffix or local_part_prefix option, Exim now
1898 does not use the address's named list lookup cache, since this
1899 contains cached lookups for the whole local part.
1901 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1902 Robert Millan. Documentation is in experimental-spec.txt.
1904 TF/04 Bugzilla 668: Fix parallel build (make -j).
1906 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1908 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1909 Patch provided by Jan Srzednicki.
1911 TF/05 Leading white space used to be stripped from $spam_report which
1912 wrecked the formatting. Now it is preserved.
1914 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1915 that they are available at delivery time.
1917 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1919 TF/08 TLS error reporting now respects the incoming_interface and
1920 incoming_port log selectors.
1922 TF/09 Produce a more useful error message if an SMTP transport's hosts
1923 setting expands to an empty string.
1925 NM/06 Bugzilla 744: EXPN did not work under TLS.
1926 Patch provided by Phil Pennock.
1928 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1929 Patch provided by Richard Godbee.
1931 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1932 acl_smtp_notquit, added index entry.
1934 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1935 Patch provided by Eugene Bujak.
1937 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1938 accept(). Patch provided by Maxim Dounin.
1940 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1941 Patch provided by Phil Pennock.
1943 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1945 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1946 Patch provided by Brad "anomie" Jorsch.
1948 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1949 Patch provided by Dean Brooks.
1951 TK/05 Add native DKIM support (does not depend on external libraries).
1953 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1954 Patch provided by Graeme Fowler.
1956 NM/16 Bugzilla 851: Documentation example syntax fix.
1958 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1960 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1963 NM/19 Bugzilla 745: TLS version reporting.
1964 Patch provided by Phil Pennock.
1966 NM/20 Bugzilla 167: bool: condition support.
1967 Patch provided by Phil Pennock.
1969 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1970 clients. Patch provided by Phil Pennock.
1972 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1973 Patch provided by Brad "anomie" Jorsch.
1975 NM/23 Bugzilla 687: Fix misparses in eximstats.
1976 Patch provided by Heiko Schlittermann.
1978 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1979 Patch provided by Heiko Schlittermann.
1981 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1982 plus update to original patch.
1984 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1986 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1987 Patch provided by David Brownlee.
1989 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1991 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1993 NM/30 Bugzilla 888: TLS documentation bugfixes.
1995 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1997 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1998 Unlike the original bugzilla I have changed all shell scripts in src tree.
2000 NM/33 Bugzilla 898: Transport filter timeout fix.
2001 Patch by Todd Rinaldo.
2003 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2004 Patch by Serge Demonchaux.
2006 NM/35 Bugzilla 39: Base64 decode bug fixes.
2007 Patch by Jakob Hirsch.
2009 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2011 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2013 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2015 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2021 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2022 ALT-N's libdkim that I have put here:
2023 http://duncanthrax.net/exim-experimental/
2025 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2026 were called 'true' and it seems that C99 defines that as a reserved
2027 keyword to be used with 'bool' variable types. That means you could
2028 not include C99-style headers which use bools without triggering
2029 build errors in sieve.c.
2031 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2032 as mailq or other aliases. Changed the --help handling significantly
2033 to do whats expected. exim_usage() emits usage/help information.
2035 SC/01 Added the -bylocaldomain option to eximstats.
2037 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2039 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2041 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2047 PH/01 Another patch from the Sieve maintainer.
2049 PH/02 When an IPv6 address is converted to a string for single-key lookup
2050 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2051 dots are used instead of colons so that keys in lsearch files need not
2052 contain colons. This was done some time before quoting was made available
2053 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2054 (notated using the quote facility) so as to distinguish them from IPv4
2055 keys. This meant that lookups for IP addresses in host lists did not work
2056 for iplsearch lookups.
2058 This has been fixed by arranging for IPv6 addresses to be expressed with
2059 colons if the lookup type is iplsearch. This is not incompatible, because
2060 previously such lookups could never work.
2062 The situation is now rather anomalous, since one *can* have colons in
2063 ordinary lsearch keys. However, making the change in all cases is
2064 incompatible and would probably break a number of configurations.
2066 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2069 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2070 conversion specification without a maximum field width, thereby enabling
2071 a rogue spamd server to cause a buffer overflow. While nobody in their
2072 right mind would setup Exim to query an untrusted spamd server, an
2073 attacker that gains access to a server running spamd could potentially
2074 exploit this vulnerability to run arbitrary code as the Exim user.
2076 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2077 $primary_hostname instead of what libspf2 thinks the hosts name is.
2079 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2080 a directory entry by the name of the lookup key. Previously, if a
2081 symlink pointed to a non-existing file or a file in a directory that
2082 Exim lacked permissions to read, a lookup for a key matching that
2083 symlink would fail. Now it is enough that a matching directory entry
2084 exists, symlink or not. (Bugzilla 503.)
2086 PH/03 The body_linecount and body_zerocount variables are now exported in the
2089 PH/04 Added the $dnslist_matched variable.
2091 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2092 This means they are set thereafter only if the connection becomes
2095 PH/06 Added the client_condition to authenticators so that some can be skipped
2096 by clients under certain conditions.
2098 PH/07 The error message for a badly-placed control=no_multiline_responses left
2099 "_responses" off the end of the name.
2101 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2103 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2104 (without spaces) instead of just copying the configuration text.
2106 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2108 PH/11 Added $max_received_linelength.
2110 PH/12 Added +ignore_defer and +include_defer to host lists.
2112 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2113 way in which PCRE > 7.0 is built.
2115 PH/14 Implemented queue_only_load_latch.
2117 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2118 MAIL command. The effect was to mangle the value on 64-bit systems.
2120 PH/16 Another patch from the Sieve maintainer.
2122 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2124 PH/18 If a system quota error occurred while trying to create the file for
2125 a maildir delivery, the message "Mailbox is full" was not appended to the
2126 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2127 only to a quota excession during the actual writing of the file.
2129 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2130 characters?) which causes problems in log lines. The DN values are now
2131 passed through string_printing() before being added to log lines.
2133 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2134 and InterBase are left for another time.)
2136 PH/21 Added message_body_newlines option.
2138 PH/22 Guard against possible overflow in moan_check_errorcopy().
2140 PH/23 POSIX allows open() to be a macro; guard against that.
2142 PH/24 If the recipient of an error message contained an @ in the local part
2143 (suitably quoted, of course), incorrect values were put in $domain and
2144 $local_part during the evaluation of errors_copy.
2150 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2151 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2154 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2155 issue a MAIL command.
2157 PH/02 In an ACL statement such as
2159 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2161 if a client was not listed at all, or was listed with a value other than
2162 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2163 the condition was not true (as it should be), so access was not denied.
2164 The bug was that the ! inversion was incorrectly passed on to the second
2165 item. This has been fixed.
2167 PH/03 Added additional dnslists conditions == and =& which are different from
2168 = and & when the dns lookup returns more than one IP address.
2170 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2171 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2173 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2174 FSYNC, which compiles an option called disable_fsync that allows for
2175 bypassing fsync(). The documentation is heavily laced with warnings.
2177 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2179 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2180 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2181 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2182 including adding "make clean"; (3) Added -fPIC when compiling the test
2183 dynamically loaded module, to get rid of a warning.
2185 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2186 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2187 = 0s. The bug is otherwise harmless.
2189 PH/07 There was a bug in the dovecot authenticator such that the value of
2190 $auth1 could be overwritten, and so not correctly preserved, after a
2191 successful authentication. This usually meant that the value preserved by
2192 the server_setid option was incorrect.
2194 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2196 PH/09 Installed PCRE release 7.0.
2198 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2199 run for batched SMTP input. It is now run at the start of every message
2200 in the batch. While fixing this I discovered that the process information
2201 (output by running exiwhat) was not always getting set for -bs and -bS
2202 input. This is fixed, and it now also says "batched" for BSMTP.
2204 PH/11 Added control=no_pipelining.
2206 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2207 patch, slightly modified), and move the expansion of helo_data till after
2208 the connection is made in the smtp transport (so it can use these
2211 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2213 PH/14 Added log_selector = +pid.
2215 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2217 PH/16 Add ${if forany and ${if forall.
2219 PH/17 Added dsn_from option to vary the From: line in DSNs.
2221 PH/18 Flush SMTP output before performing a callout, unless control =
2222 no_callout_flush is set.
2224 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2225 was true (the default) a successful delivery failed to delete the retry
2226 item, thus causing premature timeout of the address. The bug is now
2229 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2231 PH/21 Long custom messages for fakedefer and fakereject are now split up
2232 into multiline responses in the same way that messages for "deny" and
2233 other ACL rejections are.
2235 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2236 with slight modification.
2238 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2239 draft, changing the syntax and factoring some duplicate code".
2241 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2242 for deliveries of the second and subsequent messages over the same SMTP
2245 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2246 ${reduce, with only minor "tidies".
2248 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2250 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2251 expansion side effects.
2253 PH/27 When a message times out after an over-quota error from an Exim-imposed
2254 quota, the bounce message says "mailbox is full". This message was not
2255 being given when it was a system quota that was exceeded. It now should
2258 MH/03 Made $recipients available in local_scan(). local_scan() already has
2259 better access to the recipient list through recipients_list[], but
2260 $recipients can be useful in postmaster-provided expansion strings.
2262 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2263 in the case of a MAIL command with additional options following the
2264 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2265 were accidentally chopped off.
2267 PH/29 SMTP synchronization checks are implemented when a command is read -
2268 there is a check that no more input is waiting when there shouldn't be
2269 any. However, for some commands, a delay in an ACL can mean that it is
2270 some time before the response is written. In this time, more input might
2271 arrive, invalidly. So now there are extra checks after an ACL has run for
2272 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2273 pipelining has not been advertised.
2275 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2277 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2278 returned in the SMTP response when smtp_return_error_details was false.
2279 This has been fixed.
2281 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2282 instead of the C I/O that was originally supplied, because problems were
2283 reported on Solaris.
2285 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2286 Exim which did not show up earlier: it was assuming that a call to
2287 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2288 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2289 was a macro that became an assignment, so it seemed to work. This has
2290 changed to a proper function call with a void return, hence the compile
2291 error. Exim's code has been fixed.
2293 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2296 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2298 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2300 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2301 because it turns out to be too big or there is a timeout) there is no
2302 "Completed" line in the log. When some messages of this type were
2303 selected by exigrep, they were listed as "not completed". Others were
2304 picked up by some special patterns. I have improved the selection
2305 criteria to be more general.
2307 PH/38 The host_find_failed option in the manualroute router can now be set
2308 to "ignore", to completely ignore a host whose IP address cannot be
2309 found. If all hosts are ignored, the behaviour is controlled by the new
2310 host_all_ignored option.
2312 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2313 homing or because of multiple MX records with /mx) generated more than
2314 one IP address, and the following item turned out to be the local host,
2315 all the secondary addresses of the first item were incorrectly removed
2316 from the list, along with the local host and any following hosts (which
2317 is what is supposed to happen).
2319 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2320 whoever called Exim into the -H file. In the case of the daemon it was
2321 behaving confusingly. When first started, it used values for whoever
2322 started the daemon, but after a SIGHUP it used the Exim user (because it
2323 calls itself on a restart). I have changed the code so that it now always
2326 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2327 message are rejected with the same error (e.g. no authentication or bad
2328 sender address), and a DATA command is nevertheless sent (as can happen
2329 with PIPELINING or a stupid MUA), the error message that was given to the
2330 RCPT commands is included in the rejection of the DATA command. This is
2331 intended to be helpful for MUAs that show only the final error to their
2334 PH/42 Another patch from the Sieve maintainer.
2336 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2337 Eximstats - Fixed some broken HTML links and added missing column headers
2339 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2340 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2342 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2344 PH/43 Yet another patch from the Sieve maintainer.
2346 PH/44 I found a way to check for a TCP/IP connection going away before sending
2347 the response to the final '.' that terminates a message, but only in the
2348 case where the client has not sent further data following the '.'
2349 (unfortunately, this is allowed). However, in many cases there won't be
2350 any further data because there won't be any more messages to send. A call
2351 to select() can be used: if it shows that the input is "ready", there is
2352 either input waiting, or the socket has been closed. An attempt to read
2353 the next input character can distinguish the two cases. Previously, Exim
2354 would have sent an OK response which the client would never have see.
2355 This could lead to message repetition. This fix should cure that, at
2356 least in a lot of common cases.
2358 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2359 advertised in response to EHLO.
2365 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2366 fixed by 4.65/MH/01 (is this a record?) are fixed:
2368 (i) An empty string was always treated as zero by the numeric comparison
2369 operators. This behaviour has been restored.
2371 (ii) It is documented that the numeric comparison operators always treat
2372 their arguments as decimal numbers. This was broken in that numbers
2373 starting with 0 were being interpreted as octal.
2375 While fixing these problems I realized that there was another issue that
2376 hadn't been noticed. Values of message_size_limit (both the global option
2377 and the transport option) were treated as octal if they started with 0.
2378 The documentation was vague. These values are now always treated as
2379 decimal, and I will make that clear in the documentation.
2385 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2386 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2389 MH/01 Don't check that the operands of numeric comparison operators are
2390 integers when their expansion is in "skipping" mode (fixes bug
2391 introduced by 4.64-PH/07).
2393 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2394 child addresses, Exim now panics and dies. Previously, because the count
2395 is held in a short int, deliveries were likely to be lost. As such a
2396 large number of recipients for a single message is ridiculous
2397 (performance will be very, very poor), I have chosen to impose a limit
2398 rather than extend the field.
2404 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2405 leftover -K file (the existence of which was triggered by #402).
2406 While we were at it, introduced process PID as part of the -K
2407 filename. This should rule out race conditions when creating
2410 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2411 processing considerably. Previous code took too long for large mails,
2412 triggering a timeout which in turn triggers #401.
2414 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2415 in the DK code in transports.c. sendfile() is not really portable,
2416 hence the _LINUX specificness.
2418 TF/01 In the add_headers option to the mail command in an Exim filter,
2419 there was a bug that Exim would claim a syntax error in any
2420 header after the first one which had an odd number of characters
2423 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2424 callout verification, Exim cached a "reject" for the entire domain. This
2425 is correct for most verifications, but it is not correct for a recipient
2426 verification with use_sender or use_postmaster set, because in that case
2427 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2428 case of MAIL FROM:<> rejection from other early rejections (e.g.
2429 rejection of HELO). When verifying a recipient using a non-null MAIL
2430 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2431 Whatever the result of the callout, the value of the domain cache is
2432 left unchanged (for any other kind of callout, getting as far as trying
2433 RCPT means that the domain itself is ok).
2435 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2438 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2439 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2440 session, the daemon ignores ECONNECT errors and logs others; it now
2441 ignores EPIPE as well.
2443 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2444 (quoted-printable decoding).
2446 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2447 later the small subsequent patch to fix an introduced bug.
2449 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2451 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2453 PH/08 An error is now given if message_size_limit is specified negative.
2455 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2456 to be given (somewhat) arbitrary names.
2458 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2461 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2462 miscellaneous code fixes
2464 PH/10 Added the log_reject_target ACL modifier to specify where to log
2467 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2468 hostname. This is wrong, because it relates to the incoming message (and
2469 probably the interface on which it is arriving) and not to the outgoing
2470 callout (which could be using a different interface). This has been
2471 changed to use the value of the helo_data option from the smtp transport
2472 instead - this is what is used when a message is actually being sent. If
2473 there is no remote transport (possible with a router that sets up host
2474 addresses), $smtp_active_hostname is used.
2476 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2477 tweaks were necessary in order to get it to work (see also 21 below):
2478 (a) The code assumed that strncpy() returns a negative number on buffer
2479 overflow, which isn't the case. Replaced with Exim's string_format()
2481 (b) There were several signed/unsigned issues. I just did the minimum
2482 hacking in of casts. There is scope for a larger refactoring.
2483 (c) The code used strcasecmp() which is not a standard C function.
2484 Replaced with Exim's strcmpic() function.
2485 (d) The code set only $1; it now sets $auth1 as well.
2486 (e) A simple test gave the error "authentication client didn't specify
2487 service in request". It would seem that Dovecot has changed its
2488 interface. Fortunately there's a specification; I followed it and
2489 changed what the client sends and it appears to be working now.
2491 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2494 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2495 address A is aliased to B and C, where B exists and C does not. Without
2496 -v the output is "A verified" because verification stops after a
2497 successful redirection if more than one address is generated. However,
2498 with -v the child addresses are also verified. Exim was outputting "A
2499 failed to verify" and then showing the successful verification for C,
2500 with its parentage. It now outputs "B failed to verify", showing B's
2501 parentage before showing the successful verification of C.
2503 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2504 look up a TXT record in a specific list after matching in a combined
2507 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2508 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2509 they consult the DNS. I had assumed they would set it the way they
2510 wanted; and indeed my experiments on Linux seem to show that in some
2511 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2512 To be on the safe side, however, I have now made the interface to
2513 host_find_byname() similar to host_find_bydns(), with an argument
2514 containing the DNS resolver options. The host_find_byname() function now
2515 sets these options at its start, just as host_find_bydns() does. The smtp
2516 transport options dns_qualify_single and dns_search_parents are passed to
2517 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2518 of host_find_byname() use the default settings of RES_DEFNAMES
2519 (qualify_single) but not RES_DNSRCH (search_parents).
2521 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2522 spool_read_header() do less string testing, by means of a preliminary
2523 switch on the second character of optional "-foo" lines. (This is
2524 overdue, caused by the large number of possibilities that now exist.
2525 Originally there were few.) While I was there, I also converted the
2526 str(n)cmp tests so they don't re-test the leading "-" and the first
2527 character, in the hope this might squeeze out yet more improvement.
2529 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2530 flag allowing group syntax was set by the header_syntax check but not
2531 turned off, possible causing trouble later; (2) The flag was not being
2532 set at all for the header_verify test, causing "group"-style headers to
2533 be rejected. I have now set it in this case, and also caused header_
2534 verify to ignore an empty address taken from a group. While doing this, I
2535 came across some other cases where the code for allowing group syntax
2536 while scanning a header line wasn't quite right (mostly, not resetting
2537 the flag correctly in the right place). These bugs could have caused
2538 trouble for malformed header lines. I hope it is now all correct.
2540 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2541 with the "reply" argument non-NULL. The code, however (which originally
2542 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2543 but it didn't always do it. This confused somebody who was copying the
2544 code for some other use. I have removed all the tests.
2546 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2547 feature that was used to support insecure browsers during the U.S. crypto
2548 embargo. It requires special client support, and Exim is probably the
2549 only MTA that supported it -- and would never use it because real RSA is
2550 always available. This code has been removed, because it had the bad
2551 effect of slowing Exim down by computing (never used) parameters for the
2552 RSA_EXPORT functionality.
2554 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2555 authenticator to fail if there's a tab character in the incoming data
2556 (there should never be unless someone is messing about, as it's supposed
2557 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2558 if the connection is using TLS or if the remote IP is the same as the
2559 local IP, and the "valid-client-cert option" if a client certificate has
2562 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2563 authenticators. This can be used for authorization after authentication
2564 succeeds. (In the case of plaintext, it servers for both authentication
2567 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2568 if any retry times were supplied.
2570 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2571 connection, where there is no client IP address to check. In this
2572 situation, the verify now always succeeds.
2574 PH/25 Applied John Jetmore's -Mset patch.
2576 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2578 PH/27 In a string expansion for a processed (not raw) header when multiple
2579 headers of the same name were present, leading whitespace was being
2580 removed from all of them, but trailing whitespace was being removed only
2581 from the last one. Now trailing whitespace is removed from each header
2582 before concatenation. Completely empty headers in a concatenation (as
2583 before) are ignored.
2585 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2586 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2588 PH/29 [Removed. This was a change that I later backed out, and forgot to
2589 correct the ChangeLog entry (that I had efficiently created) before
2590 committing the later change.]
