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.
121 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
122 Previously it was constructed the first time a possibly-matching string
123 was met in the configuration file input during startup; now it is done
126 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
127 compatible one, to avoid the (poorly documented) possibility of a config
128 file in the working directory redirecting the DB files, possibly correpting
129 some existing file. CVE-2017-10140 assigned for BDB.
131 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
132 cache-hot. Previously, although the result was properly cached, the
133 initial verify call returned a defer.
135 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
136 the main verify for receipient in uncached-mode.
138 JH/22 Retire historical build files to an "unsupported" subdir. These are
139 defined as "ones for which we have no current evidence of testing".
141 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
142 if present. Previously it was ignored.
144 JH/24 Start using specified-initialisers in C structure init coding. This is
145 a C99 feature (it's 2017, so now considered safe).
147 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
148 if was a fixed-sized field and bitmask ops via macros; it is now more
151 PP/07 GitHub PR 56: Apply MariaDB build fix.
152 Patch provided by Jaroslav Škarvada.
154 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
155 during Coverity cleanups [4.87 JH/47]
156 Diagnosis and fix provided by Michael Fischer v. Mollard.
158 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
159 the right size to place the terminating semicolon on its own folded
160 line, the header hash was calculated to an incorrect value thanks to
161 the (relaxed) space the fold became.
163 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
164 and confused the parent.
166 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
167 which could crash as a result. This could lead to undeliverable messages.
169 JH/28 Logging: "next input sent too soon" now shows where input was truncated
172 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
173 matters on fast-turnover and PID-randomising systems, which were getting
174 out-of-order delivery.
176 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
177 a possibly-overlapping copy. The symptom was that "Remote host closed
178 connection in response to HELO" was logged instead of the actual 4xx
185 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
186 than -2003 did; needs libidn2 in addition to libidn.
188 JH/02 The path option on a pipe transport is now expanded before use.
190 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
191 Patch provided by "Björn", documentation fix added too.
193 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
194 missing a wire-to-host endian conversion.
196 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
197 close after a BDAT command line could be taken as a following command,
198 giving a synch failure. Fix by only checking for synch immediately
199 before acknowledging the chunk.
201 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
202 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
203 macro. Patches provided by Josh Soref.
205 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
206 Previously we did not; the RFC seems ambiguous and VRFY is not listed
207 by IANA as a service extension. However, John Klensin suggests that we
210 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
211 the dkim code may be unix-mode line endings rather than smtp wire-format
212 CRLF, so prepend a CR to any bare LF.
214 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
215 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
217 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
218 the first were themselves being wrongly included in the feed into dkim
219 processing; with most chunk sizes in use this resulted in an incorrect
220 body hash calculated value.
222 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
223 DKIM signature block, for verification. Although advised against by
224 standards it is specifically not ruled illegal.
226 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
228 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
229 missing a body hash (the bh= tag).
231 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
232 It seems that HAProxy sends the Proxy Protocol information in clear and
233 only then does a TLS startup, so do the same.
235 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
236 TCP connections (such as for Spamd) unless the daemon successfully set
237 Fast Open mode on its listening sockets. This fixes breakage seen on
238 too-old kernels or those not configured for Fast Open, at the cost of
239 requiring both directions being enabled for TFO, and TFO never being used
240 by non-daemon-related Exim processes.
242 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
243 endings, at least on the first header line. Try to canonify any that get
244 past that check, despite the cost.
246 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
247 now limited to an arbitrary five deep, while parsing addresses with the
248 strip_excess_angle_brackets option enabled.
250 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
251 instead leave the unprompted TLS handshake in socket buffer for the
252 TLS library to consume.
254 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
256 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
258 JH/16 Drop variables when they go out of scope. Memory management drops a whole
259 region in one operation, for speed, and this leaves assigned pointers
260 dangling. Add checks run only under the testsuite which checks all
261 variables at a store-reset and panics on a dangling pointer; add code
262 explicitly nulling out all the variables discovered. Fixes one known
263 bug: a transport crash, where a dangling pointer for $sending_ip_address
264 originally assigned in a verify callout, is re-used.
266 PP/06 Drop '.' from @INC in various Perl scripts.
268 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
270 PP/08 Reduce a number of compilation warnings under clang; building with
271 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
272 should be warning-free.
274 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
276 HS/01 Fix portability problems introduced by PP/08 for platforms where
277 realloc(NULL) is not equivalent to malloc() [SunOS et al].
279 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
280 chunk. This allows us to accept broken chunked messages. We need a more
281 general solution here.
283 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
284 already-broken messages in the queue.
286 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
288 JH/19 Fix reference counting bug in routing-generated-address tracking.
294 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
295 supports it and a size is available (ie. the sending peer gave us one).
297 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
298 years of being deprecated). The replacements are the ACLs
299 acl_smtp_mime and acl_not_smtp_mime.
301 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
302 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
303 or even an in-clear connection were permitted. Now, if the host lookup
304 was dnssec and dane was requested then the host is only used if the
305 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
306 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
307 if one fails this test.
308 This means that a poorly-configured remote DNS will make it incommunicado;
309 but it protects against a DNS-interception attack on it.
311 JH/04 Bug 1810: make continued-use of an open smtp transport connection
312 non-noisy when a race steals the message being considered.
314 JH/05 If main configuration option tls_certificate is unset, generate a
315 self-signed certificate for inbound TLS connections.
317 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
318 in rewrites and routers.
320 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
321 and logged a warning sing 4.83; now they are a configuration file error.
323 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
324 (lacking @domain). Apply the same qualification processing as RCPT.
326 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
328 JH/10 Support ${sha256:} applied to a string (as well as the previous
331 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
332 a cutthrough deliver is pending, as we always want to make a connection.
333 This also avoids re-routing the message when later placing the cutthrough
334 connection after a verify cache hit.
335 Do not update it with the verify result either.
337 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
338 when routing results in more than one destination address.
340 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
341 signing (which inhibits the cutthrough capability). Previously only
342 the presence of an option was tested; now an expansion evaluating as
343 empty is permissible (obviously it should depend only on data available
344 when the cutthrough connection is made).
346 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
347 the relevant preceding SMTP command did not note the pipelining mode.
349 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
350 Previously they were not counted.
352 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
353 as one having no matching records. Previously we deferred the message
354 that needed the lookup.
356 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
357 distinguished as "(=".
359 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
360 for missing MX records. Previously it only worked for missing A records.
362 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
364 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
365 after the data-go-ahead and data-ack. Patch from Jason Betts.
367 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
368 even for a "none" policy. Patch from Tony Meyer.
370 JH/22 Fix continued use of a connection for further deliveries. If a port was
371 specified by a router, it must also match for the delivery to be
374 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
375 When one of the recipients of a message was unsuitable for the connection
376 (has no matching addresses), we lost track of needing to mark it
377 deferred. As a result mail would be lost.
379 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
381 JH/25 Decoding ACL controls is now done using a binary search; the source code
382 takes up less space and should be simpler to maintain. Merge the ACL
383 condition decode tables also, with similar effect.
385 JH/26 Fix problem with one_time used on a redirect router which returned the
386 parent address unchanged. A retry would see the parent address marked as
387 delivered, so not attempt the (identical) child. As a result mail would
390 JH/27 Fix a possible security hole, wherein a process operating with the Exim
391 UID can gain a root shell. Credit to http://www.halfdog.net/ for
392 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
395 JH/28 Enable {spool,log} filesystem space and inode checks as default.
396 Main config options check_{log,spool}_{inodes,space} are now
397 100 inodes, 10MB unless set otherwise in the configuration.
399 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
400 Previously it only applied to the main-section connection policy
403 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
405 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
406 by me. Added RFC7919 DH primes as an alternative.
408 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
409 are not in the system include path.
411 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
412 GnuTLS, when a session startup failed (eg because the client disconnected)
413 Exim did stdio operations after fclose. This was exposed by a recent
414 change which nulled out the file handle after the fclose.
416 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
417 signed directly by the cert-signing cert, rather than an intermediate
418 OCSP-signing cert. This is the model used by LetsEncrypt.
420 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
422 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
423 an incoming connection.
425 HS/02 Bug 1802: Do not half-close the connection after sending a request
428 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
429 fallback to "prime256v1".
431 JH/34 SECURITY: Use proper copy of DATA command in error message.
432 Could leak key material. Remotely exploitable. CVE-2016-9963.
438 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
439 and 3.4.4 - once the server is enabled to respond to an OCSP request
440 it does even when not requested, resulting in a stapling non-aware
441 client dropping the TLS connection.
443 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
444 support variable-length bit vectors. No functional change.
446 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
447 The I= interface field on outgoing lines is now after the H= remote
448 host field, same as incoming lines. There is a separate
449 outgoing_interface log selector which allows you to disable the
452 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
453 If not running log_selector +smtp_connection the mainlog would be held
454 open indefinitely after a "too many connections" event, including to a
455 deleted file after a log rotate. Leave the per net connection logging
456 leaving it open for efficiency as that will be quickly detected by the
457 check on the next write.
459 HS/01 Bug 1671: Fix post transport crash.
460 Processing the wait-<transport> messages could crash the delivery
461 process if the message IDs didn't exist for some reason. When
462 using 'split_spool_directory=yes' the construction of the spool
463 file name failed already, exposing the same netto behaviour.
465 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
466 mime_regex ACL conditions.
468 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
469 to DSN fail messages (bounces): remote IP, remote greeting, remote response
470 to HELO, local diagnostic string.
472 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
473 log line to debug. Even when configured with a tls authenticator many
474 client connections are expected to not authenticate in this way, so
475 an authenticate fail is not an error.
477 HS/02 Add the Exim version string to the process info. This way exiwhat
478 gives some more detail about the running daemon.
480 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
481 matter for fast-change records such as DNSBLs.
483 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
484 retry record, even if constant. There may be multiple transports with
485 different interface settings and the retry behaviour needs to be kept
488 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
490 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
492 JH/10 Bug 840: fix log_defer_output option of pipe transport
494 JH/11 Bug 830: use same host for all RCPTS of a message, even under
495 hosts_randomize. This matters a lot when combined with mua_wrapper.
497 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
498 ${quote_pgsql:<string>} operator.
500 JH/13 Bug 1708: avoid misaligned access in cached lookup.
502 JH/14 Change header file name for freeradius-client. Relevant if compiling
503 with Radius support; from the Gentoo tree and checked under Fedora.
505 JH/15 Bug 1712: Introduce $prdr_requested flag variable
507 JH/16 Bug 1714: Permit an empty string as expansion result for transport
508 option transport_filter, meaning no filtering.
510 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
512 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
513 defaults to "*" (all hosts). The variable is now available when not built
514 with TLS, default unset, mainly to enable keeping the testsuite sane.
515 If a server certificate is not supplied (via tls_certificate) an error is
516 logged, and clients will find TLS connections fail on startup. Presumably
517 they will retry in-clear.
518 Packagers of Exim are strongly encouraged to create a server certificate
519 at installation time.
521 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
522 with the $config_file variable.
524 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
525 in transport context, after the attempt, and per-recipient. The latter type
526 is per host attempted. The event data is the error message, and the errno
527 information encodes the lookup type (A vs. MX) used for the (first) host,
528 and the trailing two digits of the smtp 4xx response.
530 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
531 to write to mainlog (or rejectlog, paniclog) in the window between file
532 creation and permissions/ownership being changed. Particularly affects
533 installations where exicyclog is run as root, rather than exim user;
534 result is that the running daemon panics and dies.
536 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
538 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
539 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
540 "pri" and "weight". Note that the previous implicit priority given by the
541 list order is no longer honoured.
543 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
546 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
547 by defining SUPPORT_SOCKS.
549 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
550 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
551 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
552 variables are renamed to proxy_{local,external}_{address,port}.
554 JH/27 Move Internationalisation support from Experimental to mainline, enabled
555 for a build by defining SUPPORT_I18N
557 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
558 of the query string, and make ${quote_redis:} do that quoting.
560 JH/29 Move Events support from Experimental to mainline, enabled by default
561 and removable for a build by defining DISABLE_EVENT.
563 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
565 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
566 cached by the daemon.
568 JH/32 Move Redis support from Experimental to mainline, enabled for a build
569 by defining LOOKUP_REDIS. The libhiredis library is required.
571 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
572 keys are given for lookup.
574 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
575 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
576 only supported when built with TLS support. The PolarSSL SHA routines
577 are still used when the TLS library is too old for convenient support.
579 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
580 openssl_options), for security. OpenSSL forces this from version 1.1.0
581 server-side so match that on older versions.
583 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
584 allocation for $value could be released as the expansion processing
585 concluded, but leaving the global pointer active for it.
587 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
588 and to use the domains and local_parts ACL conditions.
590 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
591 incorrectly not doubled on cutthrough transmission, hence seen as a
592 body-termination at the receiving system - resulting in truncated mails.
593 Commonly the sender saw a TCP-level error, and retransmitted the message
594 via the normal store-and-forward channel. This could result in duplicates
595 received - but deduplicating mailstores were liable to retain only the
596 initial truncated version.
598 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
600 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
602 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
603 we're in there, support oversigning also; bug 1309.
605 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
607 HS/04 Add support for keep_environment and add_environment options.
609 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
610 either intentional arithmetic overflow during PRNG, or testing config-
613 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
614 delivery resulted in actual delivery. Cancel cutthrough before DATA
617 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
618 rejects a recipient: pass the reject to the originator.
620 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
621 Many were false-positives and ignorable, but it's worth fixing the
624 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
625 for the new environment-manipulation done at startup. Move the routines
626 from being local to tls.c to being global via the os.c file.
628 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
629 an extract embedded as result-arg for a map, the first arg for extract
630 is unavailable so we cannot tell if this is a numbered or keyed
631 extraction. Accept either.
637 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
640 JH/02 The smtp transport option "multi_domain" is now expanded.
642 JH/03 The smtp transport now requests PRDR by default, if the server offers
645 JH/04 Certificate name checking on server certificates, when exim is a client,
646 is now done by default. The transport option tls_verify_cert_hostnames
647 can be used to disable this per-host. The build option
648 EXPERIMENTAL_CERTNAMES is withdrawn.
650 JH/05 The value of the tls_verify_certificates smtp transport and main options
651 default to the word "system" to access the system default CA bundle.
652 For GnuTLS, only version 3.0.20 or later.
654 JH/06 Verification of the server certificate for a TLS connection is now tried
655 (but not required) by default. The verification status is now logged by
656 default, for both outbound TLS and client-certificate supplying inbound
659 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
662 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
663 Status Notification (bounce) messages are now MIME format per RFC 3464.
664 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
665 under the control of the dsn_advertise_hosts option, and routers may
666 have a dsn_lasthop option.
668 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
669 default, modifiable by a malware= option. The list separator for
670 the options can now be changed in the usual way. Bug 68.
672 JH/10 The smtp_receive_timeout main option is now expanded before use.
674 JH/11 The incoming_interface log option now also enables logging of the
675 local interface on delivery outgoing connections.
677 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
678 if the interface and destination host and port all match.
680 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
683 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
684 Patch from Andrew Lewis.
686 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
687 now supports optional time-restrictions, weighting, and priority
688 modifiers per server. Patch originally by <rommer@active.by>.
690 JH/16 The spamd_address main option now supports a mixed list of local
691 and remote servers. Remote servers can be IPv6 addresses, and
692 specify a port-range.
694 JH/17 Bug 68: The spamd_address main option now supports an optional
695 timeout value per server.
697 JH/18 Bug 1581: Router and transport options headers_add/remove can
698 now have the list separator specified.
700 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
703 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
706 JH/21 Support for the A6 type of dns record is withdrawn.
708 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
709 rather than the verbs used.
711 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
712 from 255 to 1024 chars.
714 JH/24 Verification callouts now attempt to use TLS by default.
716 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
717 are generic router options now. The defaults didn't change.
719 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
720 Original patch from Alexander Shikoff, worked over by JH.
722 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
723 files if ZCAT_COMMAND is not executable.
725 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
727 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
729 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
730 Normally benign, it bites when the pair was led to by a CNAME;
731 modern usage is to not canonicalize the domain to a CNAME target
732 (and we were inconsistent anyway for A-only vs AAAA+A).
734 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
736 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
737 when evaluating $sender_host_dnssec.
739 JH/31 Check the HELO verification lookup for DNSSEC, adding new
740 $sender_helo_dnssec variable.
742 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
744 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
746 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
748 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
749 documented as working, but never had. Support all but $spam_report.
751 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
752 added for tls authenticator.
754 HS/03 Add perl_taintmode main config option
760 TL/01 When running the test suite, the README says that variables such as
761 no_msglog_check are global and can be placed anywhere in a specific
762 test's script, however it was observed that placement needed to be near
763 the beginning for it to behave that way. Changed the runtest perl
764 script to read through the entire script once to detect and set these
765 variables, reset to the beginning of the script, and then run through
766 the script parsing/test process like normal.
768 TL/02 The BSD's have an arc4random API. One of the functions to induce
769 adding randomness was arc4random_stir(), but it has been removed in
770 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
771 function when detected.
773 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
774 cause callback expansion.
776 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
777 syntax errors in an expansion can be treated as a string instead of
778 logging or causing an error, due to the internal use of bool_lax
779 instead of bool when processing it.
781 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
782 server certificates when making smtp deliveries.
784 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
786 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
788 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
790 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
791 Merged patch from Sebastian Wiedenroth.
793 JH/05 Fix results-pipe from transport process. Several recipients, combined
794 with certificate use, exposed issues where response data items split
795 over buffer boundaries were not parsed properly. This eventually
796 resulted in duplicates being sent. This issue only became common enough
797 to notice due to the introduction of connection certificate information,
798 the item size being so much larger. Found and fixed by Wolfgang Breyha.
800 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
801 size buffer was used, resulting in syntax errors when an expansion
804 JH/07 Add support for directories of certificates when compiled with a GnuTLS
805 version 3.3.6 or later.
807 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
808 is EXPERIMENTAL_EVENT, the main-configuration and transport options
809 both become "event_action", the variables become $event_name, $event_data
810 and $event_defer_errno. There is a new variable $verify_mode, usable in
811 routers, transports and related events. The tls:cert event is now also
812 raised for inbound connections, if the main configuration event_action
815 TL/06 In test suite, disable OCSP for old versions of openssl which contained
816 early OCSP support, but no stapling (appears to be less than 1.0.0).
818 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
819 server certificate names available under the smtp transport option
820 "tls_verify_cert_hostname" now do not permit multi-component wildcard
823 JH/10 Time-related extraction expansions from certificates now use the main
824 option "timezone" setting for output formatting, and are consistent
825 between OpenSSL and GnuTLS compilations. Bug 1541.
827 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
828 encoded parameter in the incoming message. Bug 1558.
830 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
831 include certificate info, eximon was claiming there were spoolfile
834 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
836 JH/14 Log delivery-related information more consistently, using the sequence
837 "H=<name> [<ip>]" wherever possible.
839 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
840 are problematic for Debian distribution, omit them from the release
843 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
845 JH/16 Fix string representation of time values on 64bit time_t architectures.
848 JH/17 Fix a null-indirection in certextract expansions when a nondefault
849 output list separator was used.
854 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
855 checkers that were complaining about end of non-void function with no
858 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
859 This was a regression introduced in 4.83 by another bugfix.
861 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
863 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
864 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
870 TF/01 Correctly close the server side of TLS when forking for delivery.
872 When a message was received over SMTP with TLS, Exim failed to clear up
873 the incoming connection properly after forking off the child process to
874 deliver the message. In some situations the subsequent outgoing
875 delivery connection happened to have the same fd number as the incoming
876 connection previously had. Exim would try to use TLS and fail, logging
877 a "Bad file descriptor" error.
879 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
880 utilities have not been installed.
882 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
883 temporary space as the ACL may create new global variables.
885 TL/01 LDAP support uses per connection or global context settings, depending
886 upon the detected version of the libraries at build time.