2592 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2593 address errors (4xx response to RCPT) over the same connection as other
2594 messages routed to the same hosts. Such deliveries are always "forced",
2595 so retry times are not inspected. This resulted in far too many retries
2596 for the affected addresses. The effect occurred only when there were more
2597 hosts than the hosts_max_try setting in the smtp transport when it had
2598 the 4xx errors. Those hosts that it had tried were not added to the list
2599 of hosts for which the message was waiting, so if all were tried, there
2600 was no problem. Two fixes have been applied:
2602 (i) If there are any address or message errors in an SMTP delivery, none
2603 of the hosts (tried or untried) are now added to the list of hosts
2604 for which the message is waiting, so the message should not be a
2605 candidate for sending over the same connection that was used for a
2606 successful delivery of some other message. This seems entirely
2607 reasonable: after all the message is NOT "waiting for some host".
2608 This is so "obvious" that I'm not sure why it wasn't done
2609 previously. Hope I haven't missed anything, but it can't do any
2610 harm, as the worst effect is to miss an optimization.
2612 (ii) If, despite (i), such a delivery is accidentally attempted, the
2613 routing retry time is respected, so at least it doesn't keep
2614 hammering the server.
2616 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2617 in ${readsocket because some servers need this prod.
2619 PH/32 Added some extra debug output when updating a wait-xxx database.
2621 PH/33 The hint "could be header name not terminated by colon", which has been
2622 given for certain expansion errors for a long time, was not being given
2623 for the ${if def:h_colon_omitted{... case.
2625 PH/34 The spec says: "With one important exception, whenever a domain list is
2626 being scanned, $domain contains the subject domain." There was at least
2627 one case where this was not true.
2629 PH/35 The error "getsockname() failed: connection reset by peer" was being
2630 written to the panic log as well as the main log, but it isn't really
2631 panic-worthy as it just means the connection died rather early on. I have
2632 removed the panic log writing for the ECONNRESET error when getsockname()
2635 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2636 runs only) independently of the message's sender address. This meant
2637 that, if the 4xx error was in fact related to the sender, a different
2638 message to the same recipient with a different sender could confuse
2639 things. In particular, this can happen when sending to a greylisting
2640 server, but other circumstances could also provoke similar problems.
2641 I have changed the default so that the retry time for these errors is now
2642 based a combination of the sender and recipient addresses. This change
2643 can be overridden by setting address_retry_include_sender=false in the
2646 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2647 remote server are returned as part of bounce messages. This was not
2648 happening for LMTP over a pipe (the lmtp transport), but now it is the
2649 same for both kinds of LMTP.
2651 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2652 in header lines that were in fact CNAMEs. This is no longer the case.
2654 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2655 and queue runs started by the daemon processed all messages. This has
2656 been fixed so that -R and -S can now usefully be given with -q<time>.
2658 PH/40 Import PCRE release 6.7 (fixes some bugs).
2660 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2662 PH/42 Give an error if -q is specified more than once.
2664 PH/43 Renamed the variables $interface_address and $interface_port as
2665 $received_ip_address and $received_port, to make it clear that these
2666 values apply to message reception, and not to the outgoing interface when
2667 a message is delivered. (The old names remain recognized, of course.)
2669 PH/44 There was no timeout on the connect() call when using a Unix domain
2670 socket in the ${readsocket expansion. There now is.
2672 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2673 be meaningful with "accept".
2675 SC/01 Eximstats V1.43
2676 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2678 SC/02 Eximstats V1.44
2679 Use a glob alias rather than an array ref in the generated
2680 parser. This improves both readability and performance.
2682 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2683 Collect SpamAssassin and rejection statistics.
2684 Don't display local sender or destination tables unless
2685 there is data to show.
2686 Added average volumes into the top table text output.
2688 SC/04 Eximstats V1.46
2689 Collect data on the number of addresses (recipients)
2690 as well as the number of messages.
2692 SC/05 Eximstats V1.47
2693 Added 'Message too big' to the list of mail rejection
2694 reasons (thanks to Marco Gaiarin).
2696 SC/06 Eximstats V1.48
2697 Mainlog lines which have GMT offsets and are too short to
2698 have a flag are now skipped.
2700 SC/07 Eximstats V1.49 (Alain Williams)
2701 Added the -emptyok flag.
2703 SC/08 Eximstats V1.50
2704 Fixes for obtaining the IP address from reject messages.
2706 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2707 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2708 whitespace changes from 4.64-PH/27
2710 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2713 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2714 are found, allow negative numbers in numeric criteria)
2716 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2718 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2719 to match changes made in 4.64-PH/43
2721 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2723 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2724 "domain" is actually an IP address, and force a failure. This locks out
2725 those revolvers/nameservers that support "A-for-A" lookups, in
2726 contravention of the specifications.
2728 PH/48 When a host name was looked up from an IP address, and the subsequent
2729 forward lookup of the name timed out, the host name was left in
2730 $sender_host_name, contrary to the specification.
2732 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2733 restricted to single-key lookups, Exim was not diagnosing an error if
2734 * or *@ was used with a query-style lookup.
2736 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2738 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2739 long ago, but noone interested enough thought of it. Let's just say that
2740 the "1.1" means that there are some new functions that weren't there at
2741 some point in the past.
2743 PH/51 Error processing for expansion failure of helo_data from an smtp
2744 transport during callout processing was broken.
2746 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2747 tested/used via the -bh/-bhc/-bs options.
2749 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2750 bug, fixed in subsequent PCRE releases).
2752 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2753 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2755 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2761 SC/01 Use a glob alias rather than an array ref in eximstats generated
2762 parser. This improves both readability and performance.
2764 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2765 Don't display local sender or destination tables in eximstats unless
2766 there is data to show.
2767 Added average volumes into the eximstats top table text output.
2769 SC/03 Collect data on the number of addresses (recipients) as well
2770 as the number of messages in eximstats.
2772 TF/01 Correct an error in the documentation for the redirect router. Exim
2773 does (usually) call initgroups() when daemonizing.
2775 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2776 with consistent privilege compared to when running as a daemon.
2778 TF/03 Note in the spec that $authenticated_id is not set for local
2779 submissions from trusted users.
2781 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2782 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2784 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2785 by adding some example configuration directives to the default
2786 configuration file. A little bit of work is required to uncomment the
2787 directives and define how usernames and passwords are checked, but
2788 there is now a framework to start from.
2790 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2791 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2792 without this. I don't know how relevant this is to other LDAP libraries.
2794 PH/02 Add the verb name to the "unknown ACL verb" error.
2796 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2798 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2800 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2801 directory not expanded when it should be if an expanded home directory
2802 was set for the address (which is overridden by the transport).
2804 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2807 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2808 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2809 because it is too late at that time, and has no effect.
2811 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2812 security issue with \' (bugzilla #107). I could not use the
2813 PQescapeStringConn() function, because it needs a PGconn value as one of
2816 PH/08 When testing addresses using -bt, indicate those final addresses that
2817 are duplicates that would not cause an additional delivery. At least one
2818 person was confused, thinking that -bt output corresponded to deliveries.
2819 (Suppressing duplicates isn't a good idea as you lose the information
2820 about possibly different redirections that led to the duplicates.)
2822 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2823 systems where poll() doesn't work, in particular OS X.
2825 PH/10 Added more information to debugging output for retry time not reached.
2827 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2828 operations in malware.c.
2830 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2833 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2834 syslog_duplication set false, log lines that would normally be written
2835 both the the main log and to the reject log were not written to syslog at
2838 PH/14 In the default configuration, change the use of "message" in ACL warn
2839 statements to "add_header".
2841 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2842 not followed by a command (e.g. "seen endif").
2844 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2845 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2848 PH/17 Added extra conditions to the default value of delay_warning_condition
2852 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2853 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2854 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2857 The Auto-Submitted: and various List- headers are standardised, whereas I
2858 don't think Precedence: ever was.
2860 PH/18 Refactored debugging code in route_finduser() to show more information,
2861 in particular, the error code if getpwnam() issues one.
2863 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2864 This is apparently needed in addition to the PH/07 change above to avoid
2865 any possible encoding problems.
2867 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2868 but not after initializing Perl.
2870 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2871 output them only if debugging. By default they are written stderr,
2872 apparently, which is not desirable.
2874 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2877 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2880 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2882 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2883 authenticated or an ident call has been made. Suppress the default
2884 values for $authenticated_id and $authenticated_sender (but permit -oMai
2885 and -oMas) when testing with -bh.
2887 PH/24 Re-jigged the order of the tests in the default configuration so that the
2888 tests for valid domains and recipients precede the DNS black list and CSA
2889 tests, on the grounds that those ones are more expensive.
2891 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2892 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2893 This bug exists in every version of Exim that I still have, right back to
2896 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2897 However, an attempt to turn on case-sensitivity in a regex key by
2898 including (?-i) didn't work because the subject string was already
2899 lowercased, and the effects were non-intuitive. It turns out that a
2900 one-line patch can be used to allow (?-i) to work as expected.
2906 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2907 other effects) broke the use of negated acl sub-conditions.
2909 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2912 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2913 "Deny" causes Exim to reject the incoming connection with a 554 error.
2914 Unfortunately, if there is a major crisis, such as a disk failure,
2915 tcp-wrappers gives "deny", whereas what one would like would be some
2916 kind of temporary error. A kludge has been added to help with this.
2917 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2918 554 error is used if errno is still zero or contains ENOENT (which occurs
2919 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2922 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2924 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2925 errors. Otherwise a message that provokes a temporary error (when other
2926 messages do not) can cause a whole host to time out.
2928 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2929 addresses were routed directly to files or pipes from a redirect router.
2930 File deliveries just didn't batch; pipe deliveries might have suffered
2933 PH/06 A failure to get a lock for a hints database would erroneously always say
2934 "Failed to get write lock", even when it was really a read lock.
2936 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2937 of 0600. This has been changed to use the value of the lockfile_mode
2938 option (which defaults to 0600).
2940 PH/08 Applied small patch from the Sieve maintainer.
2942 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2943 folder from quota calculations, a direct delivery into this folder messed
2944 up the contents of the maildirsize file. This was because the regex was
2945 used only to exclude .Trash (or whatever) when the size of the mailbox
2946 was calculated. There was no check that a delivery was happening into an
2947 excluded directory. This bug has been fixed by ignoring all quota
2948 processing for deliveries into excluded directories.
2950 PH/10 Added the maildirfolder_create_regex option to appendfile.
2956 PH/01 The code for finding all the local interface addresses on a FreeBSD
2957 system running IPv6 was broken. This may well have applied to all BSD
2958 systems, as well as to others that have similar system calls. The broken
2959 code found IPv4 interfaces correctly, but gave incorrect values for the
2960 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2961 that it would not match correctly against @[] and not recognize the IPv6
2964 PH/02 The ipliteral router was not recognizing addresses of the form user@
2965 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2967 PH/03 Added disable_ipv6.
2969 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2970 decoded form, because it is most often used to construct To: headers
2971 lines in autoreplies, and the decoded form may well be syntactically
2972 invalid. However, $reply_address has leading white space removed, and all
2973 newlines turned into spaces so that the autoreply transport does not
2976 PH/05 If group was specified without a user on a router, and no group or user
2977 was specified on a transport, the group from the router was ignored.
2979 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2980 for visible compile-time settings that can be used to change these
2981 numbers, for those that want even more. Backwards compatibility with old
2982 spool files has been maintained. However, going back to a previous Exim
2983 release will lost any variables that are in spool files.
2985 PH/07 Two small changes when running in the test harness: increase delay when
2986 passing a TCP/IP connection to a new process, in case the original
2987 process has to generate a bounce, and remove special handling of
2988 127.0.0.2 (sic), which is no longer necessary.
2990 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2991 be the same on different OS.
2993 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2996 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2997 whether --show-vars was specified or not
2999 JJ/02 exipick: Added support for new ACL variable spool format introduced
3002 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3003 syntactically invalid From: or Reply-to: line, and a filter used this to
3004 generate an autoreply, and therefore failed to obtain an address for the
3005 autoreply, Exim could try to deliver to a non-existent relative file
3006 name, causing unrelated and misleading errors. What now happens is that
3007 it logs this as a hard delivery error, but does not attempt to create a
3010 PH/11 The exinext utility has a -C option for testing purposes, but although
3011 the given file was scanned by exinext itself; it wasn't being passed on
3012 when Exim was called.
3014 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3015 an end-of-file indication when reading a command response.
3017 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3018 compiled. In many other places in Exim, IPv6 addresses are always
3019 recognized, so I have changed this. It also means that IPv4 domain
3020 literals of the form [IPV4:n.n.n.n] are now always recognized.
3022 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3023 used if the router is not running as root, for example, when verifying at
3024 ACL time, or when using -bh. The debugging output from this situation was
3025 non-existent - all you got was a failure to exec. I have made two
3028 (a) Failures to set uid/gid, the current directory, or a process leader
3029 in a subprocess such as that created by queryprogram now generate
3030 suitable debugging output when -d is set.
3032 (b) The queryprogram router detects when it is not running as root,
3033 outputs suitable debugging information if -d is set, and then runs
3034 the subprocess without attempting to change uid/gid.
3036 PH/15 Minor change to Makefile for building test_host (undocumented testing
3039 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3040 additional section of a DNS packet that returns MX or SRV records.
3041 Instead, it always explicitly searches for A/AAAA records. This avoids
3042 major problems that occur when a DNS server includes only records of one
3043 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3044 fixed another bug: if SRV records were looked up and the corresponding
3045 address records were *not* found in the additional section, the port
3046 values from the SRV records were lost.
3048 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3049 using the correct key (the original address) when searching the retry
3050 rules in order to find which one to use for generating the retry hint.
3052 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3053 adding the default one. Similarly, if it contains a Reply-To: header, the
3054 errors_reply_to option, if set, is not used.
3056 PH/19 When calculating a retry time, Exim used to measure the "time since
3057 failure" by looking at the "first failed" field in the retry record. Now
3058 it does not use this if it is later than than the arrival time of the
3059 message. Instead it uses the arrival time. This makes for better
3060 behaviour in cases where some deliveries succeed, thus re-setting the
3061 "first failed" field. An example is a quota failure for a huge message
3062 when small messages continue to be delivered. Without this change, the
3063 "time since failure" will always be short, possible causing more frequent
3064 delivery attempts for the huge message than are intended.
3065 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3067 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3068 $1, $2, $3) because the numerical variables can be reset during some
3069 expansion items (e.g. "match"), thereby losing the authentication data.
3071 PH/21 Make -bV show the size of off_t variables so that the test suite can
3072 decide whether to run tests for quotas > 2G.
3074 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3075 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3076 filecount value is greater than 2G or if a quota value is greater than 2G
3077 on a system where the size of off_t is not greater than 4, a panic error
3080 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3081 never match. The debug and -bh output now contains an explicit error
3082 message indicating a malformed IPv4 address or mask.
3084 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3085 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3086 PH/23 above applies.
3088 PH/25 Do not write to syslog when running in the test harness. The only
3089 occasion when this arises is a failure to open the main or panic logs
3090 (for which there is an explicit test).
3092 PH/26 Added the /no_tell option to "control=freeze".
3094 PH/27 If a host name lookup failed very early in a connection, for example, if
3095 the IP address matched host_lookup and the reverse lookup yielded a name
3096 that did not have a forward lookup, an error message of the form "no IP
3097 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3098 could be logged. Now it outputs the IP address instead of "NULL".
3100 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3101 allows the sender and the authenticated sender to be set when
3102 submitting a message from within Exim. Since child_open_exim() is
3103 documented for local_scan(), the new function should be too.
3105 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3106 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3107 results in an empty string is now treated as unset.
3109 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3111 PH/31 Added sender_verify_fail logging option.
3113 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3114 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3115 tidied the source and removed it altogether.
3117 PH/33 When a queue run was abandoned because the load average was too high, a
3118 log line was always written; now it is written only if the queue_run log
3119 selector is set. In addition, the log line for abandonment now contains
3120 information about the queue run such as the pid. This is always present
3121 in "start" and "stop" lines but was omitted from the "abandon" line.
3123 PH/34 Omit spaces between a header name and the colon in the error message that
3124 is given when verify = headers_syntax fails (if there are lots of them,
3125 the message gets confusing).
3127 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3128 names, as there are now some PTR records that contain slashes. This check
3129 is only to protect against broken name servers that fall over on strange
3130 characters, so the fact that it applies to all lookups doesn't matter.
3132 PH/36 Now that the new test suite is complete, we can remove some of the
3133 special code in Exim that was needed for the old test suite. For example,
3134 sorting DNS records because real resolvers return them in an arbitrary
3135 order. The new test suite's fake resolver always returns records in the
3138 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3139 bounces) except when queue_only is set, to avoid logging races between
3140 the different processes.
3142 PH/38 Panic-die if .include specifies a non-absolute path.
3144 PH/39 A tweak to the "H" retry rule from its user.
3146 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3147 a label. They prevented compilation on older perls.
3149 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3150 a warning to be raised on newish perls.
3152 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3153 on queue. Changes to match documented behaviour of showing count of
3154 messages matching specified criteria.
3156 PH/40 Changed the default ident timeout from 30s to 5s.
3158 PH/41 Added support for the use of login_cap features, on those BSD systems
3159 that have them, for controlling the resources used by pipe deliveries.
3161 PH/42 The content-scanning code uses fopen() to create files in which to put
3162 message data. Previously it was not paying any attention to the mode of
3163 the files. Exim runs with umask(0) because the rest of the code creates
3164 files with open(), and sets the required mode explicitly. Thus, these
3165 files were ending up world-writeable. This was not a big issue, because,
3166 being within the spool directory, they were not world-accessible. I have
3167 created a function called modefopen, which takes an additional mode
3168 argument. It sets umask(777), creates the file, chmods it to the required
3169 mode, then resets the umask. All the relevant calls to fopen() in the
3170 content scanning code have been changed to use this function.
3172 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3173 to 24 hours. This avoids potential overflow problems when processing G
3174 and H retry rules. I suspect nobody ever tinkers with this value.
3176 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3178 PH/45 When the plaintext authenticator is running as a client, the server's
3179 challenges are checked to ensure they are valid base64 strings. By
3180 default, the authentication attempt is cancelled if an invalid string is
3181 received. Setting client_ignore_invalid_base64 true ignores these errors.
3182 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3183 they are received. Thus, the responses can be made to depend on the
3184 challenges. If an invalid string is ignored, an empty string is placed in
3187 PH/46 Messages that are created by the autoreply transport now contains a
3188 References: header, in accordance with RFCs 2822 and 3834.
3190 PH/47 Added authenticated_sender_force to the smtp transport.
3192 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3194 PH/49 Installed latest patch from the Sieve maintainer.
3196 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3197 was also set, the appendfile transport was unnecessarily scanning a
3198 directory of message files (e.g. for maildir delivery) to find the count
3199 of files (along with the size), even though it did not need this
3200 information. It now does the scan only if it needs to find either the
3201 size of the count of files.
3203 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3205 PH/52 Two bugs concerned with error handling when the smtp transport is
3208 (i) Exim was not creating retry information for temporary errors given
3209 for individual recipients after the DATA command when the smtp transport
3210 was used in LMTP mode. This meant that they could be retried too
3211 frequently, and not timed out correctly.
3213 (ii) Exim was setting the flag that allows error details to be returned
3214 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3215 recipients that were returned after the DATA command.
3217 PH/53 This is related to PH/52, but is more general: for any failing address,
3218 when detailed error information was permitted to be returned to the
3219 sender, but the error was temporary, then after the final timeout, only
3220 "retry timeout exceeded" was returned. Now it returns the full error as
3221 well as "retry timeout exceeded".
3223 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3224 do this, and (what is worse) MTAs that accept it.
3226 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3227 will now be deprecated.
3229 PH/56 New os.c-cygwin from the Cygwin maintainer.
3231 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3232 formats (previously only available in exim formats via -bpr, -bpru,
3233 and -bpra. Now also available in native and exiqgrep formats)
3235 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3236 with very large, slow to parse queues
3238 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3240 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3242 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3243 responses to authentication challenges, though it was showing the
3244 challenges; (ii) I've removed the CR characters from the debug output for
3247 PH/58 Allow for the insertion of a newline as well as a space when a string
3248 is turned into more than one encoded-word during RFC 2047 encoding. The
3249 Sieve code now uses this.
3251 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3252 data_4xx, lost_connection, tls_required.
3254 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3255 message was being sent as an SMTP response.
3257 PH/61 Add -l and -k options to exicyclog.
3259 PH/62 When verifying, if an address was redirected to one new address, so that
3260 verification continued, and the new address failed or deferred after
3261 having set something in $address_data, the value of $address_data was not
3262 passed back to the ACL. This was different to the case when no
3263 redirection occurred. The value is now passed back in both cases.
3265 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3266 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3267 use login_cap.h, so on its own it isn't the distinguishing feature. The
3268 new name refers directly to the setclassresources() function.
3270 PH/65 Added configuration files for NetBSD3.
3272 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3274 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3275 is preferred over IPv4.
3277 PH/68 The bounce_return_message and bounce_return_body options were not being
3278 honoured for bounces generated during the reception of non-SMTP messages.
3279 In particular, this applied to messages rejected by the ACL. This bug has
3280 been fixed. However, if bounce_return_message is true and bounce_return_
3281 body is false, the headers that are returned for a non-SMTP message
3282 include only those that have been read before the error was detected.
3283 (In the case of an ACL rejection, they have all been read.)
3285 PH/69 The HTML version of the specification is now built in a directory called
3286 spec_html instead of spec.html, because the latter looks like a path with
3287 a MIME-type, and this confuses some software.
3289 PH/70 Catch two compiler warnings in sieve.c.
3291 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3292 function verify_get_ident() calls ip_connect() to connect a socket, but
3293 if the "connect()" function timed out, ip_connect() used to close the
3294 socket. However, verify_get_ident() also closes the socket later, and in
3295 between Exim writes to the log, which may get opened at this point. When
3296 the socket was closed in ip_connect(), the log could get the same file
3297 descriptor number as the socket. This naturally causes chaos. The fix is
3298 not to close the socket in ip_connect(); the socket should be closed by
3299 the function that creates it. There was only one place in the code where
3300 this was missing, in the iplookup router, which I don't think anybody now
3301 uses, but I've fixed it anyway.
3303 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3304 well as to direct DNS lookups. Otherwise the handling of names in host
3305 lists is inconsistent and therefore confusing.
3311 PH/01 Two changes to the default runtime configuration:
3313 (1) Move the checks for relay_from_hosts and authenticated clients from
3314 after to before the (commented out) DNS black list checks.
3316 (2) Add control=submission to the relay_from_hosts and authenticated
3317 clients checks, on the grounds that messages accepted by these
3318 statements are most likely to be submissions.
3320 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3322 (1) Generate an error if the third argument for the ${prvs expansion is
3325 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3328 (3) Reset the variables that are obtained from the first argument of
3329 ${prvscheck and used in the second argument before leaving the code,
3330 because their memory is reclaimed, so using them afterwards may do
3333 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3334 one (it's much easier than Tom thought :-).
3336 (5) Because of (4), we can now allow for the use of $prvscheck_result
3337 inside the third argument.
3339 PH/03 For some reason, the default setting of PATH when running a command from
3340 a pipe transport was just "/usr/bin". I have changed it to
3343 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3344 anything to be listed in the output from -bV.
3346 PH/05 When a filter generated an autoreply, the entire To: header line was
3347 quoted in the delivery log line, like this:
3349 => >A.N.Other <ano@some.domain> <original@ddress> ...
3351 This has been changed so that it extracts the operative address. There
3352 may be more than one such address. If so, they are comma-separated, like
3355 => >ano@some.domain,ona@other.domain <original@ddress> ...
3357 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3358 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3359 not being looked up in the rDNS to get a host name, Exim was showing the
3360 IP address twice in Received: lines, even though the IP addresses were
3361 identical. For example:
3363 Received: from [1.2.3.4] (helo=[1.2.3.4])
3365 However, if the real host name was known, it was omitting the HELO data
3366 if it matched the actual IP address. This has been tidied up so that it
3367 doesn't show the same IP address twice.
3369 PH/07 When both +timestamp and +memory debugging was on, the value given by
3370 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3371 called by the debug printing, thereby overwriting the timestamp buffer.
3372 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3374 PH/08 When the original message was included in an autoreply transport, it
3375 always said "this is a copy of the message, including all the headers",
3376 even if body_only or headers_only was set. It now gives an appropriate
3379 PH/09 Applied a patch from the Sieve maintainer which:
3381 o fixes some comments
3382 o adds the (disabled) notify extension core
3383 o adds some debug output for the result of if/elsif tests
3384 o points to the current vacation draft in the documentation
3385 and documents the missing references header update
3389 o fixes a bug in processing the envelope test (when testing
3390 multiple envelope elements, the last element determined the
3393 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3394 Electronic Mail") by including:
3396 Auto-submitted: auto-generated
3398 in the messages that it generates (bounce messages and others, such as
3399 warnings). In the case of bounce messages for non-SMTP messages, there was
3400 also a typo: it was using "Auto_submitted" (underscore instead of
3401 hyphen). Since every message generated by Exim is necessarily in response
3402 to another message, thes have all been changed to:
3404 Auto-Submitted: auto-replied
3406 in accordance with these statements in the RFC:
3408 The auto-replied keyword:
3410 - SHOULD be used on messages sent in direct response to another
3411 message by an automatic process,
3413 - MUST NOT be used on manually-generated messages,
3415 - MAY be used on Delivery Status Notifications (DSNs) and Message
3416 Disposition Notifications (MDNs),
3418 - MUST NOT be used on messages generated by automatic or periodic
3419 processes, except for messages which are automatic responses to
3422 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3423 to the default Received: header definition.
3425 PH/12 Added log selector acl_warn_skipped (default on).
3427 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3428 variables because (a) they are in the wrong storage pool and (b) even if
3429 they were copied, it wouldn't work properly because of the caching.
3431 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3432 checking when decoding. Apparently there are clients that generate
3433 overlong encoded strings. Why am I not surprised?
3435 PH/15 If the first argument of "${if match_address" was not empty, but did not
3436 contain an "@" character, Exim crashed. Now it writes a panic log message
3437 and treats the condition as false.
3439 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3441 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3442 extension "envelope-auth". The code is finished and in agreement with
3443 other implementations, but there is no documentation so far and in fact,
3444 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3445 not changing the active code.
3447 Print executed "if" and "elsif" statements when debugging is used. This
3448 helps a great deal to understand what a filter does.
3450 Document more things not specified clearly in RFC3028. I had all this
3451 sorted out, when out of a sudden new issues came to my mind. Oops."
3453 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3456 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3457 canonical form (as documented). However, after a host name lookup from
3458 the IP address, check_host() was doing a simple string comparison with
3459 addresses acquired from the DNS when checking that the found name did
3460 have the original IP as one of its addresses. Since any found IPv6
3461 addresses are likely to be in abbreviated form, the comparison could
3462 fail. Luckily, there already exists a function for doing the comparison
3463 by converting both addresses to binary, so now that is used instead of
3464 the text comparison.
3466 PH/20 There was another similar case to PH/19, when a complete host name was
3467 given in a host list; looking up its IP address could give an abbreviated
3468 form, whereas the current host's name might or might not be abbreviated.
3469 The same fix has been applied.
3475 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3476 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3479 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3480 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3482 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3484 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3485 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3486 Also, for "Auto-Submitted", treat anything other than "no" as
3487 non-personal, in accordance with RFC 3834. (Previously it treated
3488 anything starting "auto-" as non-personal.)
3490 TF/01 The control=submission/name=... option had a problem with syntax
3491 errors if the name included a slash character. The /name= option
3492 now slurps the rest of the string, so it can include any characters
3493 but it must come last in the list of options (after /sender_retain
3496 PH/05 Some modifications to the interface to the fake nameserver for the new
3504 TK/01 Added the "success_on_redirect" address verification option. See
3505 NewStuff for rationale and an example.
3507 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3509 PH/02 Patch to exigrep to allow it to work on syslog lines.
3511 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3512 fread() to skip over the body file's header line, because in Cygwin the
3513 header line is locked and is inaccessible.
3515 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3516 co-exist for some time) to make it clear that it is the Exim ID that is
3517 referenced, not the Message-ID: header line.
3519 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3520 string_format() function, because snprintf() does not exist on all
3523 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3524 ${stat: expansion item.
3526 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3527 protocol synchronization error", to keep the pedants happy.
3529 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3530 well as for IRIX systems, when gcc is being used. See the host.c source
3533 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3535 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3538 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3539 transport and to the smtp transport in LMTP mode.
3541 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3543 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3544 run a filter in a subprocess. This could lead to confusion in subsequent
3545 lookups in the parent process. There should also be a search_tidyup() at
3546 the end of the subprocess.
3548 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3549 only if the host matched helo_try_verify_hosts, which caused the
3550 verification to occur when the EHLO/HELO command was issued. The ACL just
3551 tested the remembered result. Now, if a previous verification attempt has
3552 not happened, "verify = helo" does it there and then.
3554 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3556 TK/03 Fix log output including CR from clamd.
3558 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3559 exist provoked a memory error which could cause a segfault.
3561 PH/15 Installed PCRE 6.2
3563 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3565 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3566 of the problem. Specifically, suggested +O2 rather than +O1 for the
3569 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3571 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3572 record, the port was not correctly logged when the outgoing_port log
3573 selector was set (it logged the transort's default port).
3575 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3576 fallback_hosts, and "hosts" in the smtp transport.
3578 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3579 host errors such as "Connection refused".
3581 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3582 authentication with radiusclient 0.4.9:
3584 - Error returned from rc_read_config was caught wrongly
3585 - Username/password not passed on to radius server due to wrong length.
3587 The presumption is that some radiusclient API changes for 4.51/PH/17
3588 were not taken care of correctly. The code is still untested by me (my
3589 Linux distribution still has 0.3.2 of radiusclient), but it was
3590 contributed by a Radius user.
3592 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3593 expanding the "port" option of the smtp transport.
3595 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3596 while reading a MIME header. Thanks to Tom Hughes for a patch.
3598 PH/24 Include config.h inside local_scan.h so that configuration settings are
3601 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3602 that in an ACL for RCPT (for example), you can examine exactly what was
3605 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3606 commands, but it was not correctly comparing the address with the actual
3607 client host address. Thus, it would show the EHLO address in Received:
3608 header lines when this was not necessary.
3610 PH/27 Added the % operator to ${eval:}.
3612 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3613 it should be ignoring failures (because with -C, for example, it has lost
3614 privilege). It wasn't ignoring creation failures other than "already
3617 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3620 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3621 because an input line was too long, either on its own, or by virtue of
3622 too many continuations, the temporary file was not being removed, and the
3623 return code was incorrect.
3625 PH/31 Missing "BOOL" in function definition in filtertest.c.
3627 PH/32 Applied Sieve patches from the maintainer.
3629 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3631 PH/33 Added "verify = not_blind".
3633 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3634 Local/Makefile (with some defaults set). These are used in built scripts
3635 such as exicyclog, but they have never been used in the exim_install
3636 script (though there are many overriding facilities there). I have
3637 arranged that the exim_install script now takes note of these two
3640 PH/35 Installed configuration files for Dragonfly.
3642 PH/36 When a locally submitted message by a trusted user did not contain a
3643 From: header, and the sender address was obtained from -f or from an SMTP
3644 MAIL command, and the trusted user did not use -F to supply a sender
3645 name, $originator_name was incorrectly used when constructing a From:
3646 header. Furthermore, $originator_name was used for submission mode
3647 messages from external hosts without From: headers in a similar way,
3648 which is clearly wrong.
3650 PH/37 Added control=suppress_local_fixups.
3652 PH/38 When log_selector = +received_sender was set, and the addition of the
3653 sender made the log line's construction buffer exactly full, or one byte
3654 less than full, an overflow happened when the terminating "\n" was
3657 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3658 when the result of a list match is failure because a DNS lookup failed.
3660 PH/40 RM_COMMAND is now used in the building process.
3662 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3663 the "build-* directories that it finds.
3665 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3666 address in a domain literal was a prefix of an interface address.
3668 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3669 when verifying a sender address, unless rewrite_headers is false.
3671 PH/44 Wrote a long comment about why errors_to addresses are verified as
3672 recipients, not senders.
3674 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3675 the ratelimit ACL was added.
3677 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3679 PH/46 Added extra information about PostgreSQL errors to the error string.
3681 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3682 avoiding the need to install special zones in a real server. This is
3683 backwards compatible; if it can't find the fake resolver, it drops back.
3684 Thus, both old and new test suites can be run.
3686 TF/02 Added util/ratelimit.pl
3688 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3689 clock is set back in time.
3691 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3692 Candler <B.Candler@pobox.com>.
3694 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3695 OK for addresses that are the result of redirections.
3697 PH/48 A number of further additions for the benefit of the new test suite,
3698 including a fake gethostbyname() that interfaces to the fake DNS resolver
3701 TF/06 The fix for widen_domains has also been applied to qualify_single and
3702 search_parents which are the other dnslookup options that can cause
3705 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3708 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3710 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3711 service" if it encountered an MX record with an empty target hostname.
3712 The message is now "an MX or SRV record indicated no SMTP service".
3714 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3715 if the DNS of the sending site is misconfigured. This is quite a
3716 common situation. This change restores the behaviour of treating a
3717 helo verification defer as a failure.
3719 PH/51 If self=fail was set on a router, the bounce message did not include the
3720 actual error message.
3726 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3728 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3729 command itself ended in error, the underlying message about the transport
3730 filter timeout was being overwritten with the pipe command error. Now the
3731 underlying error message should be appended to the second error message.
3733 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3735 PH/02 Reduce the amount of output that "make" produces by default. Full output
3736 can still be requested.
3738 PH/03 The warning log line about a condition test deferring for a "warn" verb
3739 was being output only once per connection, rather than after each
3740 occurrence (because it was using the same function as for successful
3741 "warn" verbs). This seems wrong, so I have changed it.
3743 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3744 it should not have, which might have caused a crash in the right
3745 circumstances, but probably never did.
3747 PH/04 Installed a modified version of Tony Finch's patch to make submission
3748 mode fix the return path as well as the Sender: header line, and to
3749 add a /name= option so that you can make the user's friendly name appear
3752 TF/03 Added the control = fakedefer ACL modifier.
3754 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3755 Mark Lowes for thorough testing.
3757 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3759 TK/03 Merged latest SRS patch from Miles Wilton.
3761 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3762 with the definition in sysexits.h (which is #included earlier).
3763 Fortunately, Exim does not actually use EX_OK. The code used to try to
3764 preserve the sysexits.h value, by assuming that macro definitions were
3765 scanned for macro replacements. I have been disabused of this notion,
3766 so now the code just undefines EX_OK before #including unistd.h.
3768 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3769 in the smtp transport. When a block could not be written in a single
3770 write() function, the timeout was being re-applied to each part-write.
3771 This seems wrong - if the receiver was accepting one byte at a time it
3772 would take for ever. The timeout is now adjusted when this happens. It
3773 doesn't have to be particularly precise.
3775 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3776 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3778 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3779 without a domain if the check to <postmaster@domain> fails.
3781 SC/01 Eximstats: added -xls and the ability to specify output files
3782 (patch written by Frank Heydlauf).
3784 SC/02 Eximstats: use FileHandles for outputting results.
3786 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3788 SC/04 Eximstats: fixed display of large numbers with -nvr option
3790 SC/05 Eximstats: fixed merging of reports with empty tables.
3792 SC/06 Eximstats: added the -include_original_destination flag
3794 SC/07 Eximstats: removed tabs and trailing whitespace.
3796 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3798 TK/06 MBOX spool code: Add real "From " MBOX separator line
3799 so the .eml file is really in mbox format (even though
3800 most programs do not really care). Patch from Alex Miller.
3802 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3803 The latter is generated from $received_to and is only set if the
3804 message has one envelope recipient. SA can use these headers,
3805 obviously out-of-the-box. Patch from Alex Miller.
3807 PH/08 The ${def test on a variable was returning false if the variable's
3808 value was "0", contrary to what the specification has always said!
3809 The result should be true unless the variable is empty.
3811 PH/09 The syntax error of a character other than { following "${if
3812 def:variable_name" (after optional whitespace) was not being diagnosed.
3813 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3814 accidental colon was present, for example, could give incorrect results.
3816 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3817 result is used (not including appendfile, where other changes are about
3820 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3821 This involved changing a lot of size variables from int to off_t. It
3822 should work with maildirs and everything.
3824 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3825 spamd dying while we are connected to it.
3827 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3830 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3831 function for BDB 4.3.
3833 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3835 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3836 the sources for PCRE have been re-arranged and more files are now
3839 PH/15 The code I had for printing potentially long long variables in PH/11
3840 above was not the best (it lost precision). The length of off_t variables
3841 is now inspected at build time, and an appropriate printing format (%ld
3842 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3843 to be "long long int" or "long int". This is needed for the internal
3844 formatting function string_vformat().
3846 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3847 the configuration file to be ":syslog", then the script "guesses" where
3848 the logs files are, rather than using the compiled in default. In our
3849 case the guess is not the same as the compiled default, so the script
3850 suddenly stopped working when I started to use syslog. The patch checks
3851 to see if log_file_path is "". If so, it attempts to read it from exim
3852 with no configuration file to get the compiled in version, before it
3853 falls back to the previous guessing code."
3855 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3856 implementing BATV in an Exim configuration. See NewStuff for the gory
3859 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3860 Makefile that are specific to HP-UX.
3862 PH/18 If the "use_postmaster" option was set for a recipient callout together
3863 with the "random" option, the postmaster address was used as the MAIL
3864 FROM address for the random test, but not for the subsequent recipient
3865 test. It is now used for both.
3867 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3868 patch removes a few documentation additions to RFC 3028, because the
3869 latest draft now contains them. It adds the new en;ascii-case comparator
3870 and a new error check for 8bit text in MIME parts. Comparator and
3871 require names are now matched exactly. I enabled the subaddress
3872 extension, but it is not well tested yet (read: it works for me)."
3874 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3875 rework some of the code of TK/09 above to avoid the hardwired use of
3876 "%lld" and "long long". Replaced the call to snprintf() with a call to
3879 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3880 records point to non-existent hosts", "retry timeout exceeded", and
3881 "retry time not reached for any host after a long failure period".
3883 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3884 experimental DomainKeys support:
3886 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3887 (2) On an error such as an illegally used "control", the wrong name for
3888 the control was given.
3890 These problems did NOT occur unless DomainKeys support was compiled.
3892 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3894 PH/24 Added ${if match_ip condition.
3896 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3897 databases so that it will be absolutely obvious if a crash occurs in the
3898 DB library. This is a regular occurrence (often caused by mis-matched
3901 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3902 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3903 picked up on a user's system that detects such things. There doesn't seem
3904 to be a gcc warning option for this - only an attribute that has to be
3905 put on the function's prototype. It seems that in Fedora Core 4 they have
3906 set this on a number of new functions. No doubt there will be more in due
3909 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3910 specify a transport. However, if an address that was verified by such a
3911 router was the subject of a callout, Exim crashed because it tried to
3912 read the rcpt_include_affixes from the non-existent transport. Now it
3913 just assumes that the setting of that option is false. This bug was
3914 introduced by 4.51/PH/31.
3916 PH/28 Changed -d+all to exclude +memory, because that information is very
3917 rarely of interest, but it makes the output a lot bigger. People tend to
3918 do -d+all out of habit.
3920 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3921 code in os-type was giving problems when libc.so lives in lib64, like on
3924 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3925 aren't the modern standard, and it seems that some systems' include files
3926 don't always have them. Exim was already checking for some of the newer
3927 ones like T_AAAA, and defining it itself. I've added checks for all the
3928 record types that Exim uses.
3930 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3931 not automatically generating a new one, as it is supposed to. This
3932 prevented TLS from working. If the file did exist, but contained invalid
3933 data, a new version was generated, as expected. It was only the case of a
3934 non-existent file that was broken.
3936 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3937 with a change in libdomainkeys > 0.64.
3939 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3940 from DNS. If the selector record carries the flag, it now has
3941 precedence over the domain-wide flag.