888 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
889 to extract and use the src ip:port in logging and expansions as if it
890 were a direct connection from the outside internet. PPv2 support was
891 updated based on HAProxy spec change in May 2014.
893 JH/02 Add ${listextract {number}{list}{success}{fail}}.
895 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
896 Properly escape header and check for NULL return.
898 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
901 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
903 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
904 characters in header names, implemented as a verify condition.
905 Contributed by Michael Fischer v. Mollard.
907 TL/05 Rename SPF condition results err_perm and err_temp to standardized
908 results permerror and temperror. Previous values are deprecated but
909 still accepted. In a future release, err_perm and err_temp will be
910 completely removed, which will be a backward incompatibility if the
911 ACL tests for either of these two old results. Patch contributed by
912 user bes-internal on the mailing list.
914 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
916 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
917 selectors, in both main and reject logs.
919 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
922 JH/07 Add malware type "sock" for talking to simple daemon.
924 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
926 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
927 routers/transports under cutthrough routing.
929 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
930 numbers. Touch up "bool" conditional to keep the same definition.
932 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
934 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
936 JH/12 Expand items in router/transport headers_add or headers_remove lists
937 individually rather than the list as a whole. Bug 1452.
939 Required for reasonable handling of multiple headers_ options when
940 they may be empty; requires that headers_remove items with embedded
941 colons must have them doubled (or the list-separator changed).
943 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
944 view the policy declared in the DMARC record. Currently, $dmarc_status
945 is a combined value of both the record presence and the result of the
948 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
950 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
951 dnslookup router and the smtp transport (applying to the forward
954 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
955 of ldap servers used for a specific lookup. Patch provided by Heiko
958 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
959 New variable $lookup_dnssec_authenticated for observability.
961 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
962 Patch submitted by Lars Timman.
964 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
966 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
967 Requires trusted mode and valid format message id, aborts otherwise.
968 Patch contributed by Heiko Schlichting.
970 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
971 certextract with support for various fields. Bug 1358.
973 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
974 is requested by default, modifiable by smtp transport option
977 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
978 operate on certificate variables to give certificate fingerprints
979 Also new ${sha256:cert_variable}.
981 JH/23 The PRDR feature is moved from being Experimental into the mainline.
983 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
984 Christian Aistleitner.
986 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
988 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
989 file. Patch from Wolfgang Breyha.
991 JH/25 Expand the coverage of the delivery $host and $host_address to
992 client authenticators run in verify callout. Bug 1476.
994 JH/26 Port service names are now accepted for tls_on_connect_ports, to
995 align with daemon_smtp_ports. Bug 72.
997 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
998 support and error reporting did not work properly.
1000 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1001 and is readable. Patch from Andrew Colin Kissa.
1003 TL/14 Enhance documentation of ${run expansion and how it parses the
1004 commandline after expansion, particularly in the case when an
1005 unquoted variable expansion results in an empty value.
1007 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1009 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1010 by renaming away from it. Addresses GH issue 15, reported by
1013 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1015 TL/15 SECURITY: prevent double expansion in math comparison functions
1016 (can expand unsanitized data). Not remotely exploitable.
1023 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1025 PP/02 Make -n do something, by making it not do something.
1026 When combined with -bP, the name of an option is not output.
1028 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1031 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1032 $sender_host_name and config options to manage this, and basic check
1035 PP/05 DSCP support for outbound connections and control modifier for inbound.
1037 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1038 (Only plugin which currently uses this is kerberos4, which nobody should
1039 be using, but we should make it available and other future plugins might
1040 conceivably use it, even though it would break NAT; stuff *should* be
1041 using channel bindings instead).
1043 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1044 name; added for Sendmail compatibility; requires admin caller.
1045 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1046 just ignore it); requires trusted caller.
1047 Also parse but ignore: -Ac -Am -X<logfile>
1050 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1052 TL/02 Add +smtp_confirmation as a default logging option.
1054 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1055 Patch by Magnus Holmgren from 2007-02-20.
1057 TL/04 Bugzilla 1281 - Spec typo.
1058 Bugzilla 1283 - Spec typo.
1059 Bugzilla 1290 - Spec grammar fixes.
1061 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1063 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1065 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1066 mailing list by by Dmitry Isaikin.
1068 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1070 JH/02 Support "G" suffix to numbers in ${if comparisons.
1072 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1074 NM/01 Bugzilla 1197 - Spec typo
1075 Bugzilla 1196 - Spec examples corrections
1077 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1079 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1080 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1083 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1084 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1086 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1087 "acl {{name}{arg}...}", and optional args on acl condition
1090 JH/05 Permit multiple router/transport headers_add/remove lines.
1092 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1094 JH/07 Avoid using a waiting database for a single-message-only transport.
1095 Performance patch from Paul Fisher. Bugzilla 1262.
1097 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1100 JH/09 Add $headers_added variable, with content from use of ACL modifier
1101 add_header (but not yet added to the message). Bugzilla 199.
1103 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1104 Pulled from Bugzilla 817 by Wolfgang Breyha.
1106 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1108 (nb: this is the same fix as in Exim 4.80.1)
1110 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1113 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1116 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1118 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1119 advertise SMTP AUTH mechanism to us, instead of a generic
1120 protocol violation error. Also, make Exim more robust to bad
1121 data from the Dovecot auth socket.
1123 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1125 When a queue runner is handling a message, Exim first routes the
1126 recipient addresses, during which it prunes them based on the retry
1127 hints database. After that it attempts to deliver the message to
1128 any remaining recipients. It then updates the hints database using
1131 So if a recipient address works intermittently, it can get repeatedly
1132 deferred at routing time. The retry hints record remains fresh so the
1133 address never reaches the final cutoff time.
1135 This is a fairly common occurrence when a user is bumping up against
1136 their storage quota. Exim had some logic in its local delivery code
1137 to deal with this. However it did not apply to per-recipient defers
1138 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1140 This change adds a proper retry rule check during routing so that the
1141 final cutoff time is checked against the message's age. We only do
1142 this check if there is an address retry record and there is not a
1143 domain retry record; this implies that previous attempts to handle
1144 the address had the retry_use_local_parts option turned on. We use
1145 this as an approximation for the destination being like a local
1146 delivery, as in LMTP.
1148 I suspect this new check makes the old local delivery cutoff check
1149 redundant, but I have not verified this so I left the code in place.
1151 TF/02 Correct gecos expansion when From: is a prefix of the username.
1153 Test 0254 submits a message to Exim with the header
1157 When I ran the test suite under the user fanf2, Exim expanded
1158 the header to contain my full name, whereas it should have added
1159 a Resent-Sender: header. It erroneously treats any prefix of the
1160 username as equal to the username.
1162 This change corrects that bug.
1164 GF/01 DCC debug and logging tidyup
1165 Error conditions log to paniclog rather than rejectlog.
1166 Debug lines prefixed by "DCC: " to remove any ambiguity.
1168 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1170 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1171 Bug spotted by Jeremy Harris; was flawed since initial commit.
1172 Would have resulted in OCSP responses post-SNI triggering an Exim
1173 NULL dereference and crash.
1175 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1177 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1178 Bug detection, analysis and fix by Samuel Thibault.
1179 Bugzilla 1331, Debian bug #698092.
1181 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1183 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1184 Server implementation by Todd Lyons, client by JH.
1185 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1186 config variable "prdr_enable" controls whether the server
1187 advertises the facility. If the client requests PRDR a new
1188 acl_data_smtp_prdr ACL is called once for each recipient, after
1189 the body content is received and before the acl_smtp_data ACL.
1190 The client is controlled by both of: a hosts_try_prdr option
1191 on the smtp transport, and the server advertisement.
1192 Default client logging of deliveries and rejections involving
1193 PRDR are flagged with the string "PRDR".
1195 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1196 fclose(). Diagnosis by Todd Lyons.
1198 PP/17 Update configure.default to handle IPv6 localhost better.
1199 Patch by Alain Williams (plus minor tweaks).
1202 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1203 This is now consistent with GnuTLS, and is now documented: the
1204 previous undocumented portable approach to treating the option as
1205 unset was to force an expansion failure. That still works, and
1206 an empty string is now equivalent.
1208 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1209 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1210 not performing validation itself.
1212 PP/20 Added force_command boolean option to pipe transport.
1213 Patch from Nick Koston, of cPanel Inc.
1215 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1218 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1220 PP/21 Fix eximon continuous updating with timestamped log-files.
1221 Broken in a format-string cleanup in 4.80, missed when I repaired the
1222 other false fix of the same issue.
1223 Report and fix from Heiko Schlichting.
1226 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1227 Report from Prashanth Katuri.
1229 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1230 It's SecureTransport, so affects any MacOS clients which use the
1231 system-integrated TLS libraries, including email clients.
1233 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1234 using a MIME ACL for non-SMTP local injection.
1235 Report and assistance in diagnosis by Warren Baker.
1237 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1239 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1241 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1242 last id that failed so it may be referenced in subsequent ACL's.
1244 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1247 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1248 ldap library initialization, allowing self-signed CA's to be
1249 used. Also properly sets require_cert option later in code by
1250 using NULL (global ldap config) instead of ldap handle (per
1251 session). Bug diagnosis and testing by alxgomz.
1253 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1254 the src/util/ subdirectory.
1256 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1257 renamed to Transport Post Delivery Action by Jeremy Harris, as
1260 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1261 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1262 redis_servers = needs to be configured which will be used by the redis
1263 lookup. Patch from Warren Baker, of The Packet Hub.
1265 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1267 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1268 hostname or reverse DNS when processing a host list. Used suggestions
1269 from multiple comments on this bug.
1271 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1273 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1274 Missed a few lines, added it to make the runtest require no keyboard
1277 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1278 contains upper case chars. Make router use caseful_local_part.
1280 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1281 support when GnuTLS has been built with p11-kit.
1287 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1289 This, or similar/improved, will also be change PP/11 of 4.82.
1295 PP/01 Handle short writes when writing local log-files.
1296 In practice, only affects FreeBSD (8 onwards).
1297 Bugzilla 1053, with thanks to Dmitry Isaikin.
1299 NM/01 Bugzilla 949 - Documentation tweak
1301 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1304 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1306 PP/02 Implemented gsasl authenticator.
1308 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1310 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1311 `pkg-config foo` for cflags/libs.
1313 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1314 with rest of GSASL and with heimdal_gssapi.
1316 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1317 `pkg-config foo` for cflags/libs for the TLS implementation.
1319 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1320 properties get this fed in as external SSF. A number of robustness
1321 and debugging improvements to the cyrus_sasl authenticator.
1323 PP/08 cyrus_sasl server now expands the server_realm option.
1325 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1326 Patch by Jeremy Harris.
1328 PP/10 Added dbmjz lookup type.
1330 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1332 PP/12 MAIL args handles TAB as well as SP, for better interop with
1333 non-compliant senders.
1334 Analysis and variant patch by Todd Lyons.
1336 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1337 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1338 Patch from Dirk Mueller <dmueller@suse.com>
1340 PP/13 tls_peerdn now print-escaped for spool files.
1341 Observed some $tls_peerdn in wild which contained \n, which resulted
1342 in spool file corruption.
1344 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1345 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1346 or write after TLS renegotiation, which otherwise led to messages
1349 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1350 as a tracking header (ie: a signed header comes before the signature).
1351 Patch from Wolfgang Breyha.
1353 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1354 comma-sep list; embedded commas doubled.
1356 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1358 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1360 Report and patch from Dmitry Banschikov.
1362 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1363 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1364 failures. We appear to now support TLS1.1+ with Exim.
1366 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1367 lets Exim select keys and certificates based upon TLS SNI from client.
1368 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1369 before an outbound SMTP session. New log_selector, +tls_sni.
1371 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1372 NULL dereference. Report and patch from Alun Jones.
1374 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1375 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1376 Not seeing resolver debug output on NetBSD, but suspect this is a
1377 resolver implementation change.
1379 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1380 Left warnings. Added "eximon gdb" invocation mode.
1382 PP/21 Defaulting "accept_8bitmime" to true, not false.
1384 PP/22 Added -bw for inetd wait mode support.
1386 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1387 locate the relevant includes and libraries. Made this the default.
1389 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1390 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1392 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1393 This may cause build issues on older platforms.
1395 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1396 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1397 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1398 Added SNI support via GnuTLS too.
1399 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1401 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1403 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1404 Applied second patch from Janne, implementing suggestion to default
1405 multiple-strings-in-record handling to match SPF spec.
1407 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1409 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1410 read-only, out of scope).
1411 Patch from Wolfgang Breyha, report from Stuart Northfield.
1413 PP/29 Fix three issues highlighted by clang analyser static analysis.
1414 Only crash-plausible issue would require the Cambridge-specific
1415 iplookup router and a misconfiguration.
1416 Report from Marcin Mirosław.
1418 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1420 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1421 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1422 As part of this, removing so much warning spew let me fix some minor
1423 real issues in debug logging.
1425 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1426 assignment on my part. Fixed.
1428 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1429 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1430 Janne Snabb (who went above and beyond: thank you).
1432 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1433 string otherwise requires a connection and a bunch more work and it's
1434 relatively easy to get wrong. Should also expose TLS library linkage
1437 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1438 64-bit ${eval} (JH/03).
1440 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1441 GNU libc to support some of the 64-bit stuff, should not lead to
1442 conflicts. Defined before os.h is pulled in, so if a given platform
1443 needs to override this, it can.
1445 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1446 protection layer was required, which is not implemented.
1447 Bugzilla 1254, patch from Wolfgang Breyha.
1449 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1450 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1451 tls_dhparam take prime identifiers. Also unbreak combination of
1452 OpenSSL+DH_params+TLSSNI.
1454 PP/39 Disable SSLv2 by default in OpenSSL support.
1460 PP/01 Solaris build fix for Oracle's LDAP libraries.
1461 Bugzilla 1109, patch from Stephen Usher.
1463 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1465 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1468 TF/02 Fix a couple more cases where we did not log the error message
1469 when unlink() failed. See also change 4.74-TF/03.
1471 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1472 lock up or crash if it happened to be inside a call to libc when it
1473 got a SIGUSR1 from exiwhat.
1475 The SIGUSR1 handler appends the current process status to the process
1476 log which is later printed by exiwhat. It used to use the general
1477 purpose logging code to do this, but several functions it calls are
1478 not safe for signals.
1480 The new output code in the SIGUSR1 handler is specific to the process
1481 log, and simple enough that it's easy to inspect for signal safety.
1482 Removing some special cases also simplifies the general logging code.
1483 Removing the spurious timestamps from the process log simplifies
1486 TF/04 Improved ratelimit ACL condition.
1488 The /noupdate option has been deprecated in favour of /readonly which
1489 has clearer semantics. The /leaky, /strict, and /readonly update modes
1490 are mutually exclusive. The update mode is no longer included in the
1491 database key; it just determines when the database is updated. (This
1492 means that when you upgrade Exim will forget old rate measurements.)
1494 Exim now checks that the per_* options are used with an update mode that
1495 makes sense for the current ACL. For example, when Exim is processing a
1496 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1497 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1498 must specify per_mail/readonly. If you omit the update mode it defaults to
1499 /leaky where that makes sense (as before) or /readonly where required.
1501 The /noupdate option is now undocumented but still supported for
1502 backwards compatibility. It is equivalent to /readonly except that in
1503 ACLs where /readonly is required you may specify /leaky/noupdate or
1504 /strict/noupdate which are treated the same as /readonly.
1506 A useful new feature is the /count= option. This is a generalization
1507 of the per_byte option, so that you can measure the throughput of other
1508 aggregate values. For example, the per_byte option is now equivalent
1509 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1511 The per_rcpt option has been generalized using the /count= mechanism
1512 (though it's more complicated than the per_byte equivalence). When it is
1513 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1514 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1515 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1516 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1517 Note that using per_rcpt with a non-readonly update mode in more than
1518 one ACL will cause the recipients to be double-counted. (The per_mail
1519 and per_byte options don't have this problem.)
1521 The handling of very low rates has changed slightly. If the computed rate
1522 is less than the event's count (usually one) then this event is the first
1523 after a long gap. In this case the rate is set to the same as this event's
1524 count, so that the first message of a spam run is counted properly.
1526 The major new feature is a mechanism for counting the rate of unique
1527 events. The new per_addr option counts the number of different
1528 recipients that someone has sent messages to in the last time period. It
1529 behaves like per_rcpt if all the recipient addresses are different, but
1530 duplicate recipient addresses do not increase the measured rate. Like
1531 the /count= option this is a general mechanism, so the per_addr option
1532 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1533 example, measure the rate that a client uses different sender addresses
1534 with the options per_mail/unique=$sender_address. There are further
1535 details in the main documentation.
1537 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1539 TF/06 Removed a few PCRE remnants.
1541 TF/07 Automatically extract Exim's version number from tags in the git
1542 repository when doing development or release builds.
1544 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1545 Bugzilla 879. Patch from Paul Fisher.
1547 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1548 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1551 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1553 PP/05 Variable $av_failed, true if the AV scanner deferred.
1554 Bugzilla 1078. Patch from John Horne.
1556 PP/06 Stop make process more reliably on build failure.
1557 Bugzilla 1087. Patch from Heiko Schlittermann.
1559 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1560 Bugzilla 1089. Patch from Heiko Schlittermann.
1562 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1563 Bugzilla 1131. Patch from Holger Weiß.
1565 PP/09 Handle IPv6 addresses with SPF.
1566 Bugzilla 860. Patch from Wolfgang Breyha.
1568 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1570 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1573 PP/11 match_* no longer expand right-hand-side by default.
1574 New compile-time build option, EXPAND_LISTMATCH_RHS.
1575 New expansion conditions, "inlist", "inlisti".
1577 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1579 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1581 PP/14 fix log_write() format string regression from TF/03.
1582 Bugzilla 1152. Patch from Dmitry Isaikin.
1588 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1590 PP/02 Harmonised TLS library version reporting; only show if debugging.
1591 Layout now matches that introduced for other libraries in 4.74 PP/03.
1593 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1595 PP/04 New "dns_use_edns0" global option.
1597 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1600 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1601 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1603 TK/01 Updated PolarSSL code to 0.14.2.
1604 Bugzilla 1097. Patch from Andreas Metzler.
1606 PP/07 Catch divide-by-zero in ${eval:...}.
1607 Fixes bugzilla 1102.
1609 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1612 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1613 format-string attack -- SECURITY: remote arbitrary code execution.
1615 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1616 time unintentionally subject to list matching rules, letting the header
1617 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1618 arbitrary string expansion). This allowed for information disclosure.
1620 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1621 INT_MIN/-1 -- value coerced to INT_MAX.
1627 NM/01 Workaround for PCRE version dependency in version reporting
1630 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1631 This fixes portability to compilers other than gcc, notably
1632 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1634 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1635 makefiles for portability to HP-UX and POSIX correctness.
1637 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1638 Also via indented variable definition in the Makefile.
1639 (Debugging by Oliver Heesakkers).
1641 PP/02 Restore caching of spamd results with expanded spamd_address.
1642 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1644 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1645 Improves build reliability. Fix from: Frank Elsner
1647 NM/02 Fix wide character breakage in the rfc2047 coding
1648 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1650 NM/03 Allow underscore in dnslist lookups
1651 Fixes bug 1026. Patch from Graeme Fowler
1653 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1654 Code patches from Adam Ciarcinski of NetBSD.
1656 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1659 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1660 is logged, to avoid truncation. Patch from John Horne.
1662 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1663 Patch from Jakob Hirsch.
1665 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1666 SQL string expansion failure details.
1667 Patch from Andrey Oktyabrski.
1669 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1670 Patch from Simon Arlott.
1672 PP/09 New lookups functionality failed to compile on old gcc which rejects
1673 extern declarations in function scope.
1674 Patch from Oliver Fleischmann
1676 PP/10 Use sig_atomic_t for flags set from signal handlers.
1677 Check getgroups() return and improve debugging.
1678 Fixed developed for diagnosis in bug 927 (which turned out to be
1681 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1682 Patch from Mark Zealey.