3943 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3945 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3946 the use of a port name that isn't defined in /etc/services) occurred, the
3947 message was deferred as in a normal delivery, and thus remained on the
3948 spool, instead of being failed because of the mua_wrapper setting. This
3949 is now fixed, and I tidied up some of the mua_wrapper messages at the
3952 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3953 the messages in a hash of arrays rather than using individual hashes.
3954 This is a bit cleaner and results in dramatic memory savings, albeit
3955 at a slight CPU cost.
3957 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3958 as requested by Marc Sherman.
3960 SC/10 Eximstats: added histograms for user specified patterns as requested
3963 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3965 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3966 fopen() in the content-scanning modules that did not already have it.
3972 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3973 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3975 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3977 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3979 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3980 to test Sieve filters that use "vacation".
3982 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3983 that changes the way the GnuTLS parameters are stored in the cache file.
3984 The new format can be generated externally. For backward compatibility,
3985 if the data in the cache doesn't make sense, Exim assumes it has read an
3986 old-format file, and it generates new data and writes a new file. This
3987 means that you can't go back to an older release without removing the
3990 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3991 work if there are any delivery delays because "one_time" forces the
3992 parent to be marked "delivered", so its unseen clone is never tried
3993 again. For this reason, Exim now forbids the simultaneous setting of
3996 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3997 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3998 entry if you want to know the details. The fix, however, neglected to
3999 consider the case when local delivery batching is involved. The test for
4000 "previously delivered" was not happening when checking to see if an
4001 address could be batched with a previous (undelivered) one; under
4002 certain circumstances this could lead to multiple deliveries to the same
4005 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4006 in its include files, and this causes problems building Exim.
4008 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4009 header_syntax) but Exim was just ignoring anything given after a slash.
4010 In particular, this caused confusion with an attempt to use "verify =
4011 reverse_host_lookup/defer_ok". An error is now given when options are
4012 supplied for verify items that do not have them. (Maybe reverse_host_
4013 lookup should have a defer_ok option, but that's a different point.)
4015 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4016 defined by RFC 821) to 2048, because there were problems with some AUTH
4017 commands, and RFC 1869 says the size should be increased for extended
4018 SMTP commands that take arguments.
4020 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4023 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4024 "unknown" error; now it says that the functionality isn't in the binary.
4026 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4027 an address' error message when a string expansion fails (syntax or
4028 whatever). Otherwise the password may appear in the log. Following change
4029 PH/42 below, there is no longer a chance of it appearing in a bounce
4032 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4034 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4035 of its addresses was ever tried. (Bugzilla bug #2.)
4037 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4038 the result incorrectly in the debug output. (It correctly added a newline
4039 to what was transported.)
4041 TF/01 Added $received_time.
4043 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4044 commented out examples of how to interface to a virus scanner and to
4045 SpamAssassin. Also added commented examples of av_scanner and
4046 spamd_address settings.
4048 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4049 and controls are allowed in which ACLs. There were a couple of minor
4050 errors. Some of the entries in the conditions table (which is a table of
4051 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4052 negation of where the condition IS allowed.
4054 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4056 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4057 header file does not have a version number, so I've had to invent a new
4058 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4059 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4060 radiusclient), but it was contributed by a Radius user.
4062 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4063 files or directories, for OpenSSL.
4065 PH/19 When an Exim process that is running as root has to create an Exim log
4066 file, it does so in a subprocess that runs as exim:exim so as to get the
4067 ownership right at creation (otherwise, other Exim processes might see
4068 the file with the wrong ownership). There was no test for failure of this
4069 fork() call, which would lead to the process getting stuck as it waited
4070 for a non-existent subprocess. Forks do occasionally fail when resources
4071 run out. I reviewed all the other calls to fork(); they all seem to check
4074 PH/20 When checking for unexpected SMTP input at connect time (before writing
4075 the banner), Exim was not dealing correctly with a non-positive return
4076 from the read() function. If the client had disconnected by this time,
4077 the result was a log entry for a synchronization error with an empty
4078 string after "input=" when read() returned zero. If read() returned -1
4079 (an event I could not check), uninitialized data bytes were printed.
4080 There were reports of junk text (parts of files, etc) appearing after
4083 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4085 PH/22 Added support for macro redefinition, and (re)definition in between
4086 driver and ACL definitions.
4088 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4089 forgetting to use the resulting value; it was using the unexpanded value.
4091 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4092 hadn't been configured. The fix is from Juergen Kreileder, who
4093 understands it better than I do:
4095 "Here's what I see happening with three configured cyrus_sasl
4096 authenticators configured (plain, login, cram-md5):
4098 On startup auth_cyrus_sasl_init() gets called for each of these.
4099 This means three calls to sasl_listmech() without a specified mech_list.
4100 => SASL tests which mechs of all available mechs actually work
4101 => three warnings about OTP not working
4102 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4104 With the patch, sasl_listmech() also gets called three times. But now
4105 SASL's mech_list option is set to the server_mech specified in the the
4106 authenticator. Or in other words, the answer from sasl_listmech()
4107 gets limited to just the mech you're testing for (which is different
4109 => the return list contains just 'plain' or 'login', 'cram-md5' or
4110 nothing depending on the value of ob->server_mech.
4112 I've just tested the patch: Authentication still works fine,
4113 unavailable mechs specified in the exim configuration are still
4114 caught, and the auth.log warnings about OTP are gone."
4116 PH/25 When debugging is enabled, the contents of the command line are added
4117 to the debugging output, even when log_selector=+arguments is not
4120 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4121 answer is "GNU", and only if the return is "GNU/something" is the answer
4124 PH/27 $acl_verify_message is now set immediately after the failure of a
4125 verification in an ACL, and so is available in subsequent modifiers. In
4126 particular, the message can be preserved by coding like this:
4128 warn !verify = sender
4129 set acl_m0 = $acl_verify_message
4131 Previously, $acl_verify_message was set only while expanding "message"
4132 and "log_message" when a very denied access.
4134 PH/28 Modified OS/os.c-Linux with
4136 -#ifndef OS_LOAD_AVERAGE
4137 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4139 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4140 nomenclature these days.)
4142 PH/29 Installed patch from the Sieve maintainer that adds the options
4143 sieve_useraddress and sieve_subaddress to the redirect router.
4145 PH/30 In these circumstances:
4146 . Two addresses routed to the same list of hosts;
4147 . First host does not offer TLS;
4148 . First host accepts first address;
4149 . First host gives temporary error to second address;
4150 . Second host offers TLS and a TLS session is established;
4151 . Second host accepts second address.
4152 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4153 and peerdn, if requested) that were in fact used only for the second
4156 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4157 attention to any local part prefix or suffix that was matched by the
4158 router that accepted the address. It now behaves in the same way as it
4159 does for delivery: the affixes are removed from the local part unless
4160 rcpt_include_affixes is set on the transport.
4162 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4163 timeout during the DATA phase of an incoming message.
4165 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4166 applied a patch sanctioned by the Sieve maintainer.
4168 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4169 the uid or gid is negative. A case of a negative gid caused this to be
4170 noticed. The fix allows for either to be negative.
4172 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4173 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4176 PH/36 The change PH/12 above was broken. Fixed it.
4178 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4179 the grounds that routing the same address twice would always produce the
4180 same answer. This might have been true once, but it is certainly no
4181 longer true now. Routing a child address may depend on the previous
4182 routing that produced that child. Some complicated redirection strategies
4183 went wrong when messages had multiple recipients, and made Exim's
4184 behaviour dependent on the order in which the addresses were given.
4186 I have moved the duplicate checking until after the routing is complete.
4187 Exim scans the addresses that are assigned to local and remote
4188 transports, and removes any duplicates. This means that more work will be
4189 done, as duplicates will always all be routed, but duplicates are
4190 presumably rare, so I don't expect this is of any significance.
4192 For deliveries to pipes, files, and autoreplies, the duplicate checking
4193 still happens during the routing process, since they are not going to be
4196 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4197 It corrects a timeout issue with spamd. This is Ian's comment: "The
4198 background is that sometimes spamd either never reads data from a
4199 connection it has accepted, or it never writes response data. The exiscan
4200 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4201 blindly assumes that writes won't block so it may never time out."
4203 PH/39 Allow G after quota size as well as K and M.
4205 PH/40 The value set for $authenticated_id in an authenticator may not contain
4206 binary zeroes or newlines because the value is written to log lines and
4207 to spool files. There was no check on this. Now the value is run through
4208 the string_printing() function so that such characters are converted to
4209 printable escape sequences.
4211 PH/41 $message_linecount is a new variable that contains the total number of
4212 lines in the message. Compare $body_linecount, which is the count for the
4215 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4216 bounce and delay warning messages, except in certain special cases, which
4219 (a) An SMTP error message from a remote host;
4220 (b) A message specified in a :fail: redirection;
4221 (c) A message specified in a "fail" command in a system filter;
4222 (d) A message specified in a FAIL return from the queryprogram router;
4223 (e) A message specified by the cannot_route_message router option.
4225 In these cases only, Exim does include the error details in bounce and
4226 warning messages. There are also a few cases where bland messages such
4227 as "unrouteable address" or "local delivery error" are given.
4229 PH/43 $value is now also set for the "else" part of a ${run expansion.
4231 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4232 being worked on, but at least Exim now implements the latest version to
4235 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4236 process to complete was treated as a delivery failure, a timeout while
4237 writing the message to the pipe was logged, but erroneously treated as a
4238 successful delivery. Such timeouts include transport filter timeouts. For
4239 consistency with the overall process timeout, these timeouts are now
4240 treated as errors, giving rise to delivery failures by default. However,
4241 there is now a new Boolean option for the pipe transport called
4242 timeout_defer, which, if set TRUE, converts the failures into defers for
4243 both kinds of timeout. A transport filter timeout is now identified in
4246 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4247 systems where "make" and "gmake" are different, calling "gmake" at top
4248 level broke things. I've arranged for the value of $(MAKE) to be passed
4249 from the Makefile to this script so that it can call the same version of
4253 A note about Exim versions 4.44 and 4.50
4254 ----------------------------------------
4256 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4257 changes of various kinds. As a consequence, a big documentation update was
4258 needed. This delayed the release for rather longer than seemed good, especially
4259 in the light of a couple of (minor) security issues. Therefore, the changes
4260 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4261 release. So 4.44 and 4.50 are in effect two different branches that both start
4264 I have left the 4.50 change log unchanged; it contains all the changes since
4265 4.43. The change log for 4.44 is below; many of its items are identical to
4266 those for 4.50. This seems to be the most sensible way to preserve the
4267 historical information.
4273 1. Minor wording change to the doc/README.SIEVE file.
4275 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4276 computation of the current number of files was incorrect.
4278 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4279 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4282 4. Give more explanation in the error message when the command for a transport
4283 filter fails to execute.
4285 5. There are several places where Exim runs a non-Exim command in a
4286 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4287 was being done only for the command run by the queryprogram router. It is
4288 now done for all such subprocesses. The other cases are: ${run, transport
4289 filters, and the commands run by the lmtp and pipe transports.
4291 6. Added CONFIGURE_GROUP build-time option.
4293 7. Some older OS have a limit of 256 on the maximum number of file
4294 descriptors. Exim was using setrlimit() to set 1000 as a large value
4295 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4296 systems. I've change it so that if it can't get 1000, it tries for 256.
4298 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4299 was an oversight, and furthermore, ever since the addition of extra
4300 controls (e.g. 4.43/32), the checks on when to allow different forms of
4301 "control" were broken. There should now be diagnostics for all cases when a
4302 control that does not make sense is encountered.
4304 9. Added the /retain_sender option to "control=submission".
4306 10. $recipients is now available in the predata ACL (oversight).
4308 11. Tidy the search cache before the fork to do a delivery from a message
4309 received from the command line. Otherwise the child will trigger a lookup
4310 failure and thereby defer the delivery if it tries to use (for example) a
4311 cached ldap connection that the parent has called unbind on.
4313 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4314 of $address_data from the recipient verification was clobbered by the
4315 sender verification.
4317 13. The value of address_data from a sender verification is now available in
4318 $sender_address_data in subsequent conditions in the ACL statement.
4320 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4322 15. Added a new option "connect=<time>" to callout options, to set a different
4325 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4326 was its contents. (It was OK if the option was not defined at all.)
4328 17. A "Completed" log line is now written for messages that are removed from
4329 the spool by the -Mrm option.
4331 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4332 information about exactly what failed.
4334 19. Added -dd to debug only the daemon process.
4336 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4337 handles timeouts, both on the server side and network timeouts. Renamed the
4338 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4340 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4341 It is now set to "smtps".
4343 22. $host_address is now set to the target address during the checking of
4344 ignore_target_hosts.
4346 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4347 being passed; this would have caused $sender_host_name to have been used if
4348 matching the list had actually called for a host name (not very likely,
4349 since this list is usually IP addresses). A host name is now passed as
4352 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4353 code that specifies a non-restarting handler (typically sigaction() in
4354 modern systems) in an attempt to fix a rare and obscure crash bug.
4356 25. Narrowed the window for a race in the daemon that could cause it to ignore
4357 SIGCHLD signals. This is not a major problem, because they are used only to
4358 wake it up if nothing else does.
4360 26. A malformed maildirsize file could cause Exim to calculate negative values
4361 for the mailbox size or file count. Odd effects could occur as a result.
4362 The maildirsize information is now recalculated if the size or filecount
4365 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4366 support for a long time. Removed HAVE_SYS_VFS_H.
4368 28. Installed the latest version of exipick from John Jetmore.
4370 29. In an address list, if the pattern was not a regular expression, an empty
4371 subject address (from a bounce message) matched only if the pattern was an
4372 empty string. Non-empty patterns were not even tested. This was the wrong
4373 because it is perfectly reasonable to use an empty address as part of a
4374 database query. An empty address is now tested by patterns that are
4375 lookups. However, all the other forms of pattern expect the subject to
4376 contain a local part and a domain, and therefore, for them, an empty
4377 address still always fails if the pattern is not itself empty.
4379 30. Exim went into a mad DNS loop when attempting to do a callout where the
4380 host was specified on an smtp transport, and looking it up yielded more
4381 than one IP address.
4383 31. Re-factored the code for checking spool and log partition space into a
4384 function that finds that data and another that does the check. The former
4385 is then used to implement four new variables: $spool_space, $log_space,
4386 $spool_inodes, and $log_inodes.
4388 32. The RFC2047 encoding function was originally intended for short strings
4389 such as real names; it was not keeping to the 75-character limit for
4390 encoded words that the RFC imposes. It now respects the limit, and
4391 generates multiple encoded words if necessary. To be on the safe side, I
4392 have increased the buffer size for the ${rfc2047: expansion operator from
4395 33. It is now permitted to omit both strings after an "if" condition; if the
4396 condition is true, the result is "true". As before, when the second string
4397 is omitted, a false condition yields an empty string. This makes it less
4398 cumbersome to write custom ACL and router conditions.
4400 34. Failure to deliver a bounce message always caused it to be frozen, even if
4401 there was an errors_to setting on the router. The errors_to setting is now
4404 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4405 canonical form (fully expanded) before being placed in
4406 $sender_host_address.
4408 36. The table in the code that translates DNS record types into text (T_A to
4409 "A" for instance) was missing entries for NS and CNAME. It is just possible
4410 that this could have caused confusion if both these types were looked up
4411 for the same domain, because the text type is used as part of Exim's
4412 per-process caching. But the chance of anyone hitting this buglet seems
4415 37. The dnsdb lookup has been extended in a number of ways.
4417 (1) There is a new type, "zns", which walks up the domain tree until it
4418 finds some nameserver records. It should be used with care.
4420 (2) There is a new type, "mxh", which is like "mx" except that it returns
4421 just the host names, not the priorities.
4423 (3) It is now possible to give a list of domains (or IP addresses) to be
4424 looked up. The behaviour when one of the lookups defers can be
4425 controlled by a keyword.
4427 (4) It is now possible to specify the separator character for use when
4428 multiple records are returned.
4430 38. The dnslists ACL condition has been extended: it is now possible to supply
4431 a list of IP addresses and/or domains to be looked up in a particular DNS
4434 39. Added log_selector=+queue_time_overall.
4436 40. When running the queue in the test harness, wait just a tad after forking a
4437 delivery process, to get repeatability of debugging output.
4439 41. Include certificate and key file names in error message when GnuTLS fails
4440 to set them up, because the GnuTLS error message doesn't include the name
4441 of the failing file when there is a problem reading it.
4443 42. Allow both -bf and -bF in the same test run.
4445 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4447 44. The "Exiscan patch" is now merged into the mainline Exim source.
4449 45. Sometimes the final signoff response after QUIT could fail to get
4450 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4451 < 0 before doing a fflush(). This bug looks as though it goes back to the
4452 introduction of TLS in release 3.20, but "sometimes" must have been rare
4453 because the tests only now provoked it.
4455 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4456 (this can affect the format of dates).
4458 47. exim_tidydb, when checking for the continued existence of a message for
4459 which it has found a message-specific retry record, was not finding
4460 messages that were in split spool directories. Consequently, it was
4461 deleting retry records that should have stayed in existence.
4463 48. Steve fixed some bugs in eximstats.
4465 49. The SPA authentication driver was not abandoning authentication and moving
4466 on to the next authenticator when an expansion was forced to fail,
4467 contradicting the general specification for all authenticators. Instead it
4468 was generating a temporary error. It now behaves as specified.
4470 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4471 (the order specifies the preference for clients). The order is now AES256,
4472 AES128, 3DES, ARCFOUR128.
4474 51. Small patch to Sieve code - explicitly set From: when generating an
4477 52. Exim crashed if a remote delivery caused a very long error message to be
4478 recorded - for instance if somebody sent an entire SpamAssassin report back
4479 as a large number of 550 error lines. This bug was coincidentally fixed by
4480 increasing the size of one of Exim's internal buffers (big_buffer) that
4481 happened as part of the Exiscan merge. However, to be on the safe side, I
4482 have made the code more robust (and fixed the comments that describe what
4485 53. Now that there can be additional text after "Completed" in log lines (if
4486 the queue_time_overall log selector is set), a one-byte patch to exigrep
4487 was needed to allow it to recognize "Completed" as not the last thing in
4490 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4491 patch that reportedly fixes this has been added. I am not expert enough to
4492 create a test for it. This is what the patch creator wrote:
4494 "I found a little strange behaviour of ldap code when working with
4495 Windows 2003 AD Domain, where users was placed in more than one
4496 Organization Units. When I tried to give exim partial DN, the exit code
4497 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4498 But simultaneously result of request was absolutely normal ldap result,
4499 so I produce this patch..."
4501 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4502 REFERENCE, so I have modified the code to exclude the patch when that macro
4505 55. Some experimental protocols are using DNS PTR records for new purposes. The
4506 keys for these records are domain names, not reversed IP addresses. The
4507 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4508 leaves it alone. Component reversal etc. now happens only for IP addresses.
4511 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4513 57. Double the size of the debug message buffer (to 2048) so that more of very
4514 long debug lines gets shown.
4516 58. The exicyclog utility now does better if the number of log files to keep
4517 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4519 59. Two changes related to the smtp_active_hostname option:
4521 (1) $smtp_active_hostname is now available as a variable.
4522 (2) The default for smtp_banner uses $smtp_active_hostname instead
4523 of $primary_hostname.
4525 60. The host_aton() function is supposed to be passed a string that is known
4526 to be a valid IP address. However, in the case of IPv6 addresses, it was
4527 not checking this. This is a hostage to fortune. Exim now panics and dies
4528 if the condition is not met. A case was found where this could be provoked
4529 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4530 components; fortuitously, this particular loophole had already been fixed
4531 by change 4.50/55 above.
4533 If there are any other similar loopholes, the new check in host_aton()
4534 itself should stop them being exploited. The report I received stated that
4535 data on the command line could provoke the exploit when Exim was running as
4536 exim, but did not say which command line option was involved. All I could
4537 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4538 running as the user.
4541 61. There was a buffer overflow vulnerability in the SPA authentication code
4542 (which came originally from the Samba project). I have added a test to the
4543 spa_base64_to_bits() function which I hope fixes it.