1684 PP/12 Bugzilla 1056: Improved spamd server selection.
1685 Patch from Mark Zealey.
1687 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1688 Based on patch from Heiko Schlittermann.
1690 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1691 Patch from Uwe Doering, sign-off by Michael Haardt.
1693 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1694 variable declaration deep within a block. Bug and patch from
1697 PP/15 lookups-Makefile IRIX compatibility coercion.
1699 PP/16 Make DISABLE_DKIM build knob functional.
1701 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1702 Patch by Simon Arlott
1704 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1705 variable argument macros. Our copy now differs from upstream.
1711 TF/01 Failure to get a lock on a hints database can have serious
1712 consequences so log it to the panic log.
1714 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1715 controlled using the smtp_confirmation log selector.
1717 TF/03 Include the error message when we fail to unlink a spool file.
1719 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1720 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1721 for maintaining out-of-tree patches for some time.
1723 PP/01 Bugzilla 139: Documentation and portability issues.
1724 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1725 Handle per-OS dynamic-module compilation flags.
1727 PP/02 Let /dev/null have normal permissions.
1728 The 4.73 fixes were a little too stringent and complained about the
1729 permissions on /dev/null. Exempt it from some checks.
1730 Reported by Andreas M. Kirchwitz.
1732 PP/03 Report version information for many libraries, including
1733 Exim version information for dynamically loaded libraries. Created
1734 version.h, now support a version extension string for distributors
1735 who patch heavily. Dynamic module ABI change.
1737 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1738 privilege escalation vulnerability whereby the Exim run-time user
1739 can cause root to append content of the attacker's choosing to
1742 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1745 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1746 If dropping privileges for untrusted macros, we disabled normal logging
1747 on the basis that it would fail; for the Exim run-time user, this is not
1748 the case, and it resulted in successful deliveries going unlogged.
1749 Fixed. Reported by Andreas Metzler.
1755 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1756 only prepend for the Resent-* case. Fixes regression introduced in
1757 Exim 4.70 by NM/22 for Bugzilla 607.
1759 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1760 increasing numbers of administrators be bitten by this.
1762 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1764 PP/03 Bugzilla 994: added openssl_options main configuration option.
1766 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1768 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1770 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1772 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1773 without a peer certificate, leading to a segfault because of an
1774 assumption that peers always have certificates. Be a little more
1775 paranoid. Problem reported by Martin Tscholak.
1777 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1778 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1779 NB: ClamAV planning to remove STREAM in "middle of 2010".
1780 CL also introduces -bmalware, various -d+acl logging additions and
1781 more caution in buffer sizes.
1783 PP/09 Implemented reverse_ip expansion operator.
1785 PP/10 Bugzilla 937: provide a "debug" ACL control.
1787 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1789 PP/12 Bugzilla 973: Implement --version.
1791 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1793 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1795 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1797 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1798 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1799 ignore trailing whitespace.
1801 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1803 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1806 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1807 Notification from Dr Andrew Aitchison.
1809 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1810 ExtendedDetectionInfo response format.
1811 Notification from John Horne.
1813 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1816 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1817 XSL and documented dependency on system catalogs, with examples of how
1820 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1823 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1824 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1825 configuration file which is writeable by the Exim user or group.
1827 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1828 of configuration files to cover files specified with the -C option if
1829 they are going to be used with root privileges, not just the default
1832 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1833 option (effectively making it always true).
1835 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1836 files to be used while preserving root privileges.
1838 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1839 that rogue child processes cannot use them.
1841 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1842 run-time user, instead of root.
1844 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1845 Exim run-time user without dropping privileges.
1847 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1848 result string, instead of calling string_vformat() twice with the same
1851 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1852 for other users. Others should always drop root privileges if they use
1853 -C on the command line, even for a whitelisted configure file.
1855 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1857 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1863 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1864 $data_path, and $header_path variables; fixed documentation bugs and
1867 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1868 exipick to access non-standard spools, including the "frozen" queue
1871 NM/01 Bugzilla 965: Support mysql stored procedures.
1872 Patch from Alain Williams
1874 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1876 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1877 Patch from Andreas Metzler
1879 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1880 Patch from Kirill Miazine
1882 NM/05 Bugzilla 671: Added umask to procmail example.
1884 JJ/03 installed exipick 20100323.0, fixing doc bug
1886 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1887 directory. Notification and patch from Dan Rosenberg.
1889 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1891 TK/02 Improve log output when DKIM signing operation fails.
1893 MH/01 Treat the transport option dkim_domain as a colon separated
1894 list, not as a single string, and sign the message with each element,
1895 omitting multiple occurences of the same signer.
1897 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1898 Bugzilla 985, 986. Patch by Simon Arlott
1900 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1901 Patch by Simon Arlott
1903 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1904 MBX locking. Notification from Dan Rosenberg.
1910 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1912 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1914 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1916 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1918 NM/04 Bugzilla 915: Flag broken perl installation during build.
1924 TK/01 Added patch by Johannes Berg that expands the main option
1925 "spamd_address" if it starts with a dollar sign.
1927 TK/02 Write list of recipients to X-Envelope-Sender header when building
1928 the mbox-format spool file for content scanning (suggested by Jakob
1931 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1932 (http://www.dcc-servers.net/) support via dccifd. Activated by
1933 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1935 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1936 by Mark Daniel Reidel <mr@df.eu>.
1938 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1939 When building exim an external PCRE library is now needed -
1940 PCRE is a system library on the majority of modern systems.
1941 See entry on PCRE_LIBS in EDITME file.
1943 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1944 conversation. Added nologin parameter to request.
1945 Patch contributed by Kirill Miazine.
1947 TF/01 Do not log submission mode rewrites if they do not change the address.
1949 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1951 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1952 log files in place. Contributed by Roberto Lima.
1954 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1956 TF/03 Bugzilla 615: When checking the local_parts router precondition
1957 after a local_part_suffix or local_part_prefix option, Exim now
1958 does not use the address's named list lookup cache, since this
1959 contains cached lookups for the whole local part.
1961 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1962 Robert Millan. Documentation is in experimental-spec.txt.
1964 TF/04 Bugzilla 668: Fix parallel build (make -j).
1966 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1968 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1969 Patch provided by Jan Srzednicki.
1971 TF/05 Leading white space used to be stripped from $spam_report which
1972 wrecked the formatting. Now it is preserved.
1974 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1975 that they are available at delivery time.
1977 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1979 TF/08 TLS error reporting now respects the incoming_interface and
1980 incoming_port log selectors.
1982 TF/09 Produce a more useful error message if an SMTP transport's hosts
1983 setting expands to an empty string.
1985 NM/06 Bugzilla 744: EXPN did not work under TLS.
1986 Patch provided by Phil Pennock.
1988 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1989 Patch provided by Richard Godbee.
1991 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1992 acl_smtp_notquit, added index entry.
1994 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1995 Patch provided by Eugene Bujak.
1997 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1998 accept(). Patch provided by Maxim Dounin.
2000 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2001 Patch provided by Phil Pennock.
2003 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2005 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2006 Patch provided by Brad "anomie" Jorsch.
2008 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2009 Patch provided by Dean Brooks.
2011 TK/05 Add native DKIM support (does not depend on external libraries).
2013 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2014 Patch provided by Graeme Fowler.
2016 NM/16 Bugzilla 851: Documentation example syntax fix.
2018 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2020 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2023 NM/19 Bugzilla 745: TLS version reporting.
2024 Patch provided by Phil Pennock.
2026 NM/20 Bugzilla 167: bool: condition support.
2027 Patch provided by Phil Pennock.
2029 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2030 clients. Patch provided by Phil Pennock.
2032 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2033 Patch provided by Brad "anomie" Jorsch.
2035 NM/23 Bugzilla 687: Fix misparses in eximstats.
2036 Patch provided by Heiko Schlittermann.
2038 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2039 Patch provided by Heiko Schlittermann.
2041 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2042 plus update to original patch.
2044 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2046 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2047 Patch provided by David Brownlee.
2049 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2051 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2053 NM/30 Bugzilla 888: TLS documentation bugfixes.
2055 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2057 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2058 Unlike the original bugzilla I have changed all shell scripts in src tree.
2060 NM/33 Bugzilla 898: Transport filter timeout fix.
2061 Patch by Todd Rinaldo.
2063 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2064 Patch by Serge Demonchaux.
2066 NM/35 Bugzilla 39: Base64 decode bug fixes.
2067 Patch by Jakob Hirsch.
2069 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2071 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2073 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2075 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2081 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2082 ALT-N's libdkim that I have put here:
2083 http://duncanthrax.net/exim-experimental/
2085 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2086 were called 'true' and it seems that C99 defines that as a reserved
2087 keyword to be used with 'bool' variable types. That means you could
2088 not include C99-style headers which use bools without triggering
2089 build errors in sieve.c.
2091 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2092 as mailq or other aliases. Changed the --help handling significantly
2093 to do whats expected. exim_usage() emits usage/help information.
2095 SC/01 Added the -bylocaldomain option to eximstats.
2097 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2099 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2101 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2107 PH/01 Another patch from the Sieve maintainer.
2109 PH/02 When an IPv6 address is converted to a string for single-key lookup
2110 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2111 dots are used instead of colons so that keys in lsearch files need not
2112 contain colons. This was done some time before quoting was made available
2113 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2114 (notated using the quote facility) so as to distinguish them from IPv4
2115 keys. This meant that lookups for IP addresses in host lists did not work
2116 for iplsearch lookups.
2118 This has been fixed by arranging for IPv6 addresses to be expressed with
2119 colons if the lookup type is iplsearch. This is not incompatible, because
2120 previously such lookups could never work.
2122 The situation is now rather anomalous, since one *can* have colons in
2123 ordinary lsearch keys. However, making the change in all cases is
2124 incompatible and would probably break a number of configurations.
2126 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2129 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2130 conversion specification without a maximum field width, thereby enabling
2131 a rogue spamd server to cause a buffer overflow. While nobody in their
2132 right mind would setup Exim to query an untrusted spamd server, an
2133 attacker that gains access to a server running spamd could potentially
2134 exploit this vulnerability to run arbitrary code as the Exim user.
2136 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2137 $primary_hostname instead of what libspf2 thinks the hosts name is.
2139 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2140 a directory entry by the name of the lookup key. Previously, if a
2141 symlink pointed to a non-existing file or a file in a directory that
2142 Exim lacked permissions to read, a lookup for a key matching that
2143 symlink would fail. Now it is enough that a matching directory entry
2144 exists, symlink or not. (Bugzilla 503.)
2146 PH/03 The body_linecount and body_zerocount variables are now exported in the
2149 PH/04 Added the $dnslist_matched variable.
2151 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2152 This means they are set thereafter only if the connection becomes
2155 PH/06 Added the client_condition to authenticators so that some can be skipped
2156 by clients under certain conditions.
2158 PH/07 The error message for a badly-placed control=no_multiline_responses left
2159 "_responses" off the end of the name.
2161 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2163 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2164 (without spaces) instead of just copying the configuration text.
2166 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2168 PH/11 Added $max_received_linelength.
2170 PH/12 Added +ignore_defer and +include_defer to host lists.
2172 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2173 way in which PCRE > 7.0 is built.
2175 PH/14 Implemented queue_only_load_latch.
2177 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2178 MAIL command. The effect was to mangle the value on 64-bit systems.
2180 PH/16 Another patch from the Sieve maintainer.
2182 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2184 PH/18 If a system quota error occurred while trying to create the file for
2185 a maildir delivery, the message "Mailbox is full" was not appended to the
2186 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2187 only to a quota excession during the actual writing of the file.
2189 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2190 characters?) which causes problems in log lines. The DN values are now
2191 passed through string_printing() before being added to log lines.
2193 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2194 and InterBase are left for another time.)
2196 PH/21 Added message_body_newlines option.
2198 PH/22 Guard against possible overflow in moan_check_errorcopy().
2200 PH/23 POSIX allows open() to be a macro; guard against that.
2202 PH/24 If the recipient of an error message contained an @ in the local part
2203 (suitably quoted, of course), incorrect values were put in $domain and
2204 $local_part during the evaluation of errors_copy.
2210 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2211 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2214 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2215 issue a MAIL command.
2217 PH/02 In an ACL statement such as
2219 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2221 if a client was not listed at all, or was listed with a value other than
2222 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2223 the condition was not true (as it should be), so access was not denied.
2224 The bug was that the ! inversion was incorrectly passed on to the second
2225 item. This has been fixed.
2227 PH/03 Added additional dnslists conditions == and =& which are different from
2228 = and & when the dns lookup returns more than one IP address.
2230 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2231 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2233 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2234 FSYNC, which compiles an option called disable_fsync that allows for
2235 bypassing fsync(). The documentation is heavily laced with warnings.
2237 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2239 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2240 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2241 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2242 including adding "make clean"; (3) Added -fPIC when compiling the test
2243 dynamically loaded module, to get rid of a warning.
2245 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2246 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2247 = 0s. The bug is otherwise harmless.
2249 PH/07 There was a bug in the dovecot authenticator such that the value of
2250 $auth1 could be overwritten, and so not correctly preserved, after a
2251 successful authentication. This usually meant that the value preserved by
2252 the server_setid option was incorrect.
2254 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2256 PH/09 Installed PCRE release 7.0.
2258 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2259 run for batched SMTP input. It is now run at the start of every message
2260 in the batch. While fixing this I discovered that the process information
2261 (output by running exiwhat) was not always getting set for -bs and -bS
2262 input. This is fixed, and it now also says "batched" for BSMTP.
2264 PH/11 Added control=no_pipelining.
2266 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2267 patch, slightly modified), and move the expansion of helo_data till after
2268 the connection is made in the smtp transport (so it can use these
2271 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2273 PH/14 Added log_selector = +pid.
2275 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2277 PH/16 Add ${if forany and ${if forall.
2279 PH/17 Added dsn_from option to vary the From: line in DSNs.
2281 PH/18 Flush SMTP output before performing a callout, unless control =
2282 no_callout_flush is set.
2284 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2285 was true (the default) a successful delivery failed to delete the retry
2286 item, thus causing premature timeout of the address. The bug is now
2289 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2291 PH/21 Long custom messages for fakedefer and fakereject are now split up
2292 into multiline responses in the same way that messages for "deny" and
2293 other ACL rejections are.
2295 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2296 with slight modification.
2298 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2299 draft, changing the syntax and factoring some duplicate code".
2301 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2302 for deliveries of the second and subsequent messages over the same SMTP
2305 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2306 ${reduce, with only minor "tidies".
2308 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2310 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2311 expansion side effects.
2313 PH/27 When a message times out after an over-quota error from an Exim-imposed
2314 quota, the bounce message says "mailbox is full". This message was not
2315 being given when it was a system quota that was exceeded. It now should
2318 MH/03 Made $recipients available in local_scan(). local_scan() already has
2319 better access to the recipient list through recipients_list[], but
2320 $recipients can be useful in postmaster-provided expansion strings.
2322 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2323 in the case of a MAIL command with additional options following the
2324 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2325 were accidentally chopped off.
2327 PH/29 SMTP synchronization checks are implemented when a command is read -
2328 there is a check that no more input is waiting when there shouldn't be
2329 any. However, for some commands, a delay in an ACL can mean that it is
2330 some time before the response is written. In this time, more input might
2331 arrive, invalidly. So now there are extra checks after an ACL has run for
2332 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2333 pipelining has not been advertised.
2335 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2337 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2338 returned in the SMTP response when smtp_return_error_details was false.
2339 This has been fixed.
2341 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2342 instead of the C I/O that was originally supplied, because problems were
2343 reported on Solaris.
2345 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2346 Exim which did not show up earlier: it was assuming that a call to
2347 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2348 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2349 was a macro that became an assignment, so it seemed to work. This has
2350 changed to a proper function call with a void return, hence the compile
2351 error. Exim's code has been fixed.
2353 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2356 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2358 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2360 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2361 because it turns out to be too big or there is a timeout) there is no
2362 "Completed" line in the log. When some messages of this type were
2363 selected by exigrep, they were listed as "not completed". Others were
2364 picked up by some special patterns. I have improved the selection
2365 criteria to be more general.
2367 PH/38 The host_find_failed option in the manualroute router can now be set
2368 to "ignore", to completely ignore a host whose IP address cannot be
2369 found. If all hosts are ignored, the behaviour is controlled by the new
2370 host_all_ignored option.
2372 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2373 homing or because of multiple MX records with /mx) generated more than
2374 one IP address, and the following item turned out to be the local host,
2375 all the secondary addresses of the first item were incorrectly removed
2376 from the list, along with the local host and any following hosts (which
2377 is what is supposed to happen).
2379 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2380 whoever called Exim into the -H file. In the case of the daemon it was
2381 behaving confusingly. When first started, it used values for whoever
2382 started the daemon, but after a SIGHUP it used the Exim user (because it
2383 calls itself on a restart). I have changed the code so that it now always
2386 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2387 message are rejected with the same error (e.g. no authentication or bad
2388 sender address), and a DATA command is nevertheless sent (as can happen
2389 with PIPELINING or a stupid MUA), the error message that was given to the
2390 RCPT commands is included in the rejection of the DATA command. This is
2391 intended to be helpful for MUAs that show only the final error to their
2394 PH/42 Another patch from the Sieve maintainer.
2396 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2397 Eximstats - Fixed some broken HTML links and added missing column headers
2399 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2400 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2402 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2404 PH/43 Yet another patch from the Sieve maintainer.
2406 PH/44 I found a way to check for a TCP/IP connection going away before sending
2407 the response to the final '.' that terminates a message, but only in the
2408 case where the client has not sent further data following the '.'
2409 (unfortunately, this is allowed). However, in many cases there won't be
2410 any further data because there won't be any more messages to send. A call
2411 to select() can be used: if it shows that the input is "ready", there is
2412 either input waiting, or the socket has been closed. An attempt to read
2413 the next input character can distinguish the two cases. Previously, Exim
2414 would have sent an OK response which the client would never have see.
2415 This could lead to message repetition. This fix should cure that, at
2416 least in a lot of common cases.
2418 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2419 advertised in response to EHLO.
2425 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2426 fixed by 4.65/MH/01 (is this a record?) are fixed:
2428 (i) An empty string was always treated as zero by the numeric comparison
2429 operators. This behaviour has been restored.
2431 (ii) It is documented that the numeric comparison operators always treat
2432 their arguments as decimal numbers. This was broken in that numbers
2433 starting with 0 were being interpreted as octal.
2435 While fixing these problems I realized that there was another issue that
2436 hadn't been noticed. Values of message_size_limit (both the global option
2437 and the transport option) were treated as octal if they started with 0.
2438 The documentation was vague. These values are now always treated as
2439 decimal, and I will make that clear in the documentation.
2445 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2446 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2449 MH/01 Don't check that the operands of numeric comparison operators are
2450 integers when their expansion is in "skipping" mode (fixes bug
2451 introduced by 4.64-PH/07).
2453 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2454 child addresses, Exim now panics and dies. Previously, because the count
2455 is held in a short int, deliveries were likely to be lost. As such a
2456 large number of recipients for a single message is ridiculous
2457 (performance will be very, very poor), I have chosen to impose a limit
2458 rather than extend the field.
2464 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2465 leftover -K file (the existence of which was triggered by #402).
2466 While we were at it, introduced process PID as part of the -K
2467 filename. This should rule out race conditions when creating
2470 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2471 processing considerably. Previous code took too long for large mails,
2472 triggering a timeout which in turn triggers #401.
2474 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2475 in the DK code in transports.c. sendfile() is not really portable,
2476 hence the _LINUX specificness.
2478 TF/01 In the add_headers option to the mail command in an Exim filter,
2479 there was a bug that Exim would claim a syntax error in any
2480 header after the first one which had an odd number of characters
2483 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2484 callout verification, Exim cached a "reject" for the entire domain. This
2485 is correct for most verifications, but it is not correct for a recipient
2486 verification with use_sender or use_postmaster set, because in that case
2487 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2488 case of MAIL FROM:<> rejection from other early rejections (e.g.