4546 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4547 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4549 63. The daemon start-up calls getloadavg() while still root for those OS that
4550 need the first call to be done as root, but it missed one case: when
4551 deliver_queue_load_max is set with deliver_drop_privilege. This is
4552 necessary for the benefit of the queue runner, because there is no re-exec
4553 when deliver_drop_privilege is set.
4555 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4556 This has been fixed.
4558 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4559 was in use, was not putting the data itself into the right store pool;
4560 consequently, it could be overwritten for a subsequent message in the same
4561 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4564 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4566 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4567 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4568 the function were treating the return as a boolean value, which happened to
4569 work because 0=false and not-0=true, but is not correct code.
4571 68. The host_aton() function was not handling scoped IPv6 addresses (those
4572 with, for example, "%eth0" on the end) correctly.
4574 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4575 negated items (that is, ~something) in unsigned ints. Some compilers
4576 apparently mutter when there is no cast.
4578 70. If an address verification called from an ACL failed, and did not produce a
4579 user-specific message (i.e. there was only a "system" message), nothing was
4580 put in $acl_verify_message. In this situation, it now puts the system
4583 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4584 session; change 4.31/43 added the unwanted input to the log line - except
4585 that it did not do this in the start of session case. It now does.
4587 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4588 This is wrong and can cause the other end to generate a synchronization
4589 error if it is another Exim or anything else that does the synchronization
4590 check. A QUIT command is no longer sent after a timeout.
4592 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4593 during host lookups.
4595 74. The defer_ok option of callout verification was not working if it was used
4596 when verifying addresses in header lines, that is, for this case:
4598 verify = header_sender/callout=defer_ok
4600 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4601 those file descriptors could be used for SMTP connections. If anything
4602 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4603 could be sent to the SMTP client, causing chaos. The daemon now opens
4604 stdin, stdout, and stderr to /dev/null when it puts itself into the
4607 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4608 log by default. The user can override this with suitable Perl magic.
4610 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4611 the log message when discard triggers, was not working for the DATA ACL or
4612 for the non-SMTP ACL.
4614 78. Error message wording change in sieve.c.
4616 79. If smtp_accept_max_per_host was set, the number of connections could be
4617 restricted to fewer than expected, because the daemon was trying to set up
4618 a new connection before checking whether the processes handling previous
4619 connections had finished. The check for completed processes is now done
4620 earlier. On busy systems, this bug wouldn't be noticed because something
4621 else would have woken the daemon, and it would have reaped the completed
4624 80. If a message was submitted locally by a user whose login name contained one
4625 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4626 It caused a spool format error. I have fixed the spool reading code. A
4627 related problem was that the "from" clause in the Received: line became
4628 illegal because of the space(s). It is now covered by ${quote_local_part.
4630 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4633 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4635 83. Patch from Sieve maintainer to fix unterminated string problem in
4636 "vacation" handling.
4638 84. Some minor changes to the Linux configuration files to help with other
4639 OS variants using glibc.
4641 85. One more patch for Sieve to update vacation handling to latest spec.
4644 ----------------------------------------------------
4645 See the note above about the 4.44 and 4.50 releases.
4646 ----------------------------------------------------
4652 1. Change 4.43/35 introduced a bug that caused file counts to be
4653 incorrectly computed when quota_filecount was set in an appendfile
4656 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4657 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4660 3. Give more explanation in the error message when the command for a transport
4661 filter fails to execute.
4663 4. There are several places where Exim runs a non-Exim command in a
4664 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4665 was being done only for the command run by the queryprogram router. It is
4666 now done for all such subprocesses. The other cases are: ${run, transport
4667 filters, and the commands run by the lmtp and pipe transports.
4669 5. Some older OS have a limit of 256 on the maximum number of file
4670 descriptors. Exim was using setrlimit() to set 1000 as a large value
4671 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4672 systems. I've change it so that if it can't get 1000, it tries for 256.
4674 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4675 was an oversight, and furthermore, ever since the addition of extra
4676 controls (e.g. 4.43/32), the checks on when to allow different forms of
4677 "control" were broken. There should now be diagnostics for all cases when a
4678 control that does not make sense is encountered.
4680 7. $recipients is now available in the predata ACL (oversight).
4682 8. Tidy the search cache before the fork to do a delivery from a message
4683 received from the command line. Otherwise the child will trigger a lookup
4684 failure and thereby defer the delivery if it tries to use (for example) a
4685 cached ldap connection that the parent has called unbind on.
4687 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4688 of $address_data from the recipient verification was clobbered by the
4689 sender verification.
4691 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4692 was its contents. (It was OK if the option was not defined at all.)
4694 11. A "Completed" log line is now written for messages that are removed from
4695 the spool by the -Mrm option.
4697 12. $host_address is now set to the target address during the checking of
4698 ignore_target_hosts.
4700 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4701 being passed; this would have caused $sender_host_name to have been used if
4702 matching the list had actually called for a host name (not very likely,
4703 since this list is usually IP addresses). A host name is now passed as
4706 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4707 code that specifies a non-restarting handler (typically sigaction() in
4708 modern systems) in an attempt to fix a rare and obscure crash bug.
4710 15. Narrowed the window for a race in the daemon that could cause it to ignore
4711 SIGCHLD signals. This is not a major problem, because they are used only to
4712 wake it up if nothing else does.
4714 16. A malformed maildirsize file could cause Exim to calculate negative values
4715 for the mailbox size or file count. Odd effects could occur as a result.
4716 The maildirsize information is now recalculated if the size or filecount
4719 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4720 support for a long time. Removed HAVE_SYS_VFS_H.
4722 18. Updated exipick to current release from John Jetmore.
4724 19. Allow an empty sender to be matched against a lookup in an address list.
4725 Previously the only cases considered were a regular expression, or an
4728 20. Exim went into a mad DNS lookup loop when doing a callout where the
4729 host was specified on the transport, if the DNS lookup yielded more than
4732 21. The RFC2047 encoding function was originally intended for short strings
4733 such as real names; it was not keeping to the 75-character limit for
4734 encoded words that the RFC imposes. It now respects the limit, and
4735 generates multiple encoded words if necessary. To be on the safe side, I
4736 have increased the buffer size for the ${rfc2047: expansion operator from
4739 22. Failure to deliver a bounce message always caused it to be frozen, even if
4740 there was an errors_to setting on the router. The errors_to setting is now
4743 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4744 canonical form (fully expanded) before being placed in
4745 $sender_host_address.
4747 24. Updated eximstats to version 1.33
4749 25. Include certificate and key file names in error message when GnuTLS fails
4750 to set them up, because the GnuTLS error message doesn't include the name
4751 of the failing file when there is a problem reading it.
4753 26. Expand error message when OpenSSL has problems setting up cert/key files.
4756 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4757 (this can affect the format of dates).
4759 28. exim_tidydb, when checking for the continued existence of a message for
4760 which it has found a message-specific retry record, was not finding
4761 messages that were in split spool directories. Consequently, it was
4762 deleting retry records that should have stayed in existence.
4764 29. eximstats updated to version 1.35
4765 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4766 1.35 - bugfix such that pie charts by volume are generated correctly
4768 30. The SPA authentication driver was not abandoning authentication and moving
4769 on to the next authenticator when an expansion was forced to fail,
4770 contradicting the general specification for all authenticators. Instead it
4771 was generating a temporary error. It now behaves as specified.
4773 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4774 (the order specifies the preference for clients). The order is now AES256,
4775 AES128, 3DES, ARCFOUR128.
4777 31. Small patch to Sieve code - explicitly set From: when generating an
4780 32. Exim crashed if a remote delivery caused a very long error message to be
4781 recorded - for instance if somebody sent an entire SpamAssassin report back
4782 as a large number of 550 error lines. This bug was coincidentally fixed by
4783 increasing the size of one of Exim's internal buffers (big_buffer) that
4784 happened as part of the Exiscan merge. However, to be on the safe side, I
4785 have made the code more robust (and fixed the comments that describe what
4788 33. Some experimental protocols are using DNS PTR records for new purposes. The
4789 keys for these records are domain names, not reversed IP addresses. The
4790 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4791 leaves it alone. Component reversal etc. now happens only for IP addresses.
4794 34. The host_aton() function is supposed to be passed a string that is known
4795 to be a valid IP address. However, in the case of IPv6 addresses, it was
4796 not checking this. This is a hostage to fortune. Exim now panics and dies
4797 if the condition is not met. A case was found where this could be provoked
4798 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4799 components; fortuitously, this particular loophole had already been fixed
4800 by change 4.50/55 or 4.44/33 above.
4802 If there are any other similar loopholes, the new check in host_aton()
4803 itself should stop them being exploited. The report I received stated that
4804 data on the command line could provoke the exploit when Exim was running as
4805 exim, but did not say which command line option was involved. All I could
4806 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4807 running as the user.
4810 35. There was a buffer overflow vulnerability in the SPA authentication code
4811 (which came originally from the Samba project). I have added a test to the
4812 spa_base64_to_bits() function which I hope fixes it.
4815 36. The daemon start-up calls getloadavg() while still root for those OS that
4816 need the first call to be done as root, but it missed one case: when
4817 deliver_queue_load_max is set with deliver_drop_privilege. This is
4818 necessary for the benefit of the queue runner, because there is no re-exec
4819 when deliver_drop_privilege is set.
4821 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4822 was in use, was not putting the data itself into the right store pool;
4823 consequently, it could be overwritten for a subsequent message in the same
4824 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4827 38. Sometimes the final signoff response after QUIT could fail to get
4828 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4829 < 0 before doing a fflush(). This bug looks as though it goes back to the
4830 introduction of TLS in release 3.20, but "sometimes" must have been rare
4831 because the tests only now provoked it.
4837 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4838 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4839 Now it returns an integer. A number of calls were still expecting a T/F
4840 return. Fortuitously, in all cases, the tests worked in OK situations,
4841 which is the norm. However, things would have gone wrong on any write
4842 failures on the smtp file descriptor. This function is used when sending
4843 messages over SMTP and also when doing verify callouts.
4845 2. When Exim is called to do synchronous delivery of a locally submitted
4846 message (the -odf or -odi options), it no longer closes stderr before doing
4849 3. Implemented the mua_wrapper option.
4851 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4853 5. Implemented the functions header_remove(), header_testname(),
4854 header_add_at_position(), and receive_remove_recipient(), and exported them
4857 6. If an ACL "warn" statement specified the addition of headers, Exim already
4858 inserted X-ACL-Warn: at the start if there was no header name. However, it
4859 was not making this test for the second and subsequent header lines if
4860 there were newlines in the string. This meant that an invalid header could
4861 be inserted if Exim was badly configured.
4863 7. Allow an ACL "warn" statement to add header lines at the start or after all
4864 the Received: headers, as well as at the end.
4866 8. Added the rcpt_4xx retry error code.
4868 9. Added postmaster_mailfrom=xxx to callout verification option.
4870 10. Added mailfrom=xxxx to the callout verification option, for verify=
4873 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4874 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4876 12. Inserted some casts to stop certain compilers warning when using pointer
4877 differences as field lengths or precisions in printf-type calls (mostly
4878 affecting debugging statements).
4880 13. Added optional readline() support for -be (dynamically loaded).
4882 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4883 same clock tick as a message's arrival, so that its received time was the
4884 same as the "first fail" time on the retry record, and that message
4885 remained on the queue past the ultimate address timeout, every queue runner
4886 would try a delivery (because it was past the ultimate address timeout) but
4887 after another failure, the ultimate address timeout, which should have then
4888 bounced the address, did not kick in. This was a "< instead of <=" error;
4889 in most cases the first failure would have been in the next clock tick
4890 after the received time, and all would be well.
4892 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4893 being recognized when the domain list was tested by the match_domain
4894 condition in an expansion string.
4896 16. Added the ${str2b64: operator.
4898 17. Exim was always calling setrlimit() to set a large limit for the number of
4899 processes, without checking whether the existing limit was already
4900 adequate. (It did check for the limit on file descriptors.) Furthermore,
4901 errors from getrlimit() and setrlimit() were being ignored. Now they are
4902 logged to the main and panic logs, but Exim does carry on, to try to do its
4903 job under whatever limits there are.
4905 18. Imported PCRE 5.0.
4907 19. Trivial typo in log message " temporarily refused connection" (the leading
4910 20. If the log selector return_path_on_delivery was set and an address was
4911 redirected to /dev/null, the delivery process crashed because it assumed
4912 that a return path would always be set for a "successful" delivery. In this
4913 case, the whole delivery is bypassed as an optimization, and therefore no
4916 21. Internal re-arrangement: the function for sending a challenge and reading
4917 a response while authentication was assuming a zero-terminated challenge
4918 string. It's now changed to take a pointer and a length, to allow for
4919 binary data in such strings.
4921 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4923 23. Exim was not respecting finduser_retries when seeking the login of the
4924 uid under which it was called; it was always trying 10 times. (The default
4925 setting of finduser_retries is zero.) Also, it was sleeping after the final
4926 failure, which is pointless.
4928 24. Implemented tls_on_connect_ports.
4930 25. Implemented acl_smtp_predata.
4932 26. If the domain in control=submission is set empty, Exim assumes that the
4933 authenticated id is a complete email address when it generates From: or
4934 Sender: header lines.
4936 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4937 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4938 chown and chgrp in /bin and hostname in /usr/bin.
4940 28. Exim was keeping the "process log" file open after each use, just as it
4941 does for the main log. This opens the possibility of it remaining open for
4942 long periods when the USR1 signal hits a daemon. Occasional processlog
4943 errors were reported, that could have been caused by this. Anyway, it seems
4944 much more sensible not to leave this file open at all, so that is what now
4947 29. The long-running daemon process does not normally write to the log once it
4948 has entered its main loop, and it closes the log before doing so. This is
4949 so that log files can straightforwardly be renamed and moved. However,
4950 there are a couple of unusual error situations where the daemon does write
4951 log entries, and I had neglected to close the log afterwards.
4953 30. The text of an SMTP error response that was received during a remote
4954 delivery was being truncated at 512 bytes. This is too short for some of
4955 the long messages that one sometimes sees. I've increased the limit to
4958 31. It is now possible to make retry rules that apply only when a message has a
4959 specific sender, in particular, an empty sender.
4961 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4962 it possible to be selective about when SMTP synchronization is enforced.
4964 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4966 32. Implemented hosts_connection_nolog.
4968 33. Added an ACL for QUIT.
4970 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4973 35. Added mailbox_size and mailbox_filecount to appendfile.
4975 36. Added control = no_multiline_responses to ACLs.
4977 37. There was a bug in the logic of the code that waits for the clock to tick
4978 in the case where the clock went backwards by a substantial amount such
4979 that the microsecond fraction of "now" was more than the microsecond
4980 fraction of "then" (but the whole seconds number was less).
4982 38. Added support for the libradius Radius client library this is found on
4983 FreeBSD (previously only the radiusclient library was supported).
4989 1. When certain lookups returned multiple values in the form name=value, the
4990 quoting of the values was not always being done properly. Specifically:
4991 (a) If the value started with a double quote, but contained no whitespace,
4993 (b) If the value contained whitespace other than a space character (i.e.
4994 tabs or newlines or carriage returns) it was not quoted.
4995 This fix has been applied to the mysql and pgsql lookups by writing a
4996 separate quoting function and calling it from the lookup code. The fix
4997 should probably also be applied to nisplus, ibase and oracle lookups, but
4998 since I cannot test any of those, I have not disturbed their existing code.
5000 2. A hit in the callout cache for a specific address caused a log line with no
5001 reason for rejecting RCPT. Now it says "Previous (cached) callout
5002 verification failure".
5004 3. There was an off-by-one bug in the queryprogram router. An over-long
5005 return line was truncated at 256 instead of 255 characters, thereby
5006 overflowing its buffer with the terminating zero. As well as fixing this, I
5007 have increased the buffer size to 1024 (and made a note to document this).
5009 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5010 when Exim is waiting for an SMTP response from a remote server, Exim
5011 restarts its select() call on the socket, thereby resetting its timeout.
5012 This is not a problem when such interrupts are rare. Somebody set up a cron
5013 job to run exiwhat every 2 minutes, which is less than the normal select()
5014 timeout (5 or 10 minutes). This meant that the select() timeout never
5015 kicked in because it was always reset. I have fixed this by comparing the
5016 time when an interrupt arrives with the time at the start of the first call
5017 to select(). If more time than the timeout has elapsed, the interrupt is
5018 treated as a timeout.
5020 5. Some internal re-factoring in preparation for the addition of Sieve
5021 extensions (by MH). In particular, the "personal" test is moved to a
5022 separate function, and given an option for scanning Cc: and Bcc: (which is
5023 not set for Exim filters).
5025 6. When Exim created an email address using the login of the caller as the
5026 local part (e.g. when creating a From: or Sender: header line), it was not
5027 quoting the local part when it contained special characters such as @.
5029 7. Installed new OpenBSD configuration files.
5031 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5032 try to make them clearer.
5034 9. Callout options, other than the timeout value, were being ignored when
5035 verifying sender addresses in header lines. For example, when using
5037 verify = header_sender/callout=no_cache
5039 the cache was (incorrectly) being used.
5041 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5042 only the Cygwin environment.
5044 11. When return_path_on_delivery was set as a log selector, if different remote
5045 addresses in the same message used different return paths and parallel
5046 remote delivery occurred, the wrong values would sometimes be logged.
5047 (Whenever a remote delivery process finished, the return path value from
5048 the most recently started remote delivery process was logged.)
5050 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5051 lines when AUTH and/or TLS are in use. This is the "received protocol"
5052 field. Exim used to use "asmtp" for authenticated SMTP, without any
5053 indication (in the protocol name) for TLS use. Now it follows the RFC and
5054 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5055 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5056 names appear in log lines as well as in Received: header lines.
5058 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5059 extensions, and comparison tests, and to fix some bugs.
5061 14. Changes to the "personal" filter test:
5063 (1) The test was buggy in that it was just doing the equivalent of
5064 "contains" tests on header lines. For example, if a user's address was
5065 anne@some.where, the "personal" test would incorrectly be true for
5067 To: susanne@some.where
5069 This test is now done by extracting each address from the header in turn,
5070 and checking the entire address. Other tests that are part of "personal"
5071 are now done using regular expressions (for example, to check local parts
5072 of addresses in From: header lines).
5074 (2) The list of non-personal local parts in From: addresses has been
5075 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5076 taken from the Sieve specification recommendations.
5078 (3) If the message contains any header line starting with "List-" it is
5079 treated as non-personal.
5081 (4) The test for "circular" in the Subject: header line has been removed
5082 because it now seems ill-conceived.
5084 15. Minor typos in src/EDITME comments corrected.
5086 16. Installed latest exipick from John Jetmore.
5088 17. If headers_add on a router specified a text string that was too long for
5089 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5090 of string_sprintf() is now avoided.
5092 18. $message_body_size was not set (it was always zero) when running the DATA
5093 ACL and the local_scan() function.
5095 19. For the "mail" command in an Exim filter, no default was being set for
5096 the once_repeat time, causing a random time value to be used if "once" was
5097 specified. (If the value happened to be <= 0, no repeat happened.) The
5098 default is now 0s, meaning "never repeat". The "vacation" command was OK
5099 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5100 (I found it when inspecting the code).
5102 20. There is now an overall timeout for performing a callout verification. It
5103 defaults to 4 times the callout timeout, which applies to individual SMTP
5104 commands during the callout. The overall timeout applies when there is more
5105 than one host that can be tried. The timeout is checked before trying the
5106 next host. This prevents very long delays if there are a large number of
5107 hosts and all are timing out (e.g. when the network connections are timing
5108 out). The value of the overall timeout can be changed by specifying an
5109 additional sub-option for "callout", called "maxwait". For example:
5111 verify = sender/callout=5s,maxwait=20s
5113 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5114 to the end before writing, but this should make it even safer).
5116 22. Exim was forgetting that it had advertised PIPELINING for the second and
5117 subsequent messages on an SMTP connection. It was also not resetting its
5118 memory on STARTTLS and an internal HELO.
5120 23. When Exim logs an SMTP synchronization error within a session, it now
5121 records whether PIPELINING has been advertised or not.
5123 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5124 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5125 rather than long int.
5127 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5129 26. Added the never_mail option to autoreply.
5135 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5136 crash if the getsockname() call failed; this can happen if a connection is
5137 closed very soon after it is established. The problem was simply in the
5138 order in which certain operations were done, causing Exim to try to write
5139 to the SMTP stream before it had set up the file descriptor. The bug has
5140 been fixed by making things happen in the correct order.
5146 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5147 before the connection was closed, thus losing the rejection response.
5149 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5150 some early Solaris releases, but causes trouble in current releases where
5151 socklen_t is defined.
5153 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5156 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5159 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5160 that releasing the top of it at the end releases what was used for sub-
5161 expansions (unless the block got too big). However, discard this block if
5162 the first thing is a variable or header, so that we can use its block when
5163 it is dynamic (useful for very large $message_headers, for example).
5165 6. Lookups now cache *every* query, not just the most recent. A new, separate
5166 store pool is used for this. It can be recovered when all lookup caches are
5167 flushed. Lookups now release memory at the end of their result strings.
5168 This has involved some general refactoring of the lookup sources.
5170 7. Some code has been added to the store_xxx() functions to reduce the amount
5171 of flapping under certain conditions.
5173 8. log_incoming_interface used to affect only the <= reception log lines. Now
5174 it causes the local interface and port to be added to several more SMTP log
5175 lines, for example "SMTP connection from", and rejection lines.
5177 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5179 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5181 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5182 could be overwritten at the end of the current message (or the start of a
5183 new message if it was set in a HELO ACL). The value is now preserved for
5184 the duration of the SMTP connection.
5186 12. If a transport had a headers_rewrite setting, and a matching header line
5187 contained an unqualified address, that address was qualified, even if it
5188 did not match any rewriting rules. The underlying bug was that the values
5189 of the flags that permit the existence of unqualified sender and recipient
5190 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5191 non-local messages, and by -bnq for local messages) were not being
5192 preserved with the message after it was received.
5194 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5195 "next input=" as part of the text comprising the host identity instead of
5196 the correct text. The code was using the same buffer for two different
5197 strings. However, depending on which order the printing function evaluated
5198 its arguments, the bug did not always show up. Under Linux, for example, my
5199 test suite worked just fine.
5201 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5202 doesn't work with some older versions of Perl. It has been changed to "my",
5203 which in any case is probably the better facility to use.
5205 15. A really picky compiler found some instances of statements for creating
5206 error messages that either had too many or two few arguments for the format
5209 16. The size of the buffer for calls to the DNS resolver has been increased
5210 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5211 for addresses that have a lot of PTR records. This alleviates a problem; it
5212 does not fully solve it.
5214 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5215 buffer now truncates the list and logs the incident, which is the same
5216 action as happens when Exim is looking up a host name and its aliases.
5217 Previously in this situation something unpredictable would happen;
5218 sometimes it was "internal error: store_reset failed".
5220 18. If a server dropped the connection unexpectedly when an Exim client was
5221 using GnuTLS and trying to read a response, the client delivery process
5222 crashed while trying to generate an error log message.
5224 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5225 string, for example:
5227 warn message = H1: something\nH2: something
5229 the text was added as a single header line from Exim's point of view
5230 though it ended up OK in the delivered message. However, searching for the
5231 second and subsequent header lines using $h_h2: did not work. This has been
5232 fixed. Similarly, if a system filter added multiple headers in this way,
5233 the routers could not see them.
5235 20. Expanded the error message when iplsearch is called with an invalid key to
5236 suggest using net-iplsearch in a host list.
5238 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5239 ACLs are no longer actually imposed (and a message to that effect is
5242 22. If a "gecos" field in a passwd entry contained escaped characters, in
5243 particular, if it contained a \" sequence, Exim got it wrong when building
5244 a From: or a Sender: header from that name. A second bug also caused
5245 incorrect handling when an unquoted " was present following a character
5246 that needed quoting.
5248 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5249 was not being matched caselessly.
5251 24. Arranged for all hyphens in the exim.8 source to be escaped with
5254 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5255 properly. Recipient callout cache records were still being keyed to include
5256 the sender, even when use_sender was set false. This led to far more
5257 callouts that were necessary. The sender is no longer included in the key
5258 when use_sender is false.
5260 26. Added "control = submission" modifier to ACLs.
5262 27. Added the ${base62d: operator to decode base 62 numbers.
5264 28. dnsdb lookups can now access SRV records.
5266 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5267 the configuration file.
5269 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5270 (-v) mode. This makes the output for a verbose queue run more intelligible.
5272 31. Added a use_postmaster feature to recipient callouts.
5274 32. Added the $body_zerocount variable, containing the number of binary zero
5275 bytes in the message body.
5277 33. The time of last modification of the "new" subdirectory is now used as the
5278 "mailbox time last read" when there is a quota error for a maildir
5281 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5283 35. Added +ignore_unknown as a special item in host lists.
5285 36. Code for decoding IPv6 addresses in host lists is now included, even if
5286 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5287 address was recognized as an IP address, but was then not correctly decoded
5288 into binary, causing unexpected and incorrect effects when compared with
5295 1. Very minor rewording of debugging text in manualroute to say "list of
5296 hosts" instead of "hostlist".
5298 2. If verify=header_syntax was set, and a header line with an unqualified
5299 address (no domain) and a large number of spaces between the end of the
5300 name and the colon was received, the reception process suffered a buffer
5301 overflow, and (when I tested it) crashed. This was caused by some obsolete
5302 code that should have been removed. The fix is to remove it!
5304 3. When running in the test harness, delay a bit after writing a bounce
5305 message to get a bit more predictability in the log output.
5307 4. Added a call to search_tidyup() just before forking a reception process. In
5308 theory, someone could use a lookup in the expansion of smtp_accept_max_
5309 per_host which, without the tidyup, could leave open a database connection.
5311 5. Added the variables $recipient_data and $sender_data which get set from a
5312 lookup success in an ACL "recipients" or "senders" condition, or a router
5313 "senders" option, similar to $domain_data and $local_part_data.
5315 6. Moved the writing of debug_print from before to after the "senders" test
5318 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5319 problems for message scanning, either using a data ACL, or using
5320 local_scan() because the Received: header was not generated till after they
5321 were called (in order to set the time as the time of reception completion).
5322 I have revised the way this works. The header is now generated after the
5323 body is received, but before the ACL or local_scan() are called. After they
5324 are run, the timestamp in the header is updated.
5330 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5331 before starting a queue runner without re-exec. This happened only when
5332 deliver_drop_privilege was set or when the Exim user was set to root. The
5333 effect of the bug was that timeouts during subsequent deliveries caused
5334 crashes instead of being properly handled. The handler is now left at its
5335 default (and expected) setting.
5337 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5338 message, again when deliver_drop_privilege is set or Exim is run as root.
5339 The bug described in (1) was not present in this case, but the tidying up
5340 of the other signals was missing. I have made the two cases consistent.
5342 3. The ignore_target_hosts setting on a manualroute router was being ignored
5343 for hosts that were looked up using the /MX notation.
5345 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5348 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5349 operated on the sender address. After changing the $sender_address to <>
5350 for the sender address verify, Exim was re-instated it as the original
5351 (before rewriting) address, but remembering that it had rewritten it, so it
5352 wasn't rewriting it again. This bug also had the effect of breaking the
5353 sender address verification caching when the sender address was rewritten.
5355 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5356 This has been changed so that if the ip literal address matches
5357 ignore_target_hosts, the router declines.
5359 7. Added expansion conditions match_domain, match_address, and match_local_
5360 part (NOT match_host).
5362 8. The placeholder for the Received: header didn't have a length field set.
5364 9. Added code to Exim itself and to exim_lock to test for a specific race
5365 condition that could lead to file corruption when using MBX delivery. The
5366 issue is with the lockfile that is created in /tmp. If this file is removed
5367 after a process has opened it but before that process has acquired a lock,
5368 there is the potential for a second process to recreate the file and also
5369 acquire a lock. This could lead to two Exim processes writing to the file
5370 at the same time. The added code performs the same test as UW imapd; it
5371 checks after acquiring the lock that its file descriptor still refers to
5372 the same named file.
5374 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5375 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5378 11. Added the smtp_active_hostname option. If used, this will typically be made
5379 to depend on the incoming interface address. Because $interface_address is
5380 not set up until the daemon has forked a reception process, error responses
5381 that can happen earlier (such as "too many connections") no longer contain
5384 12. If an expansion in a condition on a "warn" statement fails because a lookup
5385 defers, the "warn" statement is abandoned, and the next ACL statement is
5386 processed. Previously this caused the whole ACL to be aborted.
5388 13. Added the iplsearch lookup type.
5390 14. Added ident_timeout as a log selector.
5392 15. Added tls_certificate_verified as a log selector.
5394 16. Added a global option tls_require_ciphers (compare the smtp transport
5395 option of the same name). This controls incoming TLS connections.
5397 17. I finally figured out how to make tls_require_ciphers do a similar thing
5398 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5399 before starting the TLS session.
5401 18. Tabs are now shown as \t in -bP output.
5403 19. If the log selector return_path_on_delivery was set, Exim crashed when
5404 bouncing a message because it had too many Received: header lines.
5406 20. If two routers both had headers_remove settings, and the first one included
5407 a superfluous trailing colon, the final name in the first list and the
5408 first name in the second list were incorrectly joined into one item (with a
5409 colon in the middle).
5415 1. Added -C and -D options to the exinext utility, mainly to make it easier
5416 to include in the automated testing, but these could be helpful when
5417 multiple configurations are in use.
5419 2. The exinext utility was not formatting the output nicely when there was
5420 an alternate port involved in the retry record key, nor when there was a
5421 message id as well (for retries that were specific to a specific message
5422 and a specific host). It was also confused by IPv6 addresses, because of
5423 the additional colons they contain. I have fixed the IPv4 problem, and
5424 patched it up to do a reasonable job for IPv6.
5426 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5427 delivery, the log line now contains "pipelined" if PIPELINING was used.
5429 4. An SMTP transport process used to panic and die if the bind() call to set
5430 an explicit outgoing interface failed. This has been changed; it is now
5431 treated in the same way as a connect() failure.
5433 5. A reference to $sender_host_name in the part of a conditional expansion
5434 that was being skipped was still causing a DNS lookup. This no longer
5437 6. The def: expansion condition was not recognizing references to header lines
5438 that used bh_ and bheader_.
5440 7. Added the _cache feature to named lists.
5442 8. The code for checking quota_filecount in the appendfile transport was
5443 allowing one more file than it should have been.
5445 9. For compatibility with Sendmail, the command line option
5453 and sets the incoming protocol and host name (for trusted callers). The
5454 host name and its colon can be omitted when only the protocol is to be set.
5455 Note the Exim already has two private options, -pd and -ps, that refer to
5456 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5457 "s", but I don't think that's a major issue.
5459 10. A number of refactoring changes to the code, none of which should affect
5462 (a) The number of logging options was getting close to filling up the
5463 32-bit word that was used as a bit map. I have split them into two classes:
5464 those that are passed in the argument to log_write(), and those that are
5465 only ever tested independently outside of that function. These are now in
5466 separate 32-bit words, so there is plenty of room for expansion again.
5467 There is no change in the user interface or the logging behaviour.
5469 (b) When building, for example, log lines, the code previously used a
5470 macro that called string_cat() twice, in order to add two strings. This is
5471 not really sufficiently general. Furthermore, there was one instance where
5472 it was actually wrong because one of the argument was used twice, and in
5473 one call a function was used. (As it happened, calling the function twice
5474 did not affect the overall behaviour.) The macro has been replaced by a
5475 function that can join an arbitrary number of extra strings onto a growing
5478 (c) The code for expansion conditions now uses a table and a binary chop
5479 instead of a serial search (which was left over from when there were very
5480 few conditions). Also, it now recognizes conditions like "pam" even when
5481 the relevant support is not compiled in: a suitably worded error message is
5482 given if an attempt is made to use such a condition.
5484 11. Added ${time_interval:xxxxx}.
5486 12. A bug was causing one of the ddress fields not to be passed back correctly
5487 from remote delivery subprocesses. The field in question was not being
5488 subsequently used, so this caused to problems in practice.
5490 13. Added new log selectors queue_time and deliver_time.
5492 14. Might have fixed a bug in maildirsizefile handling that threw up
5493 "unexpected character" debug warnings, and recalculated the data
5494 unnecessarily. In any case, I expanded the warning message to give more
5497 15. Added the message "Restricted characters in address" to the statements in
5498 the default ACL that block characters like @ and % in local parts.
5500 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5501 Three changes have been made:
5503 (a) There was a serious bug; a negative response to MAIL caused the whole
5504 recipient domain to be cached as invalid, thereby blocking all messages
5505 to all local parts at the same domain, from all senders. This bug has
5506 been fixed. The domain is no longer cached after a negative response to
5507 MAIL if the sender used is not empty.
5509 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5512 (c) A new callout option, "use_sender" has been added for people who want
5513 the modified behaviour.
5519 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5522 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5523 indeed breaks things for older releases.
5525 3. Added additional logging to the case where there is a problem reading data
5526 from a filter that is running in a subprocess using a pipe, in order to
5527 try to track down a specific problem.
5529 4. Testing facility fudge: when running in the test harness and attempting
5530 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5531 getting "No route to host". Convert this to a timeout.
5533 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5536 6. Some OS don't have socklen_t but use size_t instead. This affects the
5537 fifth argument of getsockopt() amongst other things. This is now
5538 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5539 can be set for individual OS. I have set it for SunOS5, OSF1, and
5540 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5541 some earlier ones do not.
5543 7. Change 4.30/15 was not doing the test caselessly.
5545 8. The standard form for an IPv6 address literal was being rejected by address
5546 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5547 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5548 this, as well as the form without the "IPv6" on the front (but only when
5549 address literals are enabled, of course).
5551 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5553 10. Exim crashed if a message with an empty sender address specified by -f
5554 encountered a router with an errors_to setting. This could be provoked only
5555 by a command such as
5559 where an empty string was supplied; "<>" did not hit this bug.
5561 11. Installed PCRE release 4.5.
5563 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5564 remained set. It is now erased.
5566 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5567 times from message ids (which are base 36 rather than the normal 62).
5569 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5570 were being counted as actual protocol errors, and logged if the log
5571 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5572 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5573 and DATA following a set of rejected RCPTs do not count as protocol errors.
5574 In other words, Exim assumes they were pipelined, though this may not
5575 actually be the case. Of course, in all cases the client gets an
5576 appropriate error code.
5578 15. If a lookup fails in an ACL condition, a message about the failure may
5579 be available; it is used if testing the ACL cannot continue, because most
5580 such messages specify what the cause of the deferral is. However, some
5581 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5582 that caused an old message to be retained and used if a later statement
5583 caused a defer, replacing the real cause of the deferral.
5585 16. If an IP address had so many PTR records that the DNS lookup buffer
5586 was not large enough to hold them, Exim could crash while trying to process
5587 the truncated data. It now detects and logs this case.
5589 17. Further to 4.21/58, another change has been made: if (and only if) the
5590 first line of a message (the first header line) ends with CRLF, a bare LF
5591 in a subsequent header line has a space inserted after it, so as not to
5592 terminate the header.
5594 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5595 unnecessarily, used atoi() instead of strtol(), and didn't check the
5596 termination when getting file sizes from file names by regex.
5598 19. Completely re-implemented the support for maildirsize files, in the light
5599 of a number of problems with the previous contributed implementation
5600 (4.30/29). In particular:
5602 . If the quota is zero, the maildirsize file is maintained, but no quota is
5605 . If the maildir directory does not exist, it is created before any attempt
5606 to write a maildirsize file.
5608 . The quota value in the file is just a cache; if the quota is changed in
5609 the transport, the new value overrides.
5611 . A regular expression is available for excluding directories from the
5614 20. The autoreply transport checks the characters in options that define the
5615 message's headers; it allows continued headers, but it was checking with
5616 isspace() after an embedded newline instead of explicitly looking for a
5619 21. If all the "regular" hosts to which an address was routed had passed their
5620 expiry times, and had not reached their retry times, the address was
5621 bounced, even if fallback hosts were defined. Now Exim should go on to try
5624 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5625 equivalent code in the SMTP transport. Some hosts send humungous responses
5626 to HELO/EHLO, more than 1024 it seems.
5628 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5629 gives compiler warnings in some environments. I've now done it "properly",
5632 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5633 (because of problems with the built-in one) was declared to return uschar *
5634 instead of char *, causing compiler failure.
5636 25. Fixed a file descriptor leak when processing alias/forward files.
5638 26. Fixed a minor format string issue in dbfn.c.
5640 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5642 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5643 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5646 29. When a sender address is verified, it is cached, to save repeating the test
5647 when there is more than one recipient in a message. However, when the
5648 verification involves a callout, it is possible for different callout
5649 options to be set for different recipients. It is too complicated to keep
5650 track of this in the cache, so now Exim always runs a verification when a
5651 callout is required, relying on the callout cache for the optimization.
5652 The overhead is duplication of the address routing, but this should not be
5655 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5656 to be verified with callout=postmaster, and the main callout worked but the
5657 postmaster check failed, the verification correctly failed. However, if a
5658 subsequent RCPT command asked for sender verification *without* the
5659 postmaster check, incorrect caching caused this verification also to fail,
5662 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5663 it was not caching the DNS options (qualify_single, search_parents) that
5664 were used when the lookup failed. A subsequent lookup with different
5665 options therefore always gave the same answer, though there were cases
5666 where it should not have. (Example: a "domains = !$mx_any" option on a
5667 dnslookup router: the "domains" option is always processed without any
5668 widening, but the router might have qualify_single set.) Now Exim uses the
5669 cached value only when the same options are set.
5671 32. Added John Jetmore's "exipick" utility to the distribution.
5673 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5674 than a timeout (e.g. a certificate is required, and is not provided), an
5675 Exim server now closes the connection immediately. Previously it waited for
5676 the client to close - but if the client is SSL, it seems that they each
5677 wait for each other, leading to a delay before one of them times out.
5679 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5680 maintained 0.8.x compatibility because I don't think many are using it, and
5681 it is clearly obsolete.
5683 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5686 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5687 client certificate was expired. A simple patch fixes this, though I don't
5688 understand the full logic of why the verify callback is called multiple
5691 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5692 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5693 which causes problems with some clients (such as the Certicom SSL Plus
5694 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5695 disables the coutermeasure allowing Eudora to connect."
5697 38. Exim was not checking that a write() to a log file succeeded. This could
5698 lead to Bad Things if a log got too big, in particular if it hit a file
5699 size limit. Exim now panics and dies if it cannot write to a log file, just
5700 as it does if it cannot open a log file.
5702 39. Modified OS/Makefile-Linux so that it now contains
5704 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5706 The two -D definitions ensure that Exim is compiled with large file
5707 support, which makes it possible to handle log files that are bigger than
5710 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5711 instance) a domain was checked against a named list that involved a lookup,
5712 causing $domain_data to be set, then another domain was checked against the
5713 same list, then the first domain was re-checked, the value of $domain_data
5714 after the final check could be wrong. In particular, if the second check
5715 failed, it could be set empty. This bug probably also applied to
5718 41. The strip_trailing_dot option was not being applied to the address given
5719 with the -f command-line option.
5721 42. The code for reading a message's header from the spool was incrementing
5722 $received_count, but never initializing it. This meant that the value was
5723 incorrect (doubled) while delivering a message in the same process in which
5724 it was received. In the most common configuration of Exim, this never
5725 happens - a fresh exec is done - but it can happen when
5726 deliver_drop_privilege is set.
5728 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5729 it now includes up to 150 characters of the unexpected data in the log
5732 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5733 and building data strings. The size of both of these buffers was 10 000
5734 bytes - far larger than anybody would *ever* want, thought I. Needless to
5735 say, somebody hit the limit. I have increased the maximum line length to
5736 20 000 and the maximum data length of concatenated lines to 100 000. I have
5737 also fixed two bugs, because there was no checking on these buffers. Tsk,
5738 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5739 buffer is too small.