2489 rejection of HELO). When verifying a recipient using a non-null MAIL
2490 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2491 Whatever the result of the callout, the value of the domain cache is
2492 left unchanged (for any other kind of callout, getting as far as trying
2493 RCPT means that the domain itself is ok).
2495 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2498 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2499 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2500 session, the daemon ignores ECONNECT errors and logs others; it now
2501 ignores EPIPE as well.
2503 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2504 (quoted-printable decoding).
2506 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2507 later the small subsequent patch to fix an introduced bug.
2509 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2511 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2513 PH/08 An error is now given if message_size_limit is specified negative.
2515 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2516 to be given (somewhat) arbitrary names.
2518 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2521 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2522 miscellaneous code fixes
2524 PH/10 Added the log_reject_target ACL modifier to specify where to log
2527 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2528 hostname. This is wrong, because it relates to the incoming message (and
2529 probably the interface on which it is arriving) and not to the outgoing
2530 callout (which could be using a different interface). This has been
2531 changed to use the value of the helo_data option from the smtp transport
2532 instead - this is what is used when a message is actually being sent. If
2533 there is no remote transport (possible with a router that sets up host
2534 addresses), $smtp_active_hostname is used.
2536 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2537 tweaks were necessary in order to get it to work (see also 21 below):
2538 (a) The code assumed that strncpy() returns a negative number on buffer
2539 overflow, which isn't the case. Replaced with Exim's string_format()
2541 (b) There were several signed/unsigned issues. I just did the minimum
2542 hacking in of casts. There is scope for a larger refactoring.
2543 (c) The code used strcasecmp() which is not a standard C function.
2544 Replaced with Exim's strcmpic() function.
2545 (d) The code set only $1; it now sets $auth1 as well.
2546 (e) A simple test gave the error "authentication client didn't specify
2547 service in request". It would seem that Dovecot has changed its
2548 interface. Fortunately there's a specification; I followed it and
2549 changed what the client sends and it appears to be working now.
2551 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2554 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2555 address A is aliased to B and C, where B exists and C does not. Without
2556 -v the output is "A verified" because verification stops after a
2557 successful redirection if more than one address is generated. However,
2558 with -v the child addresses are also verified. Exim was outputting "A
2559 failed to verify" and then showing the successful verification for C,
2560 with its parentage. It now outputs "B failed to verify", showing B's
2561 parentage before showing the successful verification of C.
2563 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2564 look up a TXT record in a specific list after matching in a combined
2567 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2568 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2569 they consult the DNS. I had assumed they would set it the way they
2570 wanted; and indeed my experiments on Linux seem to show that in some
2571 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2572 To be on the safe side, however, I have now made the interface to
2573 host_find_byname() similar to host_find_bydns(), with an argument
2574 containing the DNS resolver options. The host_find_byname() function now
2575 sets these options at its start, just as host_find_bydns() does. The smtp
2576 transport options dns_qualify_single and dns_search_parents are passed to
2577 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2578 of host_find_byname() use the default settings of RES_DEFNAMES
2579 (qualify_single) but not RES_DNSRCH (search_parents).
2581 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2582 spool_read_header() do less string testing, by means of a preliminary
2583 switch on the second character of optional "-foo" lines. (This is
2584 overdue, caused by the large number of possibilities that now exist.
2585 Originally there were few.) While I was there, I also converted the
2586 str(n)cmp tests so they don't re-test the leading "-" and the first
2587 character, in the hope this might squeeze out yet more improvement.
2589 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2590 flag allowing group syntax was set by the header_syntax check but not
2591 turned off, possible causing trouble later; (2) The flag was not being
2592 set at all for the header_verify test, causing "group"-style headers to
2593 be rejected. I have now set it in this case, and also caused header_
2594 verify to ignore an empty address taken from a group. While doing this, I
2595 came across some other cases where the code for allowing group syntax
2596 while scanning a header line wasn't quite right (mostly, not resetting
2597 the flag correctly in the right place). These bugs could have caused
2598 trouble for malformed header lines. I hope it is now all correct.
2600 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2601 with the "reply" argument non-NULL. The code, however (which originally
2602 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2603 but it didn't always do it. This confused somebody who was copying the
2604 code for some other use. I have removed all the tests.
2606 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2607 feature that was used to support insecure browsers during the U.S. crypto
2608 embargo. It requires special client support, and Exim is probably the
2609 only MTA that supported it -- and would never use it because real RSA is
2610 always available. This code has been removed, because it had the bad
2611 effect of slowing Exim down by computing (never used) parameters for the
2612 RSA_EXPORT functionality.
2614 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2615 authenticator to fail if there's a tab character in the incoming data
2616 (there should never be unless someone is messing about, as it's supposed
2617 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2618 if the connection is using TLS or if the remote IP is the same as the
2619 local IP, and the "valid-client-cert option" if a client certificate has
2622 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2623 authenticators. This can be used for authorization after authentication
2624 succeeds. (In the case of plaintext, it servers for both authentication
2627 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2628 if any retry times were supplied.
2630 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2631 connection, where there is no client IP address to check. In this
2632 situation, the verify now always succeeds.
2634 PH/25 Applied John Jetmore's -Mset patch.
2636 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2638 PH/27 In a string expansion for a processed (not raw) header when multiple
2639 headers of the same name were present, leading whitespace was being
2640 removed from all of them, but trailing whitespace was being removed only
2641 from the last one. Now trailing whitespace is removed from each header
2642 before concatenation. Completely empty headers in a concatenation (as
2643 before) are ignored.
2645 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2646 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2648 PH/29 [Removed. This was a change that I later backed out, and forgot to
2649 correct the ChangeLog entry (that I had efficiently created) before
2650 committing the later change.]
2652 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2653 address errors (4xx response to RCPT) over the same connection as other
2654 messages routed to the same hosts. Such deliveries are always "forced",
2655 so retry times are not inspected. This resulted in far too many retries
2656 for the affected addresses. The effect occurred only when there were more
2657 hosts than the hosts_max_try setting in the smtp transport when it had
2658 the 4xx errors. Those hosts that it had tried were not added to the list
2659 of hosts for which the message was waiting, so if all were tried, there
2660 was no problem. Two fixes have been applied:
2662 (i) If there are any address or message errors in an SMTP delivery, none
2663 of the hosts (tried or untried) are now added to the list of hosts
2664 for which the message is waiting, so the message should not be a
2665 candidate for sending over the same connection that was used for a
2666 successful delivery of some other message. This seems entirely
2667 reasonable: after all the message is NOT "waiting for some host".
2668 This is so "obvious" that I'm not sure why it wasn't done
2669 previously. Hope I haven't missed anything, but it can't do any
2670 harm, as the worst effect is to miss an optimization.
2672 (ii) If, despite (i), such a delivery is accidentally attempted, the
2673 routing retry time is respected, so at least it doesn't keep
2674 hammering the server.
2676 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2677 in ${readsocket because some servers need this prod.
2679 PH/32 Added some extra debug output when updating a wait-xxx database.
2681 PH/33 The hint "could be header name not terminated by colon", which has been
2682 given for certain expansion errors for a long time, was not being given
2683 for the ${if def:h_colon_omitted{... case.
2685 PH/34 The spec says: "With one important exception, whenever a domain list is
2686 being scanned, $domain contains the subject domain." There was at least
2687 one case where this was not true.
2689 PH/35 The error "getsockname() failed: connection reset by peer" was being
2690 written to the panic log as well as the main log, but it isn't really
2691 panic-worthy as it just means the connection died rather early on. I have
2692 removed the panic log writing for the ECONNRESET error when getsockname()
2695 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2696 runs only) independently of the message's sender address. This meant
2697 that, if the 4xx error was in fact related to the sender, a different
2698 message to the same recipient with a different sender could confuse
2699 things. In particular, this can happen when sending to a greylisting
2700 server, but other circumstances could also provoke similar problems.
2701 I have changed the default so that the retry time for these errors is now
2702 based a combination of the sender and recipient addresses. This change
2703 can be overridden by setting address_retry_include_sender=false in the
2706 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2707 remote server are returned as part of bounce messages. This was not
2708 happening for LMTP over a pipe (the lmtp transport), but now it is the
2709 same for both kinds of LMTP.
2711 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2712 in header lines that were in fact CNAMEs. This is no longer the case.
2714 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2715 and queue runs started by the daemon processed all messages. This has
2716 been fixed so that -R and -S can now usefully be given with -q<time>.
2718 PH/40 Import PCRE release 6.7 (fixes some bugs).
2720 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2722 PH/42 Give an error if -q is specified more than once.
2724 PH/43 Renamed the variables $interface_address and $interface_port as
2725 $received_ip_address and $received_port, to make it clear that these
2726 values apply to message reception, and not to the outgoing interface when
2727 a message is delivered. (The old names remain recognized, of course.)
2729 PH/44 There was no timeout on the connect() call when using a Unix domain
2730 socket in the ${readsocket expansion. There now is.
2732 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2733 be meaningful with "accept".
2735 SC/01 Eximstats V1.43
2736 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2738 SC/02 Eximstats V1.44
2739 Use a glob alias rather than an array ref in the generated
2740 parser. This improves both readability and performance.
2742 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2743 Collect SpamAssassin and rejection statistics.
2744 Don't display local sender or destination tables unless
2745 there is data to show.
2746 Added average volumes into the top table text output.
2748 SC/04 Eximstats V1.46
2749 Collect data on the number of addresses (recipients)
2750 as well as the number of messages.
2752 SC/05 Eximstats V1.47
2753 Added 'Message too big' to the list of mail rejection
2754 reasons (thanks to Marco Gaiarin).
2756 SC/06 Eximstats V1.48
2757 Mainlog lines which have GMT offsets and are too short to
2758 have a flag are now skipped.
2760 SC/07 Eximstats V1.49 (Alain Williams)
2761 Added the -emptyok flag.
2763 SC/08 Eximstats V1.50
2764 Fixes for obtaining the IP address from reject messages.
2766 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2767 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2768 whitespace changes from 4.64-PH/27
2770 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2773 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2774 are found, allow negative numbers in numeric criteria)
2776 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2778 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2779 to match changes made in 4.64-PH/43
2781 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2783 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2784 "domain" is actually an IP address, and force a failure. This locks out
2785 those revolvers/nameservers that support "A-for-A" lookups, in
2786 contravention of the specifications.
2788 PH/48 When a host name was looked up from an IP address, and the subsequent
2789 forward lookup of the name timed out, the host name was left in
2790 $sender_host_name, contrary to the specification.
2792 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2793 restricted to single-key lookups, Exim was not diagnosing an error if
2794 * or *@ was used with a query-style lookup.
2796 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2798 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2799 long ago, but noone interested enough thought of it. Let's just say that
2800 the "1.1" means that there are some new functions that weren't there at
2801 some point in the past.
2803 PH/51 Error processing for expansion failure of helo_data from an smtp
2804 transport during callout processing was broken.
2806 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2807 tested/used via the -bh/-bhc/-bs options.
2809 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2810 bug, fixed in subsequent PCRE releases).
2812 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2813 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2815 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2821 SC/01 Use a glob alias rather than an array ref in eximstats generated
2822 parser. This improves both readability and performance.
2824 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2825 Don't display local sender or destination tables in eximstats unless
2826 there is data to show.
2827 Added average volumes into the eximstats top table text output.
2829 SC/03 Collect data on the number of addresses (recipients) as well
2830 as the number of messages in eximstats.
2832 TF/01 Correct an error in the documentation for the redirect router. Exim
2833 does (usually) call initgroups() when daemonizing.
2835 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2836 with consistent privilege compared to when running as a daemon.
2838 TF/03 Note in the spec that $authenticated_id is not set for local
2839 submissions from trusted users.
2841 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2842 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2844 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2845 by adding some example configuration directives to the default
2846 configuration file. A little bit of work is required to uncomment the
2847 directives and define how usernames and passwords are checked, but
2848 there is now a framework to start from.
2850 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2851 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2852 without this. I don't know how relevant this is to other LDAP libraries.
2854 PH/02 Add the verb name to the "unknown ACL verb" error.
2856 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2858 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2860 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2861 directory not expanded when it should be if an expanded home directory
2862 was set for the address (which is overridden by the transport).
2864 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2867 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2868 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2869 because it is too late at that time, and has no effect.
2871 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2872 security issue with \' (bugzilla #107). I could not use the
2873 PQescapeStringConn() function, because it needs a PGconn value as one of
2876 PH/08 When testing addresses using -bt, indicate those final addresses that
2877 are duplicates that would not cause an additional delivery. At least one
2878 person was confused, thinking that -bt output corresponded to deliveries.
2879 (Suppressing duplicates isn't a good idea as you lose the information
2880 about possibly different redirections that led to the duplicates.)
2882 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2883 systems where poll() doesn't work, in particular OS X.
2885 PH/10 Added more information to debugging output for retry time not reached.
2887 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2888 operations in malware.c.
2890 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2893 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2894 syslog_duplication set false, log lines that would normally be written
2895 both the the main log and to the reject log were not written to syslog at
2898 PH/14 In the default configuration, change the use of "message" in ACL warn
2899 statements to "add_header".
2901 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2902 not followed by a command (e.g. "seen endif").
2904 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2905 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2908 PH/17 Added extra conditions to the default value of delay_warning_condition
2912 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2913 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2914 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2917 The Auto-Submitted: and various List- headers are standardised, whereas I
2918 don't think Precedence: ever was.
2920 PH/18 Refactored debugging code in route_finduser() to show more information,
2921 in particular, the error code if getpwnam() issues one.
2923 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2924 This is apparently needed in addition to the PH/07 change above to avoid
2925 any possible encoding problems.
2927 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2928 but not after initializing Perl.
2930 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2931 output them only if debugging. By default they are written stderr,
2932 apparently, which is not desirable.
2934 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2937 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2940 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2942 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2943 authenticated or an ident call has been made. Suppress the default
2944 values for $authenticated_id and $authenticated_sender (but permit -oMai
2945 and -oMas) when testing with -bh.
2947 PH/24 Re-jigged the order of the tests in the default configuration so that the
2948 tests for valid domains and recipients precede the DNS black list and CSA
2949 tests, on the grounds that those ones are more expensive.
2951 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2952 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2953 This bug exists in every version of Exim that I still have, right back to
2956 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2957 However, an attempt to turn on case-sensitivity in a regex key by
2958 including (?-i) didn't work because the subject string was already
2959 lowercased, and the effects were non-intuitive. It turns out that a
2960 one-line patch can be used to allow (?-i) to work as expected.
2966 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2967 other effects) broke the use of negated acl sub-conditions.
2969 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2972 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2973 "Deny" causes Exim to reject the incoming connection with a 554 error.
2974 Unfortunately, if there is a major crisis, such as a disk failure,
2975 tcp-wrappers gives "deny", whereas what one would like would be some
2976 kind of temporary error. A kludge has been added to help with this.
2977 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2978 554 error is used if errno is still zero or contains ENOENT (which occurs
2979 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2982 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2984 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2985 errors. Otherwise a message that provokes a temporary error (when other
2986 messages do not) can cause a whole host to time out.
2988 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2989 addresses were routed directly to files or pipes from a redirect router.
2990 File deliveries just didn't batch; pipe deliveries might have suffered
2993 PH/06 A failure to get a lock for a hints database would erroneously always say
2994 "Failed to get write lock", even when it was really a read lock.
2996 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2997 of 0600. This has been changed to use the value of the lockfile_mode
2998 option (which defaults to 0600).
3000 PH/08 Applied small patch from the Sieve maintainer.
3002 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3003 folder from quota calculations, a direct delivery into this folder messed
3004 up the contents of the maildirsize file. This was because the regex was
3005 used only to exclude .Trash (or whatever) when the size of the mailbox
3006 was calculated. There was no check that a delivery was happening into an
3007 excluded directory. This bug has been fixed by ignoring all quota
3008 processing for deliveries into excluded directories.
3010 PH/10 Added the maildirfolder_create_regex option to appendfile.
3016 PH/01 The code for finding all the local interface addresses on a FreeBSD
3017 system running IPv6 was broken. This may well have applied to all BSD
3018 systems, as well as to others that have similar system calls. The broken
3019 code found IPv4 interfaces correctly, but gave incorrect values for the
3020 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3021 that it would not match correctly against @[] and not recognize the IPv6
3024 PH/02 The ipliteral router was not recognizing addresses of the form user@
3025 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3027 PH/03 Added disable_ipv6.
3029 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3030 decoded form, because it is most often used to construct To: headers
3031 lines in autoreplies, and the decoded form may well be syntactically
3032 invalid. However, $reply_address has leading white space removed, and all
3033 newlines turned into spaces so that the autoreply transport does not
3036 PH/05 If group was specified without a user on a router, and no group or user
3037 was specified on a transport, the group from the router was ignored.
3039 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3040 for visible compile-time settings that can be used to change these
3041 numbers, for those that want even more. Backwards compatibility with old
3042 spool files has been maintained. However, going back to a previous Exim
3043 release will lost any variables that are in spool files.
3045 PH/07 Two small changes when running in the test harness: increase delay when
3046 passing a TCP/IP connection to a new process, in case the original
3047 process has to generate a bounce, and remove special handling of
3048 127.0.0.2 (sic), which is no longer necessary.
3050 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3051 be the same on different OS.
3053 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3056 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3057 whether --show-vars was specified or not
3059 JJ/02 exipick: Added support for new ACL variable spool format introduced
3062 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3063 syntactically invalid From: or Reply-to: line, and a filter used this to
3064 generate an autoreply, and therefore failed to obtain an address for the
3065 autoreply, Exim could try to deliver to a non-existent relative file
3066 name, causing unrelated and misleading errors. What now happens is that
3067 it logs this as a hard delivery error, but does not attempt to create a
3070 PH/11 The exinext utility has a -C option for testing purposes, but although
3071 the given file was scanned by exinext itself; it wasn't being passed on
3072 when Exim was called.
3074 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3075 an end-of-file indication when reading a command response.
3077 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3078 compiled. In many other places in Exim, IPv6 addresses are always
3079 recognized, so I have changed this. It also means that IPv4 domain
3080 literals of the form [IPV4:n.n.n.n] are now always recognized.
3082 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3083 used if the router is not running as root, for example, when verifying at
3084 ACL time, or when using -bh. The debugging output from this situation was
3085 non-existent - all you got was a failure to exec. I have made two
3088 (a) Failures to set uid/gid, the current directory, or a process leader
3089 in a subprocess such as that created by queryprogram now generate
3090 suitable debugging output when -d is set.
3092 (b) The queryprogram router detects when it is not running as root,
3093 outputs suitable debugging information if -d is set, and then runs
3094 the subprocess without attempting to change uid/gid.
3096 PH/15 Minor change to Makefile for building test_host (undocumented testing
3099 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3100 additional section of a DNS packet that returns MX or SRV records.
3101 Instead, it always explicitly searches for A/AAAA records. This avoids
3102 major problems that occur when a DNS server includes only records of one
3103 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3104 fixed another bug: if SRV records were looked up and the corresponding
3105 address records were *not* found in the additional section, the port
3106 values from the SRV records were lost.
3108 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3109 using the correct key (the original address) when searching the retry
3110 rules in order to find which one to use for generating the retry hint.
3112 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3113 adding the default one. Similarly, if it contains a Reply-To: header, the
3114 errors_reply_to option, if set, is not used.
3116 PH/19 When calculating a retry time, Exim used to measure the "time since
3117 failure" by looking at the "first failed" field in the retry record. Now
3118 it does not use this if it is later than than the arrival time of the
3119 message. Instead it uses the arrival time. This makes for better
3120 behaviour in cases where some deliveries succeed, thus re-setting the
3121 "first failed" field. An example is a quota failure for a huge message
3122 when small messages continue to be delivered. Without this change, the
3123 "time since failure" will always be short, possible causing more frequent
3124 delivery attempts for the huge message than are intended.
3125 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3127 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3128 $1, $2, $3) because the numerical variables can be reset during some
3129 expansion items (e.g. "match"), thereby losing the authentication data.
3131 PH/21 Make -bV show the size of off_t variables so that the test suite can
3132 decide whether to run tests for quotas > 2G.
3134 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3135 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3136 filecount value is greater than 2G or if a quota value is greater than 2G
3137 on a system where the size of off_t is not greater than 4, a panic error
3140 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3141 never match. The debug and -bh output now contains an explicit error
3142 message indicating a malformed IPv4 address or mask.
3144 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3145 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3146 PH/23 above applies.
3148 PH/25 Do not write to syslog when running in the test harness. The only
3149 occasion when this arises is a failure to open the main or panic logs
3150 (for which there is an explicit test).
3152 PH/26 Added the /no_tell option to "control=freeze".
3154 PH/27 If a host name lookup failed very early in a connection, for example, if
3155 the IP address matched host_lookup and the reverse lookup yielded a name
3156 that did not have a forward lookup, an error message of the form "no IP
3157 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3158 could be logged. Now it outputs the IP address instead of "NULL".
3160 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3161 allows the sender and the authenticated sender to be set when
3162 submitting a message from within Exim. Since child_open_exim() is
3163 documented for local_scan(), the new function should be too.
3165 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3166 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3167 results in an empty string is now treated as unset.
3169 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3171 PH/31 Added sender_verify_fail logging option.
3173 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3174 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3175 tidied the source and removed it altogether.
3177 PH/33 When a queue run was abandoned because the load average was too high, a
3178 log line was always written; now it is written only if the queue_run log
3179 selector is set. In addition, the log line for abandonment now contains
3180 information about the queue run such as the pid. This is always present
3181 in "start" and "stop" lines but was omitted from the "abandon" line.
3183 PH/34 Omit spaces between a header name and the colon in the error message that
3184 is given when verify = headers_syntax fails (if there are lots of them,
3185 the message gets confusing).
3187 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3188 names, as there are now some PTR records that contain slashes. This check
3189 is only to protect against broken name servers that fall over on strange
3190 characters, so the fact that it applies to all lookups doesn't matter.
3192 PH/36 Now that the new test suite is complete, we can remove some of the
3193 special code in Exim that was needed for the old test suite. For example,
3194 sorting DNS records because real resolvers return them in an arbitrary
3195 order. The new test suite's fake resolver always returns records in the
3198 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3199 bounces) except when queue_only is set, to avoid logging races between
3200 the different processes.
3202 PH/38 Panic-die if .include specifies a non-absolute path.
3204 PH/39 A tweak to the "H" retry rule from its user.
3206 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3207 a label. They prevented compilation on older perls.
3209 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3210 a warning to be raised on newish perls.
3212 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3213 on queue. Changes to match documented behaviour of showing count of
3214 messages matching specified criteria.
3216 PH/40 Changed the default ident timeout from 30s to 5s.
3218 PH/41 Added support for the use of login_cap features, on those BSD systems
3219 that have them, for controlling the resources used by pipe deliveries.
3221 PH/42 The content-scanning code uses fopen() to create files in which to put
3222 message data. Previously it was not paying any attention to the mode of
3223 the files. Exim runs with umask(0) because the rest of the code creates
3224 files with open(), and sets the required mode explicitly. Thus, these
3225 files were ending up world-writeable. This was not a big issue, because,
3226 being within the spool directory, they were not world-accessible. I have
3227 created a function called modefopen, which takes an additional mode
3228 argument. It sets umask(777), creates the file, chmods it to the required
3229 mode, then resets the umask. All the relevant calls to fopen() in the
3230 content scanning code have been changed to use this function.
3232 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3233 to 24 hours. This avoids potential overflow problems when processing G
3234 and H retry rules. I suspect nobody ever tinkers with this value.
3236 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3238 PH/45 When the plaintext authenticator is running as a client, the server's
3239 challenges are checked to ensure they are valid base64 strings. By
3240 default, the authentication attempt is cancelled if an invalid string is
3241 received. Setting client_ignore_invalid_base64 true ignores these errors.
3242 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3243 they are received. Thus, the responses can be made to depend on the
3244 challenges. If an invalid string is ignored, an empty string is placed in
3247 PH/46 Messages that are created by the autoreply transport now contains a
3248 References: header, in accordance with RFCs 2822 and 3834.
3250 PH/47 Added authenticated_sender_force to the smtp transport.
3252 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3254 PH/49 Installed latest patch from the Sieve maintainer.
3256 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3257 was also set, the appendfile transport was unnecessarily scanning a
3258 directory of message files (e.g. for maildir delivery) to find the count
3259 of files (along with the size), even though it did not need this
3260 information. It now does the scan only if it needs to find either the
3261 size of the count of files.
3263 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3265 PH/52 Two bugs concerned with error handling when the smtp transport is
3268 (i) Exim was not creating retry information for temporary errors given
3269 for individual recipients after the DATA command when the smtp transport
3270 was used in LMTP mode. This meant that they could be retried too
3271 frequently, and not timed out correctly.
3273 (ii) Exim was setting the flag that allows error details to be returned
3274 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3275 recipients that were returned after the DATA command.
3277 PH/53 This is related to PH/52, but is more general: for any failing address,
3278 when detailed error information was permitted to be returned to the
3279 sender, but the error was temporary, then after the final timeout, only
3280 "retry timeout exceeded" was returned. Now it returns the full error as
3281 well as "retry timeout exceeded".
3283 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3284 do this, and (what is worse) MTAs that accept it.
3286 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3287 will now be deprecated.
3289 PH/56 New os.c-cygwin from the Cygwin maintainer.
3291 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3292 formats (previously only available in exim formats via -bpr, -bpru,
3293 and -bpra. Now also available in native and exiqgrep formats)
3295 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3296 with very large, slow to parse queues
3298 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3300 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3302 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3303 responses to authentication challenges, though it was showing the
3304 challenges; (ii) I've removed the CR characters from the debug output for
3307 PH/58 Allow for the insertion of a newline as well as a space when a string
3308 is turned into more than one encoded-word during RFC 2047 encoding. The
3309 Sieve code now uses this.
3311 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3312 data_4xx, lost_connection, tls_required.
3314 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3315 message was being sent as an SMTP response.
3317 PH/61 Add -l and -k options to exicyclog.
3319 PH/62 When verifying, if an address was redirected to one new address, so that
3320 verification continued, and the new address failed or deferred after
3321 having set something in $address_data, the value of $address_data was not
3322 passed back to the ACL. This was different to the case when no
3323 redirection occurred. The value is now passed back in both cases.
3325 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3326 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3327 use login_cap.h, so on its own it isn't the distinguishing feature. The
3328 new name refers directly to the setclassresources() function.
3330 PH/65 Added configuration files for NetBSD3.
3332 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3334 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3335 is preferred over IPv4.
3337 PH/68 The bounce_return_message and bounce_return_body options were not being
3338 honoured for bounces generated during the reception of non-SMTP messages.
3339 In particular, this applied to messages rejected by the ACL. This bug has
3340 been fixed. However, if bounce_return_message is true and bounce_return_
3341 body is false, the headers that are returned for a non-SMTP message
3342 include only those that have been read before the error was detected.
3343 (In the case of an ACL rejection, they have all been read.)
3345 PH/69 The HTML version of the specification is now built in a directory called
3346 spec_html instead of spec.html, because the latter looks like a path with
3347 a MIME-type, and this confuses some software.
3349 PH/70 Catch two compiler warnings in sieve.c.
3351 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3352 function verify_get_ident() calls ip_connect() to connect a socket, but
3353 if the "connect()" function timed out, ip_connect() used to close the
3354 socket. However, verify_get_ident() also closes the socket later, and in
3355 between Exim writes to the log, which may get opened at this point. When
3356 the socket was closed in ip_connect(), the log could get the same file
3357 descriptor number as the socket. This naturally causes chaos. The fix is
3358 not to close the socket in ip_connect(); the socket should be closed by
3359 the function that creates it. There was only one place in the code where
3360 this was missing, in the iplookup router, which I don't think anybody now
3361 uses, but I've fixed it anyway.
3363 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3364 well as to direct DNS lookups. Otherwise the handling of names in host
3365 lists is inconsistent and therefore confusing.
3371 PH/01 Two changes to the default runtime configuration:
3373 (1) Move the checks for relay_from_hosts and authenticated clients from
3374 after to before the (commented out) DNS black list checks.
3376 (2) Add control=submission to the relay_from_hosts and authenticated
3377 clients checks, on the grounds that messages accepted by these
3378 statements are most likely to be submissions.
3380 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3382 (1) Generate an error if the third argument for the ${prvs expansion is
3385 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3388 (3) Reset the variables that are obtained from the first argument of
3389 ${prvscheck and used in the second argument before leaving the code,
3390 because their memory is reclaimed, so using them afterwards may do
3393 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3394 one (it's much easier than Tom thought :-).
3396 (5) Because of (4), we can now allow for the use of $prvscheck_result
3397 inside the third argument.
3399 PH/03 For some reason, the default setting of PATH when running a command from
3400 a pipe transport was just "/usr/bin". I have changed it to
3403 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3404 anything to be listed in the output from -bV.
3406 PH/05 When a filter generated an autoreply, the entire To: header line was
3407 quoted in the delivery log line, like this:
3409 => >A.N.Other <ano@some.domain> <original@ddress> ...
3411 This has been changed so that it extracts the operative address. There
3412 may be more than one such address. If so, they are comma-separated, like
3415 => >ano@some.domain,ona@other.domain <original@ddress> ...
3417 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3418 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3419 not being looked up in the rDNS to get a host name, Exim was showing the
3420 IP address twice in Received: lines, even though the IP addresses were
3421 identical. For example:
3423 Received: from [1.2.3.4] (helo=[1.2.3.4])
3425 However, if the real host name was known, it was omitting the HELO data
3426 if it matched the actual IP address. This has been tidied up so that it
3427 doesn't show the same IP address twice.
3429 PH/07 When both +timestamp and +memory debugging was on, the value given by
3430 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3431 called by the debug printing, thereby overwriting the timestamp buffer.
3432 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3434 PH/08 When the original message was included in an autoreply transport, it
3435 always said "this is a copy of the message, including all the headers",
3436 even if body_only or headers_only was set. It now gives an appropriate
3439 PH/09 Applied a patch from the Sieve maintainer which:
3441 o fixes some comments
3442 o adds the (disabled) notify extension core
3443 o adds some debug output for the result of if/elsif tests
3444 o points to the current vacation draft in the documentation
3445 and documents the missing references header update
3449 o fixes a bug in processing the envelope test (when testing
3450 multiple envelope elements, the last element determined the
3453 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3454 Electronic Mail") by including:
3456 Auto-submitted: auto-generated
3458 in the messages that it generates (bounce messages and others, such as
3459 warnings). In the case of bounce messages for non-SMTP messages, there was
3460 also a typo: it was using "Auto_submitted" (underscore instead of
3461 hyphen). Since every message generated by Exim is necessarily in response
3462 to another message, thes have all been changed to:
3464 Auto-Submitted: auto-replied
3466 in accordance with these statements in the RFC:
3468 The auto-replied keyword:
3470 - SHOULD be used on messages sent in direct response to another
3471 message by an automatic process,
3473 - MUST NOT be used on manually-generated messages,
3475 - MAY be used on Delivery Status Notifications (DSNs) and Message
3476 Disposition Notifications (MDNs),
3478 - MUST NOT be used on messages generated by automatic or periodic
3479 processes, except for messages which are automatic responses to
3482 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3483 to the default Received: header definition.
3485 PH/12 Added log selector acl_warn_skipped (default on).
3487 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3488 variables because (a) they are in the wrong storage pool and (b) even if
3489 they were copied, it wouldn't work properly because of the caching.
3491 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3492 checking when decoding. Apparently there are clients that generate
3493 overlong encoded strings. Why am I not surprised?
3495 PH/15 If the first argument of "${if match_address" was not empty, but did not
3496 contain an "@" character, Exim crashed. Now it writes a panic log message
3497 and treats the condition as false.
3499 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3501 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3502 extension "envelope-auth". The code is finished and in agreement with
3503 other implementations, but there is no documentation so far and in fact,
3504 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3505 not changing the active code.
3507 Print executed "if" and "elsif" statements when debugging is used. This
3508 helps a great deal to understand what a filter does.
3510 Document more things not specified clearly in RFC3028. I had all this
3511 sorted out, when out of a sudden new issues came to my mind. Oops."
3513 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3516 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3517 canonical form (as documented). However, after a host name lookup from
3518 the IP address, check_host() was doing a simple string comparison with
3519 addresses acquired from the DNS when checking that the found name did
3520 have the original IP as one of its addresses. Since any found IPv6
3521 addresses are likely to be in abbreviated form, the comparison could
3522 fail. Luckily, there already exists a function for doing the comparison
3523 by converting both addresses to binary, so now that is used instead of
3524 the text comparison.
3526 PH/20 There was another similar case to PH/19, when a complete host name was
3527 given in a host list; looking up its IP address could give an abbreviated
3528 form, whereas the current host's name might or might not be abbreviated.
3529 The same fix has been applied.
3535 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3536 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3539 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3540 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3542 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3544 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3545 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3546 Also, for "Auto-Submitted", treat anything other than "no" as
3547 non-personal, in accordance with RFC 3834. (Previously it treated
3548 anything starting "auto-" as non-personal.)
3550 TF/01 The control=submission/name=... option had a problem with syntax
3551 errors if the name included a slash character. The /name= option
3552 now slurps the rest of the string, so it can include any characters
3553 but it must come last in the list of options (after /sender_retain
3556 PH/05 Some modifications to the interface to the fake nameserver for the new
3564 TK/01 Added the "success_on_redirect" address verification option. See
3565 NewStuff for rationale and an example.
3567 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3569 PH/02 Patch to exigrep to allow it to work on syslog lines.
3571 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3572 fread() to skip over the body file's header line, because in Cygwin the
3573 header line is locked and is inaccessible.
3575 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3576 co-exist for some time) to make it clear that it is the Exim ID that is
3577 referenced, not the Message-ID: header line.
3579 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3580 string_format() function, because snprintf() does not exist on all
3583 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3584 ${stat: expansion item.
3586 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3587 protocol synchronization error", to keep the pedants happy.
3589 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3590 well as for IRIX systems, when gcc is being used. See the host.c source
3593 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3595 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3598 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3599 transport and to the smtp transport in LMTP mode.
3601 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3603 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3604 run a filter in a subprocess. This could lead to confusion in subsequent
3605 lookups in the parent process. There should also be a search_tidyup() at
3606 the end of the subprocess.
3608 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3609 only if the host matched helo_try_verify_hosts, which caused the
3610 verification to occur when the EHLO/HELO command was issued. The ACL just
3611 tested the remembered result. Now, if a previous verification attempt has
3612 not happened, "verify = helo" does it there and then.
3614 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3616 TK/03 Fix log output including CR from clamd.
3618 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3619 exist provoked a memory error which could cause a segfault.
3621 PH/15 Installed PCRE 6.2
3623 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3625 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3626 of the problem. Specifically, suggested +O2 rather than +O1 for the
3629 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3631 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3632 record, the port was not correctly logged when the outgoing_port log
3633 selector was set (it logged the transort's default port).
3635 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3636 fallback_hosts, and "hosts" in the smtp transport.
3638 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3639 host errors such as "Connection refused".
3641 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3642 authentication with radiusclient 0.4.9:
3644 - Error returned from rc_read_config was caught wrongly
3645 - Username/password not passed on to radius server due to wrong length.
3647 The presumption is that some radiusclient API changes for 4.51/PH/17
3648 were not taken care of correctly. The code is still untested by me (my
3649 Linux distribution still has 0.3.2 of radiusclient), but it was
3650 contributed by a Radius user.
3652 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3653 expanding the "port" option of the smtp transport.
3655 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3656 while reading a MIME header. Thanks to Tom Hughes for a patch.
3658 PH/24 Include config.h inside local_scan.h so that configuration settings are
3661 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3662 that in an ACL for RCPT (for example), you can examine exactly what was
3665 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3666 commands, but it was not correctly comparing the address with the actual
3667 client host address. Thus, it would show the EHLO address in Received:
3668 header lines when this was not necessary.
3670 PH/27 Added the % operator to ${eval:}.
3672 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3673 it should be ignoring failures (because with -C, for example, it has lost
3674 privilege). It wasn't ignoring creation failures other than "already
3677 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3680 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3681 because an input line was too long, either on its own, or by virtue of
3682 too many continuations, the temporary file was not being removed, and the
3683 return code was incorrect.
3685 PH/31 Missing "BOOL" in function definition in filtertest.c.
3687 PH/32 Applied Sieve patches from the maintainer.
3689 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3691 PH/33 Added "verify = not_blind".
3693 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3694 Local/Makefile (with some defaults set). These are used in built scripts
3695 such as exicyclog, but they have never been used in the exim_install
3696 script (though there are many overriding facilities there). I have
3697 arranged that the exim_install script now takes note of these two
3700 PH/35 Installed configuration files for Dragonfly.
3702 PH/36 When a locally submitted message by a trusted user did not contain a
3703 From: header, and the sender address was obtained from -f or from an SMTP
3704 MAIL command, and the trusted user did not use -F to supply a sender
3705 name, $originator_name was incorrectly used when constructing a From:
3706 header. Furthermore, $originator_name was used for submission mode
3707 messages from external hosts without From: headers in a similar way,
3708 which is clearly wrong.
3710 PH/37 Added control=suppress_local_fixups.
3712 PH/38 When log_selector = +received_sender was set, and the addition of the
3713 sender made the log line's construction buffer exactly full, or one byte
3714 less than full, an overflow happened when the terminating "\n" was
3717 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3718 when the result of a list match is failure because a DNS lookup failed.
3720 PH/40 RM_COMMAND is now used in the building process.
3722 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3723 the "build-* directories that it finds.
3725 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3726 address in a domain literal was a prefix of an interface address.
3728 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3729 when verifying a sender address, unless rewrite_headers is false.
3731 PH/44 Wrote a long comment about why errors_to addresses are verified as
3732 recipients, not senders.
3734 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3735 the ratelimit ACL was added.
3737 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3739 PH/46 Added extra information about PostgreSQL errors to the error string.
3741 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3742 avoiding the need to install special zones in a real server. This is
3743 backwards compatible; if it can't find the fake resolver, it drops back.
3744 Thus, both old and new test suites can be run.
3746 TF/02 Added util/ratelimit.pl
3748 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3749 clock is set back in time.
3751 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3752 Candler <B.Candler@pobox.com>.
3754 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3755 OK for addresses that are the result of redirections.
3757 PH/48 A number of further additions for the benefit of the new test suite,
3758 including a fake gethostbyname() that interfaces to the fake DNS resolver
3761 TF/06 The fix for widen_domains has also been applied to qualify_single and
3762 search_parents which are the other dnslookup options that can cause
3765 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3768 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3770 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3771 service" if it encountered an MX record with an empty target hostname.
3772 The message is now "an MX or SRV record indicated no SMTP service".
3774 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3775 if the DNS of the sending site is misconfigured. This is quite a
3776 common situation. This change restores the behaviour of treating a
3777 helo verification defer as a failure.
3779 PH/51 If self=fail was set on a router, the bounce message did not include the
3780 actual error message.
3786 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3788 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3789 command itself ended in error, the underlying message about the transport
3790 filter timeout was being overwritten with the pipe command error. Now the
3791 underlying error message should be appended to the second error message.
3793 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3795 PH/02 Reduce the amount of output that "make" produces by default. Full output
3796 can still be requested.
3798 PH/03 The warning log line about a condition test deferring for a "warn" verb
3799 was being output only once per connection, rather than after each
3800 occurrence (because it was using the same function as for successful
3801 "warn" verbs). This seems wrong, so I have changed it.
3803 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3804 it should not have, which might have caused a crash in the right
3805 circumstances, but probably never did.
3807 PH/04 Installed a modified version of Tony Finch's patch to make submission
3808 mode fix the return path as well as the Sender: header line, and to
3809 add a /name= option so that you can make the user's friendly name appear
3812 TF/03 Added the control = fakedefer ACL modifier.
3814 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3815 Mark Lowes for thorough testing.
3817 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3819 TK/03 Merged latest SRS patch from Miles Wilton.
3821 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3822 with the definition in sysexits.h (which is #included earlier).
3823 Fortunately, Exim does not actually use EX_OK. The code used to try to
3824 preserve the sysexits.h value, by assuming that macro definitions were
3825 scanned for macro replacements. I have been disabused of this notion,
3826 so now the code just undefines EX_OK before #including unistd.h.
3828 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3829 in the smtp transport. When a block could not be written in a single
3830 write() function, the timeout was being re-applied to each part-write.
3831 This seems wrong - if the receiver was accepting one byte at a time it
3832 would take for ever. The timeout is now adjusted when this happens. It
3833 doesn't have to be particularly precise.
3835 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3836 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3838 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3839 without a domain if the check to <postmaster@domain> fails.
3841 SC/01 Eximstats: added -xls and the ability to specify output files
3842 (patch written by Frank Heydlauf).
3844 SC/02 Eximstats: use FileHandles for outputting results.
3846 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3848 SC/04 Eximstats: fixed display of large numbers with -nvr option
3850 SC/05 Eximstats: fixed merging of reports with empty tables.
3852 SC/06 Eximstats: added the -include_original_destination flag
3854 SC/07 Eximstats: removed tabs and trailing whitespace.
3856 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3858 TK/06 MBOX spool code: Add real "From " MBOX separator line
3859 so the .eml file is really in mbox format (even though
3860 most programs do not really care). Patch from Alex Miller.
3862 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3863 The latter is generated from $received_to and is only set if the
3864 message has one envelope recipient. SA can use these headers,
3865 obviously out-of-the-box. Patch from Alex Miller.
3867 PH/08 The ${def test on a variable was returning false if the variable's
3868 value was "0", contrary to what the specification has always said!
3869 The result should be true unless the variable is empty.
3871 PH/09 The syntax error of a character other than { following "${if
3872 def:variable_name" (after optional whitespace) was not being diagnosed.
3873 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3874 accidental colon was present, for example, could give incorrect results.
3876 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3877 result is used (not including appendfile, where other changes are about
3880 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3881 This involved changing a lot of size variables from int to off_t. It
3882 should work with maildirs and everything.
3884 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3885 spamd dying while we are connected to it.
3887 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3890 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3891 function for BDB 4.3.
3893 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3895 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3896 the sources for PCRE have been re-arranged and more files are now
3899 PH/15 The code I had for printing potentially long long variables in PH/11
3900 above was not the best (it lost precision). The length of off_t variables
3901 is now inspected at build time, and an appropriate printing format (%ld
3902 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3903 to be "long long int" or "long int". This is needed for the internal
3904 formatting function string_vformat().
3906 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3907 the configuration file to be ":syslog", then the script "guesses" where
3908 the logs files are, rather than using the compiled in default. In our
3909 case the guess is not the same as the compiled default, so the script
3910 suddenly stopped working when I started to use syslog. The patch checks
3911 to see if log_file_path is "". If so, it attempts to read it from exim
3912 with no configuration file to get the compiled in version, before it
3913 falls back to the previous guessing code."
3915 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3916 implementing BATV in an Exim configuration. See NewStuff for the gory
3919 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3920 Makefile that are specific to HP-UX.
3922 PH/18 If the "use_postmaster" option was set for a recipient callout together
3923 with the "random" option, the postmaster address was used as the MAIL
3924 FROM address for the random test, but not for the subsequent recipient
3925 test. It is now used for both.
3927 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3928 patch removes a few documentation additions to RFC 3028, because the
3929 latest draft now contains them. It adds the new en;ascii-case comparator
3930 and a new error check for 8bit text in MIME parts. Comparator and
3931 require names are now matched exactly. I enabled the subaddress
3932 extension, but it is not well tested yet (read: it works for me)."
3934 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3935 rework some of the code of TK/09 above to avoid the hardwired use of
3936 "%lld" and "long long". Replaced the call to snprintf() with a call to
3939 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3940 records point to non-existent hosts", "retry timeout exceeded", and
3941 "retry time not reached for any host after a long failure period".
3943 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3944 experimental DomainKeys support:
3946 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3947 (2) On an error such as an illegally used "control", the wrong name for
3948 the control was given.
3950 These problems did NOT occur unless DomainKeys support was compiled.
3952 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3954 PH/24 Added ${if match_ip condition.
3956 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3957 databases so that it will be absolutely obvious if a crash occurs in the
3958 DB library. This is a regular occurrence (often caused by mis-matched
3961 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3962 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3963 picked up on a user's system that detects such things. There doesn't seem
3964 to be a gcc warning option for this - only an attribute that has to be
3965 put on the function's prototype. It seems that in Fedora Core 4 they have
3966 set this on a number of new functions. No doubt there will be more in due
3969 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3970 specify a transport. However, if an address that was verified by such a
3971 router was the subject of a callout, Exim crashed because it tried to
3972 read the rcpt_include_affixes from the non-existent transport. Now it
3973 just assumes that the setting of that option is false. This bug was
3974 introduced by 4.51/PH/31.
3976 PH/28 Changed -d+all to exclude +memory, because that information is very
3977 rarely of interest, but it makes the output a lot bigger. People tend to
3978 do -d+all out of habit.
3980 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3981 code in os-type was giving problems when libc.so lives in lib64, like on
3984 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3985 aren't the modern standard, and it seems that some systems' include files
3986 don't always have them. Exim was already checking for some of the newer
3987 ones like T_AAAA, and defining it itself. I've added checks for all the
3988 record types that Exim uses.
3990 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3991 not automatically generating a new one, as it is supposed to. This
3992 prevented TLS from working. If the file did exist, but contained invalid
3993 data, a new version was generated, as expected. It was only the case of a
3994 non-existent file that was broken.
3996 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3997 with a change in libdomainkeys > 0.64.
3999 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4000 from DNS. If the selector record carries the flag, it now has
4001 precedence over the domain-wide flag.
4003 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4005 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4006 the use of a port name that isn't defined in /etc/services) occurred, the
4007 message was deferred as in a normal delivery, and thus remained on the
4008 spool, instead of being failed because of the mua_wrapper setting. This
4009 is now fixed, and I tidied up some of the mua_wrapper messages at the
4012 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4013 the messages in a hash of arrays rather than using individual hashes.
4014 This is a bit cleaner and results in dramatic memory savings, albeit
4015 at a slight CPU cost.
4017 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4018 as requested by Marc Sherman.
4020 SC/10 Eximstats: added histograms for user specified patterns as requested
4023 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4025 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4026 fopen() in the content-scanning modules that did not already have it.
4032 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4033 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4035 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4037 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4039 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4040 to test Sieve filters that use "vacation".
4042 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4043 that changes the way the GnuTLS parameters are stored in the cache file.
4044 The new format can be generated externally. For backward compatibility,
4045 if the data in the cache doesn't make sense, Exim assumes it has read an
4046 old-format file, and it generates new data and writes a new file. This
4047 means that you can't go back to an older release without removing the
4050 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4051 work if there are any delivery delays because "one_time" forces the
4052 parent to be marked "delivered", so its unseen clone is never tried
4053 again. For this reason, Exim now forbids the simultaneous setting of
4056 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4057 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4058 entry if you want to know the details. The fix, however, neglected to
4059 consider the case when local delivery batching is involved. The test for
4060 "previously delivered" was not happening when checking to see if an
4061 address could be batched with a previous (undelivered) one; under
4062 certain circumstances this could lead to multiple deliveries to the same
4065 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4066 in its include files, and this causes problems building Exim.
4068 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4069 header_syntax) but Exim was just ignoring anything given after a slash.
4070 In particular, this caused confusion with an attempt to use "verify =
4071 reverse_host_lookup/defer_ok". An error is now given when options are
4072 supplied for verify items that do not have them. (Maybe reverse_host_
4073 lookup should have a defer_ok option, but that's a different point.)
4075 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4076 defined by RFC 821) to 2048, because there were problems with some AUTH
4077 commands, and RFC 1869 says the size should be increased for extended
4078 SMTP commands that take arguments.
4080 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4083 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4084 "unknown" error; now it says that the functionality isn't in the binary.
4086 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4087 an address' error message when a string expansion fails (syntax or
4088 whatever). Otherwise the password may appear in the log. Following change
4089 PH/42 below, there is no longer a chance of it appearing in a bounce
4092 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4094 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4095 of its addresses was ever tried. (Bugzilla bug #2.)
4097 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4098 the result incorrectly in the debug output. (It correctly added a newline
4099 to what was transported.)
4101 TF/01 Added $received_time.
4103 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4104 commented out examples of how to interface to a virus scanner and to
4105 SpamAssassin. Also added commented examples of av_scanner and
4106 spamd_address settings.
4108 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4109 and controls are allowed in which ACLs. There were a couple of minor
4110 errors. Some of the entries in the conditions table (which is a table of
4111 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4112 negation of where the condition IS allowed.
4114 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4116 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4117 header file does not have a version number, so I've had to invent a new
4118 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4119 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4120 radiusclient), but it was contributed by a Radius user.
4122 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4123 files or directories, for OpenSSL.
4125 PH/19 When an Exim process that is running as root has to create an Exim log
4126 file, it does so in a subprocess that runs as exim:exim so as to get the
4127 ownership right at creation (otherwise, other Exim processes might see
4128 the file with the wrong ownership). There was no test for failure of this
4129 fork() call, which would lead to the process getting stuck as it waited
4130 for a non-existent subprocess. Forks do occasionally fail when resources
4131 run out. I reviewed all the other calls to fork(); they all seem to check
4134 PH/20 When checking for unexpected SMTP input at connect time (before writing
4135 the banner), Exim was not dealing correctly with a non-positive return
4136 from the read() function. If the client had disconnected by this time,
4137 the result was a log entry for a synchronization error with an empty
4138 string after "input=" when read() returned zero. If read() returned -1
4139 (an event I could not check), uninitialized data bytes were printed.
4140 There were reports of junk text (parts of files, etc) appearing after
4143 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4145 PH/22 Added support for macro redefinition, and (re)definition in between
4146 driver and ACL definitions.
4148 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4149 forgetting to use the resulting value; it was using the unexpanded value.
4151 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4152 hadn't been configured. The fix is from Juergen Kreileder, who
4153 understands it better than I do:
4155 "Here's what I see happening with three configured cyrus_sasl
4156 authenticators configured (plain, login, cram-md5):
4158 On startup auth_cyrus_sasl_init() gets called for each of these.
4159 This means three calls to sasl_listmech() without a specified mech_list.
4160 => SASL tests which mechs of all available mechs actually work
4161 => three warnings about OTP not working
4162 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4164 With the patch, sasl_listmech() also gets called three times. But now
4165 SASL's mech_list option is set to the server_mech specified in the the
4166 authenticator. Or in other words, the answer from sasl_listmech()
4167 gets limited to just the mech you're testing for (which is different
4169 => the return list contains just 'plain' or 'login', 'cram-md5' or
4170 nothing depending on the value of ob->server_mech.
4172 I've just tested the patch: Authentication still works fine,
4173 unavailable mechs specified in the exim configuration are still
4174 caught, and the auth.log warnings about OTP are gone."
4176 PH/25 When debugging is enabled, the contents of the command line are added
4177 to the debugging output, even when log_selector=+arguments is not
4180 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4181 answer is "GNU", and only if the return is "GNU/something" is the answer
4184 PH/27 $acl_verify_message is now set immediately after the failure of a
4185 verification in an ACL, and so is available in subsequent modifiers. In
4186 particular, the message can be preserved by coding like this:
4188 warn !verify = sender
4189 set acl_m0 = $acl_verify_message
4191 Previously, $acl_verify_message was set only while expanding "message"
4192 and "log_message" when a very denied access.
4194 PH/28 Modified OS/os.c-Linux with
4196 -#ifndef OS_LOAD_AVERAGE
4197 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4199 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4200 nomenclature these days.)
4202 PH/29 Installed patch from the Sieve maintainer that adds the options
4203 sieve_useraddress and sieve_subaddress to the redirect router.
4205 PH/30 In these circumstances:
4206 . Two addresses routed to the same list of hosts;
4207 . First host does not offer TLS;
4208 . First host accepts first address;
4209 . First host gives temporary error to second address;
4210 . Second host offers TLS and a TLS session is established;
4211 . Second host accepts second address.
4212 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4213 and peerdn, if requested) that were in fact used only for the second
4216 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4217 attention to any local part prefix or suffix that was matched by the
4218 router that accepted the address. It now behaves in the same way as it
4219 does for delivery: the affixes are removed from the local part unless
4220 rcpt_include_affixes is set on the transport.
4222 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4223 timeout during the DATA phase of an incoming message.
4225 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4226 applied a patch sanctioned by the Sieve maintainer.
4228 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4229 the uid or gid is negative. A case of a negative gid caused this to be
4230 noticed. The fix allows for either to be negative.
4232 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4233 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4236 PH/36 The change PH/12 above was broken. Fixed it.
4238 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4239 the grounds that routing the same address twice would always produce the
4240 same answer. This might have been true once, but it is certainly no
4241 longer true now. Routing a child address may depend on the previous
4242 routing that produced that child. Some complicated redirection strategies
4243 went wrong when messages had multiple recipients, and made Exim's
4244 behaviour dependent on the order in which the addresses were given.
4246 I have moved the duplicate checking until after the routing is complete.
4247 Exim scans the addresses that are assigned to local and remote
4248 transports, and removes any duplicates. This means that more work will be
4249 done, as duplicates will always all be routed, but duplicates are
4250 presumably rare, so I don't expect this is of any significance.
4252 For deliveries to pipes, files, and autoreplies, the duplicate checking
4253 still happens during the routing process, since they are not going to be
4256 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4257 It corrects a timeout issue with spamd. This is Ian's comment: "The
4258 background is that sometimes spamd either never reads data from a
4259 connection it has accepted, or it never writes response data. The exiscan
4260 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4261 blindly assumes that writes won't block so it may never time out."
4263 PH/39 Allow G after quota size as well as K and M.
4265 PH/40 The value set for $authenticated_id in an authenticator may not contain
4266 binary zeroes or newlines because the value is written to log lines and
4267 to spool files. There was no check on this. Now the value is run through
4268 the string_printing() function so that such characters are converted to
4269 printable escape sequences.
4271 PH/41 $message_linecount is a new variable that contains the total number of
4272 lines in the message. Compare $body_linecount, which is the count for the
4275 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4276 bounce and delay warning messages, except in certain special cases, which
4279 (a) An SMTP error message from a remote host;
4280 (b) A message specified in a :fail: redirection;
4281 (c) A message specified in a "fail" command in a system filter;
4282 (d) A message specified in a FAIL return from the queryprogram router;
4283 (e) A message specified by the cannot_route_message router option.
4285 In these cases only, Exim does include the error details in bounce and
4286 warning messages. There are also a few cases where bland messages such
4287 as "unrouteable address" or "local delivery error" are given.
4289 PH/43 $value is now also set for the "else" part of a ${run expansion.
4291 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4292 being worked on, but at least Exim now implements the latest version to
4295 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4296 process to complete was treated as a delivery failure, a timeout while
4297 writing the message to the pipe was logged, but erroneously treated as a
4298 successful delivery. Such timeouts include transport filter timeouts. For
4299 consistency with the overall process timeout, these timeouts are now
4300 treated as errors, giving rise to delivery failures by default. However,
4301 there is now a new Boolean option for the pipe transport called
4302 timeout_defer, which, if set TRUE, converts the failures into defers for
4303 both kinds of timeout. A transport filter timeout is now identified in
4306 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4307 systems where "make" and "gmake" are different, calling "gmake" at top
4308 level broke things. I've arranged for the value of $(MAKE) to be passed
4309 from the Makefile to this script so that it can call the same version of
4313 A note about Exim versions 4.44 and 4.50
4314 ----------------------------------------
4316 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4317 changes of various kinds. As a consequence, a big documentation update was
4318 needed. This delayed the release for rather longer than seemed good, especially
4319 in the light of a couple of (minor) security issues. Therefore, the changes
4320 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4321 release. So 4.44 and 4.50 are in effect two different branches that both start
4324 I have left the 4.50 change log unchanged; it contains all the changes since
4325 4.43. The change log for 4.44 is below; many of its items are identical to
4326 those for 4.50. This seems to be the most sensible way to preserve the
4327 historical information.
4333 1. Minor wording change to the doc/README.SIEVE file.
4335 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4336 computation of the current number of files was incorrect.
4338 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4339 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4342 4. Give more explanation in the error message when the command for a transport
4343 filter fails to execute.
4345 5. There are several places where Exim runs a non-Exim command in a
4346 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4347 was being done only for the command run by the queryprogram router. It is
4348 now done for all such subprocesses. The other cases are: ${run, transport
4349 filters, and the commands run by the lmtp and pipe transports.
4351 6. Added CONFIGURE_GROUP build-time option.
4353 7. Some older OS have a limit of 256 on the maximum number of file
4354 descriptors. Exim was using setrlimit() to set 1000 as a large value
4355 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4356 systems. I've change it so that if it can't get 1000, it tries for 256.
4358 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4359 was an oversight, and furthermore, ever since the addition of extra
4360 controls (e.g. 4.43/32), the checks on when to allow different forms of
4361 "control" were broken. There should now be diagnostics for all cases when a
4362 control that does not make sense is encountered.
4364 9. Added the /retain_sender option to "control=submission".
4366 10. $recipients is now available in the predata ACL (oversight).
4368 11. Tidy the search cache before the fork to do a delivery from a message
4369 received from the command line. Otherwise the child will trigger a lookup
4370 failure and thereby defer the delivery if it tries to use (for example) a
4371 cached ldap connection that the parent has called unbind on.
4373 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4374 of $address_data from the recipient verification was clobbered by the
4375 sender verification.
4377 13. The value of address_data from a sender verification is now available in
4378 $sender_address_data in subsequent conditions in the ACL statement.
4380 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4382 15. Added a new option "connect=<time>" to callout options, to set a different
4385 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4386 was its contents. (It was OK if the option was not defined at all.)
4388 17. A "Completed" log line is now written for messages that are removed from
4389 the spool by the -Mrm option.
4391 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4392 information about exactly what failed.
4394 19. Added -dd to debug only the daemon process.
4396 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4397 handles timeouts, both on the server side and network timeouts. Renamed the
4398 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4400 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4401 It is now set to "smtps".
4403 22. $host_address is now set to the target address during the checking of
4404 ignore_target_hosts.
4406 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4407 being passed; this would have caused $sender_host_name to have been used if
4408 matching the list had actually called for a host name (not very likely,
4409 since this list is usually IP addresses). A host name is now passed as
4412 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4413 code that specifies a non-restarting handler (typically sigaction() in
4414 modern systems) in an attempt to fix a rare and obscure crash bug.
4416 25. Narrowed the window for a race in the daemon that could cause it to ignore
4417 SIGCHLD signals. This is not a major problem, because they are used only to
4418 wake it up if nothing else does.
4420 26. A malformed maildirsize file could cause Exim to calculate negative values
4421 for the mailbox size or file count. Odd effects could occur as a result.
4422 The maildirsize information is now recalculated if the size or filecount
4425 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4426 support for a long time. Removed HAVE_SYS_VFS_H.
4428 28. Installed the latest version of exipick from John Jetmore.
4430 29. In an address list, if the pattern was not a regular expression, an empty
4431 subject address (from a bounce message) matched only if the pattern was an
4432 empty string. Non-empty patterns were not even tested. This was the wrong
4433 because it is perfectly reasonable to use an empty address as part of a
4434 database query. An empty address is now tested by patterns that are
4435 lookups. However, all the other forms of pattern expect the subject to
4436 contain a local part and a domain, and therefore, for them, an empty
4437 address still always fails if the pattern is not itself empty.
4439 30. Exim went into a mad DNS loop when attempting to do a callout where the
4440 host was specified on an smtp transport, and looking it up yielded more
4441 than one IP address.
4443 31. Re-factored the code for checking spool and log partition space into a
4444 function that finds that data and another that does the check. The former
4445 is then used to implement four new variables: $spool_space, $log_space,
4446 $spool_inodes, and $log_inodes.
4448 32. The RFC2047 encoding function was originally intended for short strings
4449 such as real names; it was not keeping to the 75-character limit for
4450 encoded words that the RFC imposes. It now respects the limit, and
4451 generates multiple encoded words if necessary. To be on the safe side, I
4452 have increased the buffer size for the ${rfc2047: expansion operator from
4455 33. It is now permitted to omit both strings after an "if" condition; if the
4456 condition is true, the result is "true". As before, when the second string
4457 is omitted, a false condition yields an empty string. This makes it less
4458 cumbersome to write custom ACL and router conditions.
4460 34. Failure to deliver a bounce message always caused it to be frozen, even if
4461 there was an errors_to setting on the router. The errors_to setting is now
4464 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4465 canonical form (fully expanded) before being placed in
4466 $sender_host_address.
4468 36. The table in the code that translates DNS record types into text (T_A to
4469 "A" for instance) was missing entries for NS and CNAME. It is just possible
4470 that this could have caused confusion if both these types were looked up
4471 for the same domain, because the text type is used as part of Exim's
4472 per-process caching. But the chance of anyone hitting this buglet seems
4475 37. The dnsdb lookup has been extended in a number of ways.
4477 (1) There is a new type, "zns", which walks up the domain tree until it
4478 finds some nameserver records. It should be used with care.
4480 (2) There is a new type, "mxh", which is like "mx" except that it returns
4481 just the host names, not the priorities.
4483 (3) It is now possible to give a list of domains (or IP addresses) to be
4484 looked up. The behaviour when one of the lookups defers can be
4485 controlled by a keyword.
4487 (4) It is now possible to specify the separator character for use when
4488 multiple records are returned.
4490 38. The dnslists ACL condition has been extended: it is now possible to supply
4491 a list of IP addresses and/or domains to be looked up in a particular DNS
4494 39. Added log_selector=+queue_time_overall.
4496 40. When running the queue in the test harness, wait just a tad after forking a
4497 delivery process, to get repeatability of debugging output.
4499 41. Include certificate and key file names in error message when GnuTLS fails
4500 to set them up, because the GnuTLS error message doesn't include the name
4501 of the failing file when there is a problem reading it.
4503 42. Allow both -bf and -bF in the same test run.
4505 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4507 44. The "Exiscan patch" is now merged into the mainline Exim source.
4509 45. Sometimes the final signoff response after QUIT could fail to get
4510 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4511 < 0 before doing a fflush(). This bug looks as though it goes back to the
4512 introduction of TLS in release 3.20, but "sometimes" must have been rare
4513 because the tests only now provoked it.
4515 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4516 (this can affect the format of dates).
4518 47. exim_tidydb, when checking for the continued existence of a message for
4519 which it has found a message-specific retry record, was not finding
4520 messages that were in split spool directories. Consequently, it was
4521 deleting retry records that should have stayed in existence.
4523 48. Steve fixed some bugs in eximstats.
4525 49. The SPA authentication driver was not abandoning authentication and moving
4526 on to the next authenticator when an expansion was forced to fail,
4527 contradicting the general specification for all authenticators. Instead it
4528 was generating a temporary error. It now behaves as specified.
4530 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4531 (the order specifies the preference for clients). The order is now AES256,
4532 AES128, 3DES, ARCFOUR128.
4534 51. Small patch to Sieve code - explicitly set From: when generating an
4537 52. Exim crashed if a remote delivery caused a very long error message to be
4538 recorded - for instance if somebody sent an entire SpamAssassin report back
4539 as a large number of 550 error lines. This bug was coincidentally fixed by
4540 increasing the size of one of Exim's internal buffers (big_buffer) that
4541 happened as part of the Exiscan merge. However, to be on the safe side, I
4542 have made the code more robust (and fixed the comments that describe what
4545 53. Now that there can be additional text after "Completed" in log lines (if
4546 the queue_time_overall log selector is set), a one-byte patch to exigrep
4547 was needed to allow it to recognize "Completed" as not the last thing in
4550 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4551 patch that reportedly fixes this has been added. I am not expert enough to
4552 create a test for it. This is what the patch creator wrote:
4554 "I found a little strange behaviour of ldap code when working with
4555 Windows 2003 AD Domain, where users was placed in more than one
4556 Organization Units. When I tried to give exim partial DN, the exit code
4557 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4558 But simultaneously result of request was absolutely normal ldap result,
4559 so I produce this patch..."
4561 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4562 REFERENCE, so I have modified the code to exclude the patch when that macro
4565 55. Some experimental protocols are using DNS PTR records for new purposes. The
4566 keys for these records are domain names, not reversed IP addresses. The
4567 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4568 leaves it alone. Component reversal etc. now happens only for IP addresses.
4571 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4573 57. Double the size of the debug message buffer (to 2048) so that more of very
4574 long debug lines gets shown.
4576 58. The exicyclog utility now does better if the number of log files to keep
4577 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4579 59. Two changes related to the smtp_active_hostname option:
4581 (1) $smtp_active_hostname is now available as a variable.
4582 (2) The default for smtp_banner uses $smtp_active_hostname instead
4583 of $primary_hostname.
4585 60. The host_aton() function is supposed to be passed a string that is known
4586 to be a valid IP address. However, in the case of IPv6 addresses, it was
4587 not checking this. This is a hostage to fortune. Exim now panics and dies
4588 if the condition is not met. A case was found where this could be provoked
4589 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4590 components; fortuitously, this particular loophole had already been fixed
4591 by change 4.50/55 above.
4593 If there are any other similar loopholes, the new check in host_aton()
4594 itself should stop them being exploited. The report I received stated that
4595 data on the command line could provoke the exploit when Exim was running as
4596 exim, but did not say which command line option was involved. All I could
4597 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4598 running as the user.
4601 61. There was a buffer overflow vulnerability in the SPA authentication code
4602 (which came originally from the Samba project). I have added a test to the
4603 spa_base64_to_bits() function which I hope fixes it.
4606 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4607 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4609 63. The daemon start-up calls getloadavg() while still root for those OS that
4610 need the first call to be done as root, but it missed one case: when
4611 deliver_queue_load_max is set with deliver_drop_privilege. This is
4612 necessary for the benefit of the queue runner, because there is no re-exec
4613 when deliver_drop_privilege is set.
4615 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4616 This has been fixed.
4618 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4619 was in use, was not putting the data itself into the right store pool;
4620 consequently, it could be overwritten for a subsequent message in the same
4621 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4624 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4626 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4627 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4628 the function were treating the return as a boolean value, which happened to
4629 work because 0=false and not-0=true, but is not correct code.
4631 68. The host_aton() function was not handling scoped IPv6 addresses (those
4632 with, for example, "%eth0" on the end) correctly.
4634 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4635 negated items (that is, ~something) in unsigned ints. Some compilers
4636 apparently mutter when there is no cast.
4638 70. If an address verification called from an ACL failed, and did not produce a
4639 user-specific message (i.e. there was only a "system" message), nothing was
4640 put in $acl_verify_message. In this situation, it now puts the system
4643 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4644 session; change 4.31/43 added the unwanted input to the log line - except
4645 that it did not do this in the start of session case. It now does.
4647 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4648 This is wrong and can cause the other end to generate a synchronization
4649 error if it is another Exim or anything else that does the synchronization
4650 check. A QUIT command is no longer sent after a timeout.
4652 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4653 during host lookups.
4655 74. The defer_ok option of callout verification was not working if it was used
4656 when verifying addresses in header lines, that is, for this case:
4658 verify = header_sender/callout=defer_ok
4660 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4661 those file descriptors could be used for SMTP connections. If anything
4662 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4663 could be sent to the SMTP client, causing chaos. The daemon now opens
4664 stdin, stdout, and stderr to /dev/null when it puts itself into the
4667 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4668 log by default. The user can override this with suitable Perl magic.
4670 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4671 the log message when discard triggers, was not working for the DATA ACL or
4672 for the non-SMTP ACL.
4674 78. Error message wording change in sieve.c.
4676 79. If smtp_accept_max_per_host was set, the number of connections could be
4677 restricted to fewer than expected, because the daemon was trying to set up
4678 a new connection before checking whether the processes handling previous
4679 connections had finished. The check for completed processes is now done
4680 earlier. On busy systems, this bug wouldn't be noticed because something
4681 else would have woken the daemon, and it would have reaped the completed
4684 80. If a message was submitted locally by a user whose login name contained one
4685 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4686 It caused a spool format error. I have fixed the spool reading code. A
4687 related problem was that the "from" clause in the Received: line became
4688 illegal because of the space(s). It is now covered by ${quote_local_part.
4690 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4693 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4695 83. Patch from Sieve maintainer to fix unterminated string problem in
4696 "vacation" handling.
4698 84. Some minor changes to the Linux configuration files to help with other
4699 OS variants using glibc.
4701 85. One more patch for Sieve to update vacation handling to latest spec.
4704 ----------------------------------------------------
4705 See the note above about the 4.44 and 4.50 releases.
4706 ----------------------------------------------------
4712 1. Change 4.43/35 introduced a bug that caused file counts to be
4713 incorrectly computed when quota_filecount was set in an appendfile
4716 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4717 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4720 3. Give more explanation in the error message when the command for a transport
4721 filter fails to execute.
4723 4. There are several places where Exim runs a non-Exim command in a
4724 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4725 was being done only for the command run by the queryprogram router. It is
4726 now done for all such subprocesses. The other cases are: ${run, transport
4727 filters, and the commands run by the lmtp and pipe transports.
4729 5. Some older OS have a limit of 256 on the maximum number of file
4730 descriptors. Exim was using setrlimit() to set 1000 as a large value
4731 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4732 systems. I've change it so that if it can't get 1000, it tries for 256.
4734 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4735 was an oversight, and furthermore, ever since the addition of extra
4736 controls (e.g. 4.43/32), the checks on when to allow different forms of
4737 "control" were broken. There should now be diagnostics for all cases when a
4738 control that does not make sense is encountered.
4740 7. $recipients is now available in the predata ACL (oversight).
4742 8. Tidy the search cache before the fork to do a delivery from a message
4743 received from the command line. Otherwise the child will trigger a lookup
4744 failure and thereby defer the delivery if it tries to use (for example) a
4745 cached ldap connection that the parent has called unbind on.
4747 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4748 of $address_data from the recipient verification was clobbered by the
4749 sender verification.
4751 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4752 was its contents. (It was OK if the option was not defined at all.)
4754 11. A "Completed" log line is now written for messages that are removed from
4755 the spool by the -Mrm option.
4757 12. $host_address is now set to the target address during the checking of
4758 ignore_target_hosts.
4760 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4761 being passed; this would have caused $sender_host_name to have been used if
4762 matching the list had actually called for a host name (not very likely,
4763 since this list is usually IP addresses). A host name is now passed as
4766 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4767 code that specifies a non-restarting handler (typically sigaction() in
4768 modern systems) in an attempt to fix a rare and obscure crash bug.
4770 15. Narrowed the window for a race in the daemon that could cause it to ignore
4771 SIGCHLD signals. This is not a major problem, because they are used only to
4772 wake it up if nothing else does.
4774 16. A malformed maildirsize file could cause Exim to calculate negative values
4775 for the mailbox size or file count. Odd effects could occur as a result.
4776 The maildirsize information is now recalculated if the size or filecount
4779 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4780 support for a long time. Removed HAVE_SYS_VFS_H.
4782 18. Updated exipick to current release from John Jetmore.
4784 19. Allow an empty sender to be matched against a lookup in an address list.
4785 Previously the only cases considered were a regular expression, or an
4788 20. Exim went into a mad DNS lookup loop when doing a callout where the
4789 host was specified on the transport, if the DNS lookup yielded more than
4792 21. The RFC2047 encoding function was originally intended for short strings
4793 such as real names; it was not keeping to the 75-character limit for
4794 encoded words that the RFC imposes. It now respects the limit, and
4795 generates multiple encoded words if necessary. To be on the safe side, I
4796 have increased the buffer size for the ${rfc2047: expansion operator from
4799 22. Failure to deliver a bounce message always caused it to be frozen, even if
4800 there was an errors_to setting on the router. The errors_to setting is now
4803 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4804 canonical form (fully expanded) before being placed in
4805 $sender_host_address.
4807 24. Updated eximstats to version 1.33
4809 25. Include certificate and key file names in error message when GnuTLS fails
4810 to set them up, because the GnuTLS error message doesn't include the name
4811 of the failing file when there is a problem reading it.
4813 26. Expand error message when OpenSSL has problems setting up cert/key files.
4816 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4817 (this can affect the format of dates).
4819 28. exim_tidydb, when checking for the continued existence of a message for
4820 which it has found a message-specific retry record, was not finding
4821 messages that were in split spool directories. Consequently, it was
4822 deleting retry records that should have stayed in existence.
4824 29. eximstats updated to version 1.35
4825 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4826 1.35 - bugfix such that pie charts by volume are generated correctly
4828 30. The SPA authentication driver was not abandoning authentication and moving
4829 on to the next authenticator when an expansion was forced to fail,
4830 contradicting the general specification for all authenticators. Instead it
4831 was generating a temporary error. It now behaves as specified.
4833 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4834 (the order specifies the preference for clients). The order is now AES256,
4835 AES128, 3DES, ARCFOUR128.
4837 31. Small patch to Sieve code - explicitly set From: when generating an
4840 32. Exim crashed if a remote delivery caused a very long error message to be
4841 recorded - for instance if somebody sent an entire SpamAssassin report back
4842 as a large number of 550 error lines. This bug was coincidentally fixed by
4843 increasing the size of one of Exim's internal buffers (big_buffer) that
4844 happened as part of the Exiscan merge. However, to be on the safe side, I
4845 have made the code more robust (and fixed the comments that describe what
4848 33. Some experimental protocols are using DNS PTR records for new purposes. The
4849 keys for these records are domain names, not reversed IP addresses. The
4850 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4851 leaves it alone. Component reversal etc. now happens only for IP addresses.
4854 34. The host_aton() function is supposed to be passed a string that is known
4855 to be a valid IP address. However, in the case of IPv6 addresses, it was
4856 not checking this. This is a hostage to fortune. Exim now panics and dies
4857 if the condition is not met. A case was found where this could be provoked
4858 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4859 components; fortuitously, this particular loophole had already been fixed
4860 by change 4.50/55 or 4.44/33 above.
4862 If there are any other similar loopholes, the new check in host_aton()
4863 itself should stop them being exploited. The report I received stated that
4864 data on the command line could provoke the exploit when Exim was running as
4865 exim, but did not say which command line option was involved. All I could
4866 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4867 running as the user.
4870 35. There was a buffer overflow vulnerability in the SPA authentication code
4871 (which came originally from the Samba project). I have added a test to the
4872 spa_base64_to_bits() function which I hope fixes it.
4875 36. The daemon start-up calls getloadavg() while still root for those OS that
4876 need the first call to be done as root, but it missed one case: when
4877 deliver_queue_load_max is set with deliver_drop_privilege. This is
4878 necessary for the benefit of the queue runner, because there is no re-exec
4879 when deliver_drop_privilege is set.
4881 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4882 was in use, was not putting the data itself into the right store pool;
4883 consequently, it could be overwritten for a subsequent message in the same
4884 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4887 38. Sometimes the final signoff response after QUIT could fail to get
4888 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4889 < 0 before doing a fflush(). This bug looks as though it goes back to the
4890 introduction of TLS in release 3.20, but "sometimes" must have been rare
4891 because the tests only now provoked it.
4897 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4898 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4899 Now it returns an integer. A number of calls were still expecting a T/F
4900 return. Fortuitously, in all cases, the tests worked in OK situations,
4901 which is the norm. However, things would have gone wrong on any write
4902 failures on the smtp file descriptor. This function is used when sending
4903 messages over SMTP and also when doing verify callouts.
4905 2. When Exim is called to do synchronous delivery of a locally submitted
4906 message (the -odf or -odi options), it no longer closes stderr before doing
4909 3. Implemented the mua_wrapper option.
4911 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4913 5. Implemented the functions header_remove(), header_testname(),
4914 header_add_at_position(), and receive_remove_recipient(), and exported them
4917 6. If an ACL "warn" statement specified the addition of headers, Exim already
4918 inserted X-ACL-Warn: at the start if there was no header name. However, it
4919 was not making this test for the second and subsequent header lines if
4920 there were newlines in the string. This meant that an invalid header could
4921 be inserted if Exim was badly configured.
4923 7. Allow an ACL "warn" statement to add header lines at the start or after all
4924 the Received: headers, as well as at the end.
4926 8. Added the rcpt_4xx retry error code.
4928 9. Added postmaster_mailfrom=xxx to callout verification option.
4930 10. Added mailfrom=xxxx to the callout verification option, for verify=
4933 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4934 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4936 12. Inserted some casts to stop certain compilers warning when using pointer
4937 differences as field lengths or precisions in printf-type calls (mostly
4938 affecting debugging statements).
4940 13. Added optional readline() support for -be (dynamically loaded).
4942 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4943 same clock tick as a message's arrival, so that its received time was the
4944 same as the "first fail" time on the retry record, and that message
4945 remained on the queue past the ultimate address timeout, every queue runner
4946 would try a delivery (because it was past the ultimate address timeout) but
4947 after another failure, the ultimate address timeout, which should have then
4948 bounced the address, did not kick in. This was a "< instead of <=" error;
4949 in most cases the first failure would have been in the next clock tick
4950 after the received time, and all would be well.
4952 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4953 being recognized when the domain list was tested by the match_domain
4954 condition in an expansion string.
4956 16. Added the ${str2b64: operator.
4958 17. Exim was always calling setrlimit() to set a large limit for the number of
4959 processes, without checking whether the existing limit was already
4960 adequate. (It did check for the limit on file descriptors.) Furthermore,
4961 errors from getrlimit() and setrlimit() were being ignored. Now they are
4962 logged to the main and panic logs, but Exim does carry on, to try to do its
4963 job under whatever limits there are.
4965 18. Imported PCRE 5.0.
4967 19. Trivial typo in log message " temporarily refused connection" (the leading
4970 20. If the log selector return_path_on_delivery was set and an address was
4971 redirected to /dev/null, the delivery process crashed because it assumed
4972 that a return path would always be set for a "successful" delivery. In this
4973 case, the whole delivery is bypassed as an optimization, and therefore no
4976 21. Internal re-arrangement: the function for sending a challenge and reading
4977 a response while authentication was assuming a zero-terminated challenge
4978 string. It's now changed to take a pointer and a length, to allow for
4979 binary data in such strings.
4981 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4983 23. Exim was not respecting finduser_retries when seeking the login of the
4984 uid under which it was called; it was always trying 10 times. (The default
4985 setting of finduser_retries is zero.) Also, it was sleeping after the final
4986 failure, which is pointless.
4988 24. Implemented tls_on_connect_ports.
4990 25. Implemented acl_smtp_predata.
4992 26. If the domain in control=submission is set empty, Exim assumes that the
4993 authenticated id is a complete email address when it generates From: or
4994 Sender: header lines.
4996 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4997 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4998 chown and chgrp in /bin and hostname in /usr/bin.
5000 28. Exim was keeping the "process log" file open after each use, just as it
5001 does for the main log. This opens the possibility of it remaining open for
5002 long periods when the USR1 signal hits a daemon. Occasional processlog
5003 errors were reported, that could have been caused by this. Anyway, it seems
5004 much more sensible not to leave this file open at all, so that is what now
5007 29. The long-running daemon process does not normally write to the log once it
5008 has entered its main loop, and it closes the log before doing so. This is
5009 so that log files can straightforwardly be renamed and moved. However,
5010 there are a couple of unusual error situations where the daemon does write
5011 log entries, and I had neglected to close the log afterwards.
5013 30. The text of an SMTP error response that was received during a remote
5014 delivery was being truncated at 512 bytes. This is too short for some of
5015 the long messages that one sometimes sees. I've increased the limit to
5018 31. It is now possible to make retry rules that apply only when a message has a
5019 specific sender, in particular, an empty sender.
5021 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5022 it possible to be selective about when SMTP synchronization is enforced.
5024 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5026 32. Implemented hosts_connection_nolog.
5028 33. Added an ACL for QUIT.
5030 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5033 35. Added mailbox_size and mailbox_filecount to appendfile.
5035 36. Added control = no_multiline_responses to ACLs.
5037 37. There was a bug in the logic of the code that waits for the clock to tick
5038 in the case where the clock went backwards by a substantial amount such
5039 that the microsecond fraction of "now" was more than the microsecond
5040 fraction of "then" (but the whole seconds number was less).
5042 38. Added support for the libradius Radius client library this is found on
5043 FreeBSD (previously only the radiusclient library was supported).
5049 1. When certain lookups returned multiple values in the form name=value, the
5050 quoting of the values was not always being done properly. Specifically:
5051 (a) If the value started with a double quote, but contained no whitespace,
5053 (b) If the value contained whitespace other than a space character (i.e.
5054 tabs or newlines or carriage returns) it was not quoted.
5055 This fix has been applied to the mysql and pgsql lookups by writing a
5056 separate quoting function and calling it from the lookup code. The fix
5057 should probably also be applied to nisplus, ibase and oracle lookups, but
5058 since I cannot test any of those, I have not disturbed their existing code.
5060 2. A hit in the callout cache for a specific address caused a log line with no
5061 reason for rejecting RCPT. Now it says "Previous (cached) callout
5062 verification failure".
5064 3. There was an off-by-one bug in the queryprogram router. An over-long
5065 return line was truncated at 256 instead of 255 characters, thereby
5066 overflowing its buffer with the terminating zero. As well as fixing this, I
5067 have increased the buffer size to 1024 (and made a note to document this).
5069 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5070 when Exim is waiting for an SMTP response from a remote server, Exim
5071 restarts its select() call on the socket, thereby resetting its timeout.
5072 This is not a problem when such interrupts are rare. Somebody set up a cron
5073 job to run exiwhat every 2 minutes, which is less than the normal select()
5074 timeout (5 or 10 minutes). This meant that the select() timeout never
5075 kicked in because it was always reset. I have fixed this by comparing the
5076 time when an interrupt arrives with the time at the start of the first call
5077 to select(). If more time than the timeout has elapsed, the interrupt is
5078 treated as a timeout.
5080 5. Some internal re-factoring in preparation for the addition of Sieve
5081 extensions (by MH). In particular, the "personal" test is moved to a
5082 separate function, and given an option for scanning Cc: and Bcc: (which is
5083 not set for Exim filters).
5085 6. When Exim created an email address using the login of the caller as the
5086 local part (e.g. when creating a From: or Sender: header line), it was not
5087 quoting the local part when it contained special characters such as @.
5089 7. Installed new OpenBSD configuration files.
5091 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5092 try to make them clearer.
5094 9. Callout options, other than the timeout value, were being ignored when
5095 verifying sender addresses in header lines. For example, when using
5097 verify = header_sender/callout=no_cache
5099 the cache was (incorrectly) being used.
5101 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5102 only the Cygwin environment.
5104 11. When return_path_on_delivery was set as a log selector, if different remote
5105 addresses in the same message used different return paths and parallel
5106 remote delivery occurred, the wrong values would sometimes be logged.
5107 (Whenever a remote delivery process finished, the return path value from
5108 the most recently started remote delivery process was logged.)
5110 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5111 lines when AUTH and/or TLS are in use. This is the "received protocol"
5112 field. Exim used to use "asmtp" for authenticated SMTP, without any
5113 indication (in the protocol name) for TLS use. Now it follows the RFC and
5114 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5115 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5116 names appear in log lines as well as in Received: header lines.
5118 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5119 extensions, and comparison tests, and to fix some bugs.
5121 14. Changes to the "personal" filter test:
5123 (1) The test was buggy in that it was just doing the equivalent of
5124 "contains" tests on header lines. For example, if a user's address was
5125 anne@some.where, the "personal" test would incorrectly be true for
5127 To: susanne@some.where
5129 This test is now done by extracting each address from the header in turn,
5130 and checking the entire address. Other tests that are part of "personal"
5131 are now done using regular expressions (for example, to check local parts
5132 of addresses in From: header lines).
5134 (2) The list of non-personal local parts in From: addresses has been
5135 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5136 taken from the Sieve specification recommendations.
5138 (3) If the message contains any header line starting with "List-" it is
5139 treated as non-personal.
5141 (4) The test for "circular" in the Subject: header line has been removed
5142 because it now seems ill-conceived.
5144 15. Minor typos in src/EDITME comments corrected.
5146 16. Installed latest exipick from John Jetmore.
5148 17. If headers_add on a router specified a text string that was too long for
5149 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5150 of string_sprintf() is now avoided.
5152 18. $message_body_size was not set (it was always zero) when running the DATA
5153 ACL and the local_scan() function.
5155 19. For the "mail" command in an Exim filter, no default was being set for
5156 the once_repeat time, causing a random time value to be used if "once" was
5157 specified. (If the value happened to be <= 0, no repeat happened.) The
5158 default is now 0s, meaning "never repeat". The "vacation" command was OK
5159 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5160 (I found it when inspecting the code).
5162 20. There is now an overall timeout for performing a callout verification. It
5163 defaults to 4 times the callout timeout, which applies to individual SMTP
5164 commands during the callout. The overall timeout applies when there is more
5165 than one host that can be tried. The timeout is checked before trying the
5166 next host. This prevents very long delays if there are a large number of
5167 hosts and all are timing out (e.g. when the network connections are timing
5168 out). The value of the overall timeout can be changed by specifying an
5169 additional sub-option for "callout", called "maxwait". For example:
5171 verify = sender/callout=5s,maxwait=20s
5173 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5174 to the end before writing, but this should make it even safer).
5176 22. Exim was forgetting that it had advertised PIPELINING for the second and
5177 subsequent messages on an SMTP connection. It was also not resetting its
5178 memory on STARTTLS and an internal HELO.
5180 23. When Exim logs an SMTP synchronization error within a session, it now
5181 records whether PIPELINING has been advertised or not.
5183 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5184 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5185 rather than long int.
5187 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5189 26. Added the never_mail option to autoreply.
5195 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5196 crash if the getsockname() call failed; this can happen if a connection is
5197 closed very soon after it is established. The problem was simply in the
5198 order in which certain operations were done, causing Exim to try to write
5199 to the SMTP stream before it had set up the file descriptor. The bug has
5200 been fixed by making things happen in the correct order.
5206 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5207 before the connection was closed, thus losing the rejection response.
5209 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5210 some early Solaris releases, but causes trouble in current releases where
5211 socklen_t is defined.
5213 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5216 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5219 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5220 that releasing the top of it at the end releases what was used for sub-
5221 expansions (unless the block got too big). However, discard this block if
5222 the first thing is a variable or header, so that we can use its block when
5223 it is dynamic (useful for very large $message_headers, for example).
5225 6. Lookups now cache *every* query, not just the most recent. A new, separate
5226 store pool is used for this. It can be recovered when all lookup caches are
5227 flushed. Lookups now release memory at the end of their result strings.
5228 This has involved some general refactoring of the lookup sources.
5230 7. Some code has been added to the store_xxx() functions to reduce the amount
5231 of flapping under certain conditions.
5233 8. log_incoming_interface used to affect only the <= reception log lines. Now
5234 it causes the local interface and port to be added to several more SMTP log
5235 lines, for example "SMTP connection from", and rejection lines.
5237 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5239 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5241 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5242 could be overwritten at the end of the current message (or the start of a
5243 new message if it was set in a HELO ACL). The value is now preserved for
5244 the duration of the SMTP connection.
5246 12. If a transport had a headers_rewrite setting, and a matching header line
5247 contained an unqualified address, that address was qualified, even if it
5248 did not match any rewriting rules. The underlying bug was that the values
5249 of the flags that permit the existence of unqualified sender and recipient
5250 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5251 non-local messages, and by -bnq for local messages) were not being
5252 preserved with the message after it was received.
5254 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5255 "next input=" as part of the text comprising the host identity instead of
5256 the correct text. The code was using the same buffer for two different
5257 strings. However, depending on which order the printing function evaluated
5258 its arguments, the bug did not always show up. Under Linux, for example, my
5259 test suite worked just fine.
5261 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5262 doesn't work with some older versions of Perl. It has been changed to "my",
5263 which in any case is probably the better facility to use.
5265 15. A really picky compiler found some instances of statements for creating
5266 error messages that either had too many or two few arguments for the format
5269 16. The size of the buffer for calls to the DNS resolver has been increased
5270 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5271 for addresses that have a lot of PTR records. This alleviates a problem; it
5272 does not fully solve it.
5274 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5275 buffer now truncates the list and logs the incident, which is the same
5276 action as happens when Exim is looking up a host name and its aliases.
5277 Previously in this situation something unpredictable would happen;
5278 sometimes it was "internal error: store_reset failed".
5280 18. If a server dropped the connection unexpectedly when an Exim client was
5281 using GnuTLS and trying to read a response, the client delivery process
5282 crashed while trying to generate an error log message.
5284 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5285 string, for example:
5287 warn message = H1: something\nH2: something
5289 the text was added as a single header line from Exim's point of view
5290 though it ended up OK in the delivered message. However, searching for the
5291 second and subsequent header lines using $h_h2: did not work. This has been
5292 fixed. Similarly, if a system filter added multiple headers in this way,
5293 the routers could not see them.
5295 20. Expanded the error message when iplsearch is called with an invalid key to
5296 suggest using net-iplsearch in a host list.
5298 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5299 ACLs are no longer actually imposed (and a message to that effect is
5302 22. If a "gecos" field in a passwd entry contained escaped characters, in
5303 particular, if it contained a \" sequence, Exim got it wrong when building
5304 a From: or a Sender: header from that name. A second bug also caused
5305 incorrect handling when an unquoted " was present following a character
5306 that needed quoting.
5308 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5309 was not being matched caselessly.
5311 24. Arranged for all hyphens in the exim.8 source to be escaped with
5314 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5315 properly. Recipient callout cache records were still being keyed to include
5316 the sender, even when use_sender was set false. This led to far more
5317 callouts that were necessary. The sender is no longer included in the key
5318 when use_sender is false.
5320 26. Added "control = submission" modifier to ACLs.
5322 27. Added the ${base62d: operator to decode base 62 numbers.
5324 28. dnsdb lookups can now access SRV records.
5326 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5327 the configuration file.
5329 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5330 (-v) mode. This makes the output for a verbose queue run more intelligible.
5332 31. Added a use_postmaster feature to recipient callouts.
5334 32. Added the $body_zerocount variable, containing the number of binary zero
5335 bytes in the message body.
5337 33. The time of last modification of the "new" subdirectory is now used as the
5338 "mailbox time last read" when there is a quota error for a maildir
5341 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5343 35. Added +ignore_unknown as a special item in host lists.
5345 36. Code for decoding IPv6 addresses in host lists is now included, even if
5346 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5347 address was recognized as an IP address, but was then not correctly decoded
5348 into binary, causing unexpected and incorrect effects when compared with
5355 1. Very minor rewording of debugging text in manualroute to say "list of
5356 hosts" instead of "hostlist".
5358 2. If verify=header_syntax was set, and a header line with an unqualified
5359 address (no domain) and a large number of spaces between the end of the
5360 name and the colon was received, the reception process suffered a buffer
5361 overflow, and (when I tested it) crashed. This was caused by some obsolete
5362 code that should have been removed. The fix is to remove it!
5364 3. When running in the test harness, delay a bit after writing a bounce
5365 message to get a bit more predictability in the log output.
5367 4. Added a call to search_tidyup() just before forking a reception process. In
5368 theory, someone could use a lookup in the expansion of smtp_accept_max_
5369 per_host which, without the tidyup, could leave open a database connection.
5371 5. Added the variables $recipient_data and $sender_data which get set from a
5372 lookup success in an ACL "recipients" or "senders" condition, or a router
5373 "senders" option, similar to $domain_data and $local_part_data.
5375 6. Moved the writing of debug_print from before to after the "senders" test
5378 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5379 problems for message scanning, either using a data ACL, or using
5380 local_scan() because the Received: header was not generated till after they
5381 were called (in order to set the time as the time of reception completion).
5382 I have revised the way this works. The header is now generated after the
5383 body is received, but before the ACL or local_scan() are called. After they
5384 are run, the timestamp in the header is updated.
5390 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5391 before starting a queue runner without re-exec. This happened only when
5392 deliver_drop_privilege was set or when the Exim user was set to root. The
5393 effect of the bug was that timeouts during subsequent deliveries caused
5394 crashes instead of being properly handled. The handler is now left at its
5395 default (and expected) setting.
5397 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5398 message, again when deliver_drop_privilege is set or Exim is run as root.
5399 The bug described in (1) was not present in this case, but the tidying up
5400 of the other signals was missing. I have made the two cases consistent.
5402 3. The ignore_target_hosts setting on a manualroute router was being ignored
5403 for hosts that were looked up using the /MX notation.
5405 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5408 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5409 operated on the sender address. After changing the $sender_address to <>
5410 for the sender address verify, Exim was re-instated it as the original
5411 (before rewriting) address, but remembering that it had rewritten it, so it
5412 wasn't rewriting it again. This bug also had the effect of breaking the
5413 sender address verification caching when the sender address was rewritten.
5415 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5416 This has been changed so that if the ip literal address matches
5417 ignore_target_hosts, the router declines.
5419 7. Added expansion conditions match_domain, match_address, and match_local_
5420 part (NOT match_host).
5422 8. The placeholder for the Received: header didn't have a length field set.
5424 9. Added code to Exim itself and to exim_lock to test for a specific race
5425 condition that could lead to file corruption when using MBX delivery. The
5426 issue is with the lockfile that is created in /tmp. If this file is removed
5427 after a process has opened it but before that process has acquired a lock,
5428 there is the potential for a second process to recreate the file and also
5429 acquire a lock. This could lead to two Exim processes writing to the file
5430 at the same time. The added code performs the same test as UW imapd; it
5431 checks after acquiring the lock that its file descriptor still refers to
5432 the same named file.
5434 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5435 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5438 11. Added the smtp_active_hostname option. If used, this will typically be made
5439 to depend on the incoming interface address. Because $interface_address is
5440 not set up until the daemon has forked a reception process, error responses
5441 that can happen earlier (such as "too many connections") no longer contain
5444 12. If an expansion in a condition on a "warn" statement fails because a lookup
5445 defers, the "warn" statement is abandoned, and the next ACL statement is
5446 processed. Previously this caused the whole ACL to be aborted.
5448 13. Added the iplsearch lookup type.
5450 14. Added ident_timeout as a log selector.
5452 15. Added tls_certificate_verified as a log selector.
5454 16. Added a global option tls_require_ciphers (compare the smtp transport
5455 option of the same name). This controls incoming TLS connections.
5457 17. I finally figured out how to make tls_require_ciphers do a similar thing
5458 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5459 before starting the TLS session.
5461 18. Tabs are now shown as \t in -bP output.
5463 19. If the log selector return_path_on_delivery was set, Exim crashed when
5464 bouncing a message because it had too many Received: header lines.
5466 20. If two routers both had headers_remove settings, and the first one included
5467 a superfluous trailing colon, the final name in the first list and the
5468 first name in the second list were incorrectly joined into one item (with a
5469 colon in the middle).
5475 1. Added -C and -D options to the exinext utility, mainly to make it easier
5476 to include in the automated testing, but these could be helpful when
5477 multiple configurations are in use.
5479 2. The exinext utility was not formatting the output nicely when there was
5480 an alternate port involved in the retry record key, nor when there was a
5481 message id as well (for retries that were specific to a specific message
5482 and a specific host). It was also confused by IPv6 addresses, because of
5483 the additional colons they contain. I have fixed the IPv4 problem, and
5484 patched it up to do a reasonable job for IPv6.
5486 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5487 delivery, the log line now contains "pipelined" if PIPELINING was used.
5489 4. An SMTP transport process used to panic and die if the bind() call to set
5490 an explicit outgoing interface failed. This has been changed; it is now
5491 treated in the same way as a connect() failure.
5493 5. A reference to $sender_host_name in the part of a conditional expansion
5494 that was being skipped was still causing a DNS lookup. This no longer
5497 6. The def: expansion condition was not recognizing references to header lines
5498 that used bh_ and bheader_.
5500 7. Added the _cache feature to named lists.
5502 8. The code for checking quota_filecount in the appendfile transport was
5503 allowing one more file than it should have been.
5505 9. For compatibility with Sendmail, the command line option
5513 and sets the incoming protocol and host name (for trusted callers). The
5514 host name and its colon can be omitted when only the protocol is to be set.
5515 Note the Exim already has two private options, -pd and -ps, that refer to
5516 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5517 "s", but I don't think that's a major issue.
5519 10. A number of refactoring changes to the code, none of which should affect
5522 (a) The number of logging options was getting close to filling up the
5523 32-bit word that was used as a bit map. I have split them into two classes:
5524 those that are passed in the argument to log_write(), and those that are
5525 only ever tested independently outside of that function. These are now in
5526 separate 32-bit words, so there is plenty of room for expansion again.
5527 There is no change in the user interface or the logging behaviour.
5529 (b) When building, for example, log lines, the code previously used a
5530 macro that called string_cat() twice, in order to add two strings. This is
5531 not really sufficiently general. Furthermore, there was one instance where
5532 it was actually wrong because one of the argument was used twice, and in
5533 one call a function was used. (As it happened, calling the function twice
5534 did not affect the overall behaviour.) The macro has been replaced by a
5535 function that can join an arbitrary number of extra strings onto a growing
5538 (c) The code for expansion conditions now uses a table and a binary chop
5539 instead of a serial search (which was left over from when there were very
5540 few conditions). Also, it now recognizes conditions like "pam" even when
5541 the relevant support is not compiled in: a suitably worded error message is
5542 given if an attempt is made to use such a condition.
5544 11. Added ${time_interval:xxxxx}.
5546 12. A bug was causing one of the ddress fields not to be passed back correctly
5547 from remote delivery subprocesses. The field in question was not being
5548 subsequently used, so this caused to problems in practice.
5550 13. Added new log selectors queue_time and deliver_time.
5552 14. Might have fixed a bug in maildirsizefile handling that threw up
5553 "unexpected character" debug warnings, and recalculated the data
5554 unnecessarily. In any case, I expanded the warning message to give more
5557 15. Added the message "Restricted characters in address" to the statements in
5558 the default ACL that block characters like @ and % in local parts.
5560 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5561 Three changes have been made:
5563 (a) There was a serious bug; a negative response to MAIL caused the whole
5564 recipient domain to be cached as invalid, thereby blocking all messages
5565 to all local parts at the same domain, from all senders. This bug has
5566 been fixed. The domain is no longer cached after a negative response to
5567 MAIL if the sender used is not empty.
5569 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5572 (c) A new callout option, "use_sender" has been added for people who want
5573 the modified behaviour.
5579 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5582 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5583 indeed breaks things for older releases.
5585 3. Added additional logging to the case where there is a problem reading data
5586 from a filter that is running in a subprocess using a pipe, in order to
5587 try to track down a specific problem.
5589 4. Testing facility fudge: when running in the test harness and attempting
5590 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5591 getting "No route to host". Convert this to a