5741 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5742 lsearch lookups. Now it does.
5744 46. When parsing a route_list item in a manualroute router, a fixed-length
5745 buffer was used for the list of hosts. I made this 1024 bytes long,
5746 thinking that nobody would ever have a list of hosts that long. Wrong.
5747 Somebody had a whole pile of complicated expansion conditions, and the
5748 string was silently truncated, leading to an expansion error. It turns out
5749 that it is easier to change to an unlimited length (owing to other changes
5750 that have happened since this code was originally written) than to build
5751 structure for giving a limitation error. The length of the item that
5752 expands into the list of hosts is now unlimited.
5754 47. The lsearch lookup could not handle data where the length of text line was
5755 more than 4095 characters. Such lines were truncated, leading to shortened
5756 data being returned. It should now handle lines of any length.
5758 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5759 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5762 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5763 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5764 no longer happens in comment lines. A list of replacements is now placed
5765 at the head of all of the source files, except those whose only change is
5766 to replace PERL_COMMAND in the very first #! line.
5768 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5769 messages on the queue, with a bottom-up merge sort, using code contributed
5770 by Michael Haardt. This should make operations like -bp somewhat faster on
5771 large queues. It won't affect queue runners, except when queue_run_in_order
5774 51. Installed eximstats 1.31 in the distribution.
5776 52. Added support for SRV lookups to the dnslookup router.
5778 53. If an ACL referred to $message_body or $message_body_end, the value was not
5779 reset for any messages that followed in the same SMTP session.
5781 54. The store-handling optimization for building very long strings was not
5782 differentiating between the different store pools. I don't think this
5783 actually made any difference in practice, but I've tidied it.
5785 55. While running the routers to verify a sender address, $sender_address
5786 was still set to the sender address. This is wrong, because when routing to
5787 send a bounce to the sender, it would be empty. Therefore, I have changed
5788 it so that, while verifying a sender address, $sender_address is set to <>.
5789 (There is no change to what happens when verifying a recipient address.)
5791 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5792 target A or AAAA records (if not already returned) without resetting the
5793 qualify_single or search_parents options of the DNS resolver. These are
5794 inappropriate in this case because the targets of MX and SRV records must
5795 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5796 target that, when qualified, matched something in the local domain. These
5797 two options are now turned off when doing these lookups.
5799 57. It seems that at least some releases of Reiserfs (which does not have the
5800 concept of a fixed number of inodes) returns zero and not -1 for the
5801 number of available inodes. This interacted badly with check_spool_inodes,
5802 which assumed that -1 was the "no such thing" setting. What I have done is
5803 to check that the total number of inodes is greater than zero before doing
5804 the test of how many are available.
5806 58. When a "warn" ACL statement has a log_message modifier, the message is
5807 remembered, and not repeated. This is to avoid a lot of repetition when a
5808 message has many recipients that cause the same warning to be written.
5809 However, Exim was preserving the list of already written lines for an
5810 entire SMTP session, which doesn't seem right. The memory is now reset if a
5811 new message is started.
5813 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5814 debugging output unless log_rewrite was also set.
5816 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5817 of (int)(handle) when we know that handle contains (void *)(-1).
5819 61. The Exim daemon panic-logs an error return when it closes the incoming
5820 connection. However "connection reset by peer" seems to be common, and
5821 isn't really an error worthy of noting specially, so that particular error
5824 62. When Exim is trying to find all the local interfaces, it used to panic and
5825 die if the ioctl to get the interface flags failed. However, it seems that
5826 on at least one OS (Solaris 9) it is possible to have an interface that is
5827 included in the list of interfaces, but for which you get a failure error
5828 for this call. This happens when the interface is not "plumbed" into a
5829 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5830 failure of the "get flags" call assumes that the interface is down.
5832 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5833 makes life easier for people who are doing arithmetic on fields extracted
5834 from dates, where you often get leading zeros that should not be
5835 interpreted as octal.
5837 64. Added qualify_domain to the redirect router, to override the global
5840 65. If a pathologically long header line contained very many addresses (the
5841 report of this problem mentioned 10 000) and each of them was rewritten,
5842 Exim could use up a very large amount of memory. (It kept on making new
5843 copies of the header line as it rewrote, and never released the old ones.)
5844 At the expense of a bit more processing, the header rewriting function has
5845 been changed so that it no longer eats memory in this way.
5847 66. The generation of the Received: header has been moved from the time that a
5848 message starts to be received, to the time that it finishes. The timestamp
5849 in the Received: header should now be very close to that of the <= log
5850 line. There are two side-effects of this change:
5852 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5853 logged header lines no longer include the local Received: line, because
5854 it has not yet been created. The same applies to a copy of the message
5855 that is returned to a non-SMTP sender when a message is rejected.
5857 (b) When a filter file is tested using -bf, no additional Received: header
5858 is added to the test message. After some thought, I decided that this
5861 This change does not affect the value of $received_for. It is still set
5862 after address rewriting, but before local_scan() is called.
5864 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5866 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5867 gave an unhelpful panic error message, and a defer error. I have managed to
5868 change this behaviour so that it now rejects any supplied certificate,
5869 which seems right, as the list of acceptable certificates is empty.
5871 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5872 gave an unhelpful defer error. I have not managed to make this reject any
5873 supplied certificates, but the error message it gives is "no certificate
5874 supplied", which is not helpful.
5876 70. exigrep's output now also includes lines that are not associated with any
5877 message, but which match the given pattern. Implemented by a patch from
5878 Martin Sluka, which also tidied up the Perl a bit.
5880 71. Recipient callout verification, like sender verification, was using <> in
5881 the MAIL FROM command. This isn't really the right thing, since the actual
5882 sender may affect whether the remote host accepts the recipient or not. I
5883 have changed it to use the actual sender in the callout; this means that
5884 the cache record is now keyed on a recipient/sender pair, not just the
5885 recipient address. There doesn't seem to be a real danger of callout loops,
5886 since a callout by the remote host to check the sender would use <>.
5887 [SEE ABOVE: changed after hitting problems.]
5889 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5890 temporary errors. However, in the case of such a code being given after
5891 the end of a data transmission (i.e. after ".") Exim was failing to write
5892 a retry record for the message. (Yes, there was some broken host that was
5893 actually sending 8xx at this point.)
5895 73. An unknown lookup type in a host list could cause Exim to panic-die when
5896 the list was checked. (An example that provoked this was putting <; in the
5897 middle of a list instead of at the start.) If this happened during a DATA
5898 ACL check, a -D file could be left lying around. This kind of configuration
5899 error no longer causes Exim to die; instead it causes a defer error. The
5900 incident is still logged to the main and panic logs.
5902 74. Buglet left over from Exim 3 conversion. The message "too many messages
5903 in one connection" was written to the rejectlog but not the mainlog, except
5904 when address rewriting (yes!) was being logged.
5906 75. Added write_rejectlog option.
5908 76. When a system filter was run not as root (that is, when system_filter_user
5909 was set), the values of the $n variables were not being returned to the
5910 main process; thus, they were not subsequently available in the $sn
5913 77. Added +return_path_on_delivery log selector.
5915 78. A connection timeout was being treated differently from recipients deferred
5916 when testing hosts_max_try with a message that was older than the host's
5917 retry timeout. (The host should not be counted, thus allowing all hosts to
5918 be tried at least once before bouncing.) This may have been the cause of an
5919 occasionally reported bug whereby a message would remain on the queue
5920 longer than the retry timeout, but would be bounced if a delivery was
5921 forced. I say "may" because I never totally pinned down the problem;
5922 setting up timeout/retry tests is difficult. See also the next item.
5924 79. The ultimate address timeout was not being applied to errors that involved
5925 a combination of host plus message (for example, a timeout on a MAIL
5926 command). When an address resolved to a number of possible hosts, and they
5927 were not all tried for each delivery (e.g. because of hosts_max_try), a
5928 message could remain on the queue longer than the retry timeout.
5930 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5933 81. Fixed an obscure SMTP outgoing bug which required at least the following
5934 conditions: (a) there was another message waiting for the same server;
5935 (b) the server returned 5xx to all RCPT commands in the first message so
5936 that the message was not completed; (c) the server dropped the connection
5937 or gave a negative response to the RSET that Exim sends to abort the
5938 transaction. The observed case was a dropped connection after DATA that had
5939 been sent in pipelining mode. That is, the server had advertised PIPELINING
5940 but was not implementing it correctly. The effect of the bug was incorrect
5941 behaviour, such as trying another host, and this could lead to a crash.
5947 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5948 and daemon.c were passed as pointers to ints; they should have been
5949 pointers to socklen_t variables (which are typically unsigned ints).
5951 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5954 3. Fixed a really odd bug that affected only the testing scheme; patching a
5955 certain fixed string in the binary changed the value of another string that
5956 happened to be identical to the end of the original first string.
5958 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5959 name", it returns that address as the IP address. On some operating
5960 systems (e.g. Solaris), it also passes back the IP address string as the
5961 "host name". However, on others (e.g. Linux), it passes back an empty
5962 string. Exim wasn't checking for this, and was changing the host name to an
5963 empty string, assuming it had been canonicalized.
5965 5. Although rare, it is permitted to have more than one PTR record for a given
5966 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5967 all the names associated with an address, because they do in Solaris.
5968 However, it seems that they do not in Linux for data that comes from the
5969 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5970 I found this out when I moved to a new Linux workstation and tried to run
5971 the Exim test suite.
5973 To get round this problem I have changed the code so that it now does its
5974 own call to the DNS to look up PTR records when searching for a host name.
5975 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5976 addresses that are only in /etc/hosts are still found.
5978 This behaviour is, however, controlled by an option called host_lookup_
5979 order, which defaults to "bydns:byaddr". If people want to use the other
5980 order, or indeed, just use one or the other means of lookup, they can
5981 specify it in this variable.
5983 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5984 some operating systems, this comes back from gethostbyaddr() as an empty
5985 string, and this is what Exim used to test for. However, it seems that in
5986 other systems, "." is yielded. Exim now tests for this case too.
5988 7. The values of check_spool_space and check_log_space are now held internally
5989 as a number of kilobytes instead of an absolute number of bytes. If a
5990 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5991 kilobyte. This means that much larger values can be stored.
5993 8. Exim monitor: an attempt to get the action menu when not actually pointing
5994 at a message produces an empty menu entitled "No message selected". This
5995 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5996 no entries in it ("Shell widget menu has zero width and/or height"). So I
5997 have added a single, blank menu entry in this case.
5999 9. Added ${quote_local_part.
6001 10. MIME decoding is now applied to the contents of Subject: header lines when
6004 11. Now that a reference to $sender_host_address automatically causes a reverse
6005 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6006 host lookup before query-style lookups in lists that might use this
6007 variable. This has therefore been abolished, and the "net-" prefix is no
6008 longer necessary for query-style lookups.
6010 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6011 have been a typo for LFLAGS, so it has been changed.
6013 13. The install script calls Exim with "-C /dev/null" in order to find the
6014 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6015 to be output. However, since Exim outputs its version number before the
6016 error, it didn't break the script. It just looked ugly. I fixed this by
6017 always allowing "-C /dev/null" if the caller is root.
6019 14. Ignore overlarge ACL variable number when reading spool file - insurance
6020 against a later release with more variables having written the file.
6022 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6023 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6024 this, as well as the form without the "IPv6" on the front.
6026 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6027 OS/Makefile-Darwin file.
6029 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6030 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6032 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6033 variable. However, some OS use ints for inodes. I've added cast to long int
6034 to get rid of the compiler warning.
6036 19. I had forgotten to lock out "/../" in configuration file names when
6037 ALT_CONFIG_PREFIX was set.
6039 20. Routers used for verification do not need to specify transports. However,
6040 if such a router generated a host list, and callout was configured, Exim
6041 crashed, because it could not find a port number from the (non-existent)
6042 transport. It now assumes port 25 in this circumstance.
6044 21. Added the -t option to exigrep.
6046 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6047 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6048 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6049 them is set, LOOKUP_LSEARCH is forced.
6051 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6053 24. Added sender and host information to the "rejected by local_scan()" log
6054 line; previously there was no indication of these.
6056 25. Added .include_if_exists.
6058 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6059 Linux. It turns out that not all file systems support this. Apparently some
6060 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6061 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6062 not supported on the file descriptor, is now ignored when Exim is trying to
6063 sync a directory. This applies only to Linux.
6065 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6067 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6068 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6071 29. Added support for maildirsize files from supplied patch (modified a bit).
6073 30. The use of :fail: followed by an empty string could lead Exim to respond to
6074 sender verification failures with (e.g.):
6076 550 Verification failed for <xxx>
6077 550 Sender verify failed
6079 where the first response line was missing the '-' that indicates it is not
6080 the final line of the response.
6082 31. The loop for finding the name of the user that called Exim had a hardwired
6083 limit of 10; it now uses the value of finduser_retries, which is used for
6084 all other user lookups.
6086 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6089 33. Exim was neglecting to zero errno before one call of strtol() when
6090 expanding a string and expecting an integer value. On some systems this
6091 resulted in spurious "integer overflow" errors. Also, it was casting the
6092 result into an int without checking.
6094 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6095 did not work. The code looks as if it has *never* worked, though it appears
6096 to have been documented since at least release 1.62. I have made it work.
6098 35. The "timeout_DNS" error in retry rules, also documented since at least
6099 1.62, also never worked. As it isn't clear exactly what this means, and
6100 clearly it isn't a major issue, I have abolished the feature by treating it
6101 as "timeout", and writing a warning to the main and panic logs.
6103 36. The display of retry rules for -brt wasn't always showing the error code
6106 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6107 timeout_connect_A, timeout_connect_MX.
6109 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6110 to the empty sender.
6112 39. The daemon was not analysing the content of -oX till after it had closed
6113 stderr and disconnected from the controlling terminal. This meant that any
6114 syntax errors were only noted on the panic log, and the return code from
6115 the command was 0. By re-arranging the code a little, I've made the
6116 decoding happen first, so such errors now appear on stderr, and the return
6117 code is 1. However, the actual setting up of the sockets still happens in
6118 the disconnected process, so errors there are still only recorded on the
6121 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6122 connections (as happens on some IP stacks) was logged at start up time as
6123 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6124 it from "IPv6 and IPv4", which means that two separate sockets are being
6127 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6128 says whether AF_INET or AF_INET6 was passed as an argument.
6130 42. Exiwhat output was messed up when time zones were included in log
6133 43. Exiwhat now gives more information about the daemon's listening ports,
6134 and whether -tls-on-connect was used.
6136 44. The "port" option of the smtp transport is now expanded.
6138 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6139 silently ignored. Now an error message is written to the main and panic
6142 46. There's a new ACL modifier called "logwrite" which writes to a log file
6143 as soon as it is encountered.
6145 47. Added $local_user_uid and $local_user_gid at routing time.
6147 48. Exim crashed when trying to verify a sender address that was being
6150 49. Exim was recognizing only a space character after ".include". It now also
6151 recognizes a tab character.
6153 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6154 extracting the relevant information from the specification. The man page no
6155 longer contains scrambled data for the -d option, and I've added a section
6156 at the front about calling Exim under different names.
6158 51. Added "extra_headers" argument to the "mail" command in filter files.
6160 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6163 53. Installed eximstats 1.29.
6165 54. Added transport_filter_timeout as a generic transport option.
6167 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6168 Cc: header lines. This was required by RFC 822, but it not required by RFC
6171 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6172 incoming messages that did not have them. Now it does so only if the
6173 message originates locally, that is, if there is no associated remote host
6174 address. When Resent- header lines are present, this applies to the Resent-
6175 lines rather than the non-Resent- lines.
6177 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6178 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6180 58. Messages for configuration errors now include the name of the main
6181 configuration file - useful now that there may be more than one file in a
6182 list (.included file names were always shown).
6184 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6185 for those rare installations that do not start the daemon as root or run it
6186 setuid root. I've cut out the call to initgroups() if the daemon is not
6189 60. The Exim user and group can now be bound into the binary as text strings
6190 that are looked up at the start of Exim's processing.
6192 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6194 62. Added $mailstore_basename variable.
6196 63. Installed patch to sieve.c from Michael Haardt.
6198 64. When Exim failed to open the panic log after failing to open the main log,
6199 the original message it was trying to log was written to stderr and debug
6200 output, but if they were not available (the usual case in production), it
6201 was lost. Now it is written to syslog before the two lines that record the
6202 failures to open the logs.
6204 65. Users' Exim filters run in subprocesses under the user's uid. It is
6205 possible for a "deliver" command or an alias in a "personal" command to
6206 provoke an address rewrite. If logging of address rewriting is configured,
6207 this fails because the process is not running as root or exim. There may be
6208 a better way of dealing with this, but for the moment (because 4.30 needs
6209 to be released), I have disabled address rewrite logging when running a
6210 filter in a non-root, non-exim process.
6216 1. The buildconfig auxiliary program wasn't quoting the value set for
6217 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6218 not defined. This bug was masked in 4.22 by the effect that was fixed in
6221 2. Some messages that were rejected after a message id was allocated were
6222 shown as "incomplete" by exigrep. It no longer does this for messages that
6223 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6225 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6226 have allow_domain_literals set, the ID did not get logged in the <= line.
6227 Domain literals are now always recognized in Message-ID: header lines.
6229 4. The first argument for a ${extract expansion item is the key name or field
6230 number. Leading and trailing spaces in this item were not being ignored,
6231 causing some misleading effects.
6233 5. When deliver_drop_privilege was set, single queue runner processes started
6234 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6235 same command in the process it spins off) were not dropping privilege.
6237 6. When the daemon running as "exim" started a queue runner, it always
6238 re-executed Exim in the spun-off process. This is a waste of effort when
6239 deliver_drop_privilege is set. The new process now just calls the
6240 queue-runner function directly.
6246 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6249 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6250 ignored. Though the use of -oP was forcing the writing of a pid file, it
6251 was always written to the default place.
6253 3. If the message "no IP address found for host xxxx" is generated during
6254 incoming verification, it is now followed by identification of the incoming
6255 connection (so you can more easily find what provoked it).
6257 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6259 5. Added some features to "harden" Exim a bit more against certain attacks:
6261 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6262 be put in Local/Makefile. This is like the never_users runtime option,
6263 but it cannot be overridden. The default setting is "root".
6265 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6266 prefix string with which any file named in a -C command line option
6269 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6270 is retained for -C and -D only if the caller of Exim is root. Without
6271 it, the exim user may also use -C and -D and retain privilege.
6273 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6274 command line option is disabled.
6276 6. Macro names set by the -D option must start with an upper case letter, just
6277 like macro names defined in the configuration file.
6279 7. Added "dereference=" facility to LDAP.
6281 8. Two instances of the typo "uknown" in the source files are fixed.
6283 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6284 the Configure-Makefile script screwed up while processing it.
6286 10. Incorporated PCRE 4.4.
6288 11. The SMTP synchronization check was not operating right at the start of an
6289 SMTP session. For example, it could not catch a HELO sent before the client
6290 waited for the greeting. There is now a check for outstanding input at the
6291 point when the greeting is written. Because of the duplex, asynchronous
6292 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6293 way, but not yet received, when the check is performed.
6295 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6296 on TCP/IP sockets, because this apparently causes some broken clients to
6299 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6300 unchanged) from the Cygwin maintainer.
6302 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6303 was supported instead of testing for mbx. Effectively a typo.
6305 15. The spa authenticator server code was not checking that the input it
6306 received was valid base64.
6308 16. The debug output line for the "set" modifier in ACLs was not showing the
6309 name of the variable that was being set.
6311 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6313 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6314 lookup on its own. Something else was needed to trigger the lookup. For
6315 example, a match in host_lookup or the need for a host name in a host list.
6316 Now, if $sender_host_name is referenced and the host name has not yet been
6317 looked up, a lookup is performed. If the lookup fails, the variable remains
6318 empty, and $host_lookup_failed is set to "1".
6320 19. Added "eqi" as a case-independent comparison operator.
6322 20. The saslauthd authentication condition could segfault if neither service
6323 nor realm was specified.
6325 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6326 error message that was logged was misleading, and incoming SMTP
6327 connections were dropped. The message is now more accurate, and temporary
6328 errors are given to SMTP connections.
6330 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6331 (except RSET) with a 421 error, until QUIT is received. However, it was
6332 failing to send a response to QUIT.
6334 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6335 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6336 ACL is now run after the helo_try_verify_hosts code.
6338 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6339 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6340 case-independent, so other case variants are also recognized). Apparently
6341 some systems use these upper case variants.
6343 25. If more than two messages were waiting for the same host, and a transport
6344 filter was specified for the transport, Exim sent two messages over the
6345 same TCP/IP connection, and then failed with "socket operation on non-
6346 socket" when it tried to send the third.
6348 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6350 27. The extern definition of crypt16() in expand.c was not being excluded when
6351 the OS had its own crypt16() function.
6353 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6354 as a preferred synonym for return_size_limit, both as an option and as an
6357 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6359 30. Changed the default configuration ACL to relax the local part checking rule
6360 for addresses that are not in any local domains. For these addresses,
6361 slashes and pipe symbols are allowed within local parts, but the sequence
6362 /../ is explicitly forbidden.
6364 31. SPA server authentication was not clearing the challenge buffer before
6367 32. log_message in a "warn" ACL statement was writing to the reject log as
6368 well as to the main log, which contradicts the documentation and doesn't
6369 seem right (because no rejection is happening). So I have stopped it.
6371 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6372 However, I am unable to do any testing of this.
6374 34. Fixed an infelicity in the appendfile transport. When checking directories
6375 for a mailbox, to see if any needed to be created, it was accidentally
6376 using path names with one or more superfluous leading slashes; tracing
6377 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6379 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6380 contents are added to the log line that is written for every discarded
6381 recipient. (Previously a log_message setting was ignored.)
6383 36. The ${quote: operator now quotes the string if it is empty.
6385 37. The install script runs exim in order to find its version number. If for
6386 some reason other than non-existence or emptiness, which it checks, it
6387 could not run './exim', it was installing it with an empty version number,
6388 i.e. as "exim-". This error state is now caught, and the installation is
6391 38. An argument was missing from the function that creates an error message
6392 when Exim fails to connect to the socket for saslauthd authentication.
6393 This could cause Exim to crash, or give a corrupted message.
6395 39. Added isip, isip4, and isip6 to ${if conditions.
6397 40. The ACL variables $acl_xx are now saved with the message, and can be
6398 accessed later in routers, transports, and filters.
6400 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6401 strings in the file are not string-expanded.
6403 42. If a MAIL command specified a SIZE value that was too large to fit into an
6404 int variable, the check against message_size_limit failed. Such values are
6405 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6406 day this will have to be increased, but I don't think I want to be around
6407 when emails are that large.
6414 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6415 iconv() is not standard in FreeBSD.
6417 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6418 IPv6 enabled. The observed symptom was a segmentation fault on return from
6419 the function os_common_find_running_interfaces() in src/os.c.
6421 3. In the check_special_case() function in daemon.c I had used "errno" as an
6422 argument name, which causes warnings on some systems. This was basically a
6423 typo, since it was named "eno" in the comments!
6425 4. The code that waits for the clock to tick (at a resolution of some fraction
6426 of a second) so as to ensure message-id uniqueness was always waiting for
6427 at least one whole tick, when it could have waited for less. [This is
6428 almost certainly not relevant at current processor speeds, where it is
6429 unlikely to ever wait at all. But we try to future-proof.]
6431 5. The function that sleeps for a time interval that includes fractions of a
6432 second contained a race. It did not block SIGALRM between setting the
6433 timer, and suspending (a couple of lines later). If the interval was short
6434 and the sigsuspend() was delayed until after it had expired, the suspension
6435 never ended. On busy systems this could lead to processes getting stuck for
6438 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6439 process, before it forks any delivery processes. The open lookup caching
6440 mechanism meant that the open file or database connection was passed into
6441 the delivery process. The problem was that delivery processes always tidy
6442 up cached lookup data. This could cause a problem for the next delivery
6443 process started by the queue runner, because the external queue runner
6444 process does not know about the closure. So the next delivery process
6445 still has data in the lookup cache. In the case of a file lookup, there was
6446 no problem because closing a file descriptor in a subprocess doesn't affect
6447 the parent. However, if the lookup was caching a connection to a database,
6448 the connection was closed, and the second delivery process was likely to
6449 see errors such as "PGSQL: query failed: server closed the connection
6450 unexpectedly". The problem has been fixed by closing all cached lookups
6451 in a queue runner before running a delivery process.
6453 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6454 seem to have the "const" qualifier which it has on other OS. I've
6457 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6458 *of the same type* (client or server) with the same public name that an
6459 error should be diagnosed.
6461 9. When Exim looked up a host name for an IP address, but failed to find the
6462 original IP address when looking up the host name (a safety check), it
6463 output the message "<ip address> does not match any IP for NULL", which was
6464 confusing, to say the least. The bug was that the host name should have
6465 appeared instead of "NULL".
6467 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6468 user that is built into the binary, and the -C or -D options is used, root
6469 privilege is dropped before the configuration file is read. In addition,
6470 logging is switched to stderr instead of the normal log files. If the
6471 configuration then re-defines the Exim user, the unprivileged environment
6472 is probably not what is expected, so Exim logs a panic warning message (but
6475 However, if deliver_drop_privilege is set, the unprivileged state may well
6476 be exactly what is intended, so the warning has been cut out in that case,
6477 and Exim is allowed to try to write to its normal log files.
6483 1. smtp_return_error_details was not giving details for temporary sender
6484 or receiver verification errors.
6486 2. Diagnose a configuration error if two authenticators have the same public
6489 3. Exim used not to create the message log file for a message until the first
6490 delivery attempt. This could be confusing when incoming messages were held
6491 for policy or load reasons. The message log file is now created at the time
6492 the message is received, and an initial "Received" line is written to it.
6494 4. The automatically generated man page for command line options had a minor
6495 bug that caused no ill effects; however, a more serious problem was that
6496 the procedure for building the man page automatically didn't always
6497 operate. Consequently, release 4.20 contains an out-of-date version. This
6498 shouldn't happen again.
6500 5. When building Exim with embedded Perl support, the script that builds the
6501 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6502 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6504 6. The freeze_tell option was not being used for messages that were frozen on
6505 arrival, either by an ACL or by local_scan().
6507 7. Added the smtp_incomplete_transaction log selector.
6509 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6510 was accepting AUTH without a new EHLO.
6512 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6513 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6516 10. Logging of TCP/IP connections (when configured) now happens in the main
6517 daemon process instead of the child process, so that the TCP/IP connection
6518 count is more accurate (but it can never be perfect).
6520 11. The use of "drop" in a nested ACL was not being handled correctly in the
6521 outer ACL. Now, if condition failure induced by the nested "drop" causes
6522 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6523 or "require"), the connection is dropped.
6525 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6526 that yield "discard" can now be used with an "accept" or a "discard" verb,
6527 but an error is generated for any others (because I can't see a useful way
6528 to define what should happen).
6530 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6531 are now processed in the same way as lines in the Exim configuration file.
6532 In particular, continuation lines are supported.
6534 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6536 15. Added -ti meaning -t -i.
6538 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6539 domains, and warn by logging if others are found.
6541 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6542 returned from GIFCONF when Exim is trying to find the list of interfaces on
6543 a host. The code in os.c has been modified to copy each ifreq to an aligned
6544 structure in all cases.
6546 Also, in some cases, the returned ifreq's were being copied to a 'struct
6547 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6548 means the last couple of bytes of an IPv6 address could be chopped if the
6549 ifreq contained only a normal sockaddr (14 bytes storage).
6551 18. Named domain lists were not supported in the hosts_treat_as_local option.
6552 An entry such as +xxxx was not recognized, and was treated as a literal
6555 19. Ensure that header lines added by a DATA ACL are included in the reject log
6556 if the ACL subsequently rejects the message.
6558 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6559 MD5 (which is deprecated).
6561 21. When testing a filter file using -bf, Exim was writing a message when it
6562 took the sender from a "From " line in the message, but it was not doing so
6563 when it took $return_path from a Return-Path: header line. It now does.
6565 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6566 with a valid header line field name (a series of printing characters
6567 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6569 23. Changed "disc" in the source to "disk" to conform to the documentation and
6570 the book and for uniformity.
6572 24. Ignore Sendmail's -Ooption=value command line item.
6574 25. When execve() failed while trying to run a command in a pipe transport,
6575 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6576 could be confused with a return value of 69 from the command itself. This
6577 has been changed to 127, the value the shell returns if it is asked to run
6578 a non-existent command. The wording for the related log line suggests a
6579 non-existent command as the problem.
6581 26. If received_header_text expands to an empty string, do not add a Received:
6582 header line to the message. (Well, it adds a token one on the spool, but
6583 marks it "old" so that it doesn't get used or transmitted.)
6585 27. Installed eximstats 1.28 (addition of -nt option).
6587 28. There was no check for failure on the call to getsockname() in the daemon
6588 code. This can fail if there is a shortage of resources on the system, with
6589 ENOMEM, for example. A temporary error is now given on failure.
6591 29. Contrary to the C standard, it seems that in some environments, the
6592 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6593 program. Exim now does this explicitly; it affects the formatting of
6594 timestamps using strftime().
6596 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6597 complaints. I've now initialized all the variables, to avoid this.
6599 32. Header lines added by a system filter were not being "seen" during
6600 transport-time rewrites.
6602 33. The info_callback() function passed to OpenSSL is set up with type void
6603 (*)(SSL *, int, int), as described somewhere. However, when calling the
6604 function (actually a macro) that sets it up, the type void(*)() is
6605 expected. I've put in a cast to prevent warnings from picky compilers.
6607 34. If a DNS black list lookup found a CNAME record, but there were no A
6608 records associated with the domain it pointed at, Exim crashed.
6610 35. If a DNS black list lookup returned more than one A record, Exim ignored
6611 all but the first. It now scans all returned addresses if a particular IP
6612 value is being sought. In this situation, the contents of the
6613 $dnslist_value variable are a list of all the addresses, separated by a
6616 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6617 to accept a host name and all its aliases if the forward lookup for any of
6618 them yielded the IP address of the incoming connection. Now it accepts only
6619 those names whose forward lookup yields the correct IP address. Any other
6620 names are discarded. This closes a loophole whereby a rogue DNS
6621 administrator could create reverse DNS records to break through a
6622 wildcarded host restriction in an ACL.
6624 37. If a user filter or a system filter that ran in a subprocess used any of
6625 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6626 the wrong values were passed to the pipe command ($thisaddress had the
6627 value of $0, $0 had the value of $1, etc). This bug was introduced by
6628 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6630 38. Improved the line breaking for long SMTP error messages from ACLs.
6631 Previously, if there was no break point between 40 and 75 characters, Exim
6632 left the rest of the message alone. Two changes have been made: (a) I've
6633 reduced the minimum length to 35 characters; (b) if it can't find a break
6634 point between 35 and 75 characters, it looks ahead and uses the first one
6635 that it finds. This may give the occasional overlong line, but at least the
6636 remaining text gets split now.
6638 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6639 file descriptors might be low, and that setting 1000 would always raise it.
6640 It turns out that in some environments, the limit is already over 1000 and
6641 that lowering it causes trouble. So now Exim takes care not to decrease it.
6643 40. When delivering a message, the value of $return_path is set to $sender_
6644 address at the start of routing (routers may change the value). By an
6645 oversight, this default was not being set up when an address was tested by
6646 -bt or -bv, which affected the outcome if any router or filter referred to
6649 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6650 log only when "message" or "log_message" are set. However, if one of the
6651 conditions was an address verification, or a call to a nested ACL, the
6652 messages generated by the underlying test were being passed through. This
6653 no longer happens. The underlying message is available in $acl_verify_
6654 message for both "message" and "log_message" expansions, so it can be
6655 passed through if needed.
6657 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6658 new expansion $bh_ to give the encoded byte string without charset
6659 translation. Translation happens only if iconv() is available; HAVE_ICONV
6660 indicates this at build time. HEADERS_CHARSET gives the charset to
6661 translate to; headers_charset can change it in the configuration, and
6662 "headers charset" can change it in an individual filter file.
6664 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6665 that creates RFC 2047 encoded "words" labels them as that charset instead
6666 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6667 expansion operator; (ii) when Exim creates a From: line for a local
6668 message; (iii) when a header line is rewritten to include a "phrase" part.
6670 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6671 buggy, causing it to skip the first lines of messages whose message ID
6672 ended in 'D'. This would not have bitten before Exim release 4.14, because
6673 message IDs were unlikely to end in 'D' before then. The effect was to have
6674 incorrect size information for certain domains.
6676 45. #include "config.h" was missing at the start of the crypt16.c module. This
6677 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6680 46. If there was a timeout during a "random" callout check, Exim treated it as
6681 a failure of the random address, and carried on sending RSET and the real
6682 address. If the delay was just some slowness somewhere, the response to the
6683 original RCPT would be taken as a response to RSET and so on, causing
6684 mayhem of various kinds.
6686 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6687 when I implemented it. It didn't allow for the fact that some option values
6688 may legitimately be negative (e.g. size_addition), and it didn't even do
6689 the right test for positive values.
6691 48. Domain names in DNS records are case-independent. Exim always looks them up
6692 in lower case. Some resolvers return domain names in exactly the case they
6693 appear in the zone file, that is, they may contain uppercase letters. Not
6694 all resolvers do this - some return always lower case. Exim was treating a
6695 change of case by a resolver as a change of domain, similar to a widening
6696 of a domain abbreviation. This triggered its re-routing code and so it was
6697 trying to route what was effectively the same domain again. This normally
6698 caused routing to fail (because the router wouldn't handle the domain
6699 twice). Now Exim checks for this case specially, and just changes the
6700 casing of the domain that it ultimately uses when it transmits the message
6703 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6706 50. If a filter generated a file delivery with a non-absolute name (possible if
6707 no home directory exists for the router), the forbid_file option was not
6710 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6711 the existing equality matching.
6713 52. Exim was using ints instead of ino_t variables in some places where it was
6714 dealing with inode numbers.
6716 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6717 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6718 called TMPDIR, and if it finds it is different, it changes its value.
6720 54. The smtp_printf() function is now made available to local_scan() so
6721 additional output lines can be written before returning. There is also an
6722 smtp_fflush() function to enable the detection of a dropped connection.
6723 The variables smtp_input and smtp_batched_input are exported to
6726 55. Changed the default runtime configuration: the message "Unknown user"
6727 has been removed from the ACL, and instead placed on the localuser router,
6728 using the cannot_route_message feature. This means that any verification
6729 failures that generate their own messages won't get overridden. Similarly,
6730 the "Unrouteable address" message that was in the ACL for unverifiable
6731 relay addresses has also been removed.
6733 56. Added hosts_avoid_esmtp to the smtp transport.
6735 57. The exicyclog script was not checking for the esoteric option
6736 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6737 will work only if exicyclog is run under the appropriate euid.
6739 58. Following a discussion on the list, the rules by which Exim recognises line
6740 endings on incoming messages have been changed. The -dropcr and drop_cr
6741 options are now no-ops, retained only for backwards compatibility. The
6742 following line terminators are recognized: LF CRLF CR. However, special
6743 processing applies to CR:
6745 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6746 nor a local message in the state where . is a terminator.
6748 (ii) If a bare CR is encountered in a header line, an extra space is added
6749 after the line terminator so as not to end the header. The reasoning
6750 behind this is that bare CRs in header lines are most likely either
6751 to be mistakes, or people trying to play silly games.
6753 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6754 was being incorrectly given as 18 bytes larger than it should have been.
6755 This is a VOB (very old bug).
6757 60. This may never have affected anything current, but just in case it has:
6758 When the local host is found other than at the start of a list of hosts,
6759 the local host, those with the same MX, and any that follow, are discarded.
6760 When the list in question was part of a longer list of hosts, the following
6761 hosts (not currently being processed) were also being discarded. This no
6762 longer happens. I'm not sure if this situation could ever has previously
6765 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6768 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6769 header. This is something that is recommended in a new Internet Draft, and
6770 is something that is documented as being done by Sendmail. There are two
6771 possible values. For messages generated by the autoreply transport, Exim
6774 Auto-Submitted: auto-replied
6776 whereas for all other generated messages (e.g. bounces) it adds
6778 Auto-Submitted: auto-generated
6780 63. The "personal" condition in filters now includes a test for the
6781 Auto-Submitted: header. If it contains the string "auto-" the message it
6782 not considered personal.
6784 64. Added rcpt_include_affixes as a generic transport option.
6786 65. Added queue_only_override (default true).
6788 66. Added the syslog_duplication option.
6790 67. If what should have been the first header line of a message consisted of
6791 a space followed by a colon, Exim was mis-interpreting it as a header line.
6792 It isn't of course - it is syntactically invalid and should therefore be
6793 treated as the start of the message body. The misbehaviour could have
6794 caused a number of strange effects, including loss of data in subsequent
6795 header lines, and spool format errors.
6797 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6798 client host had authenticated. This control can now be exercised by an ACL
6799 for more flexibility.
6801 69. By default, callouts do not happen when testing with -bh. There is now a
6802 variant, -bhc, which does actually run the callout code, including
6803 consulting and updating the callout cache.
6805 70. Added support for saslauthd authentication, courtesy of Alexander
6808 71. If statvfs() failed on the spool or log directories while checking their
6809 size for availability, Exim confusingly gave the error "space shortage".
6810 Furthermore, in debugging mode it crashed with a floating point exception.
6811 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6812 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6813 serious problem, Exim now writes to the main and panic logs when this
6814 happens, with details of the failure. It then refuses to accept the
6815 incoming message, giving the message "spool directory problem" or "log
6816 directory problem" with a 421 code for SMTP messages.
6818 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6819 0, 1, and 2 exist, because some OS complain for execs without them (see
6820 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6821 descriptors. However, the code omitted to check that the open succeeded,
6822 causing mysterious errors if for some reason the permissions on /dev/null
6823 got screwed. Now Exim writes a message to the main and panic logs, and
6824 bombs out if it can't open /dev/null.
6826 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6827 interact so that it is all more flexible. It is supposed to remain
6828 backwards compatible. Also added extra_local_interfaces.
6830 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6831 to bomb out with an assertion failure - to the client this appears as a
6832 connection drop. This problem occurs in the part of the code that was taken
6833 from the Samba project. Fortunately, the assertion is in a very simple
6834 function, so I have fixed this by reproducing the function inline in the
6835 one place where it is called, and arranging for authentication to fail
6836 instead of killing the process with assert().
6838 75. The SPA client code was not working when the server requested OEM rather
6839 than Unicode encoding.
6841 76. Added code to make require_files with a specific uid setting more usable in
6842 the case where statting the file as root fails - usually a non-root-mounted
6843 NFS file system. When this happens and the failure is EACCES, Exim now
6844 forks a subprocess and does the per-uid checking as the relevant uid.
6846 77. Added process_log_path.
6848 78. If log_file_path was not explicitly set, a setting of check_log_space or
6849 check_log_inodes was ignored.
6851 79. If a space check for the spool or log partitions fails, the incident is now
6852 logged. Of course, in the latter case the data may get lost...
6854 80. Added the %p formatting code to string_format() so that it can be used to
6855 print addresses in debug_print(). Adjusted all the address printing in the
6856 debugging in store.c to use %p rather than %d.
6858 81. There was a concern that a line of code in smtp_in.c could overflow a
6859 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6860 initially expressed, the concern was not well-founded, because trailing
6861 spaces are removed early. However, if the trailing spaces were followed by
6862 a NULL, they did not get removed, so the overflow was possible. Two fixes
6865 (a) I re-wrote the offending code in a cleaner fashion.
6866 (b) If an incoming SMTP command contains a NULL character, it is rejected
6869 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6870 runs initgroups(), so that if the Exim user is in any additional groups,
6871 they will be used during message reception.
6877 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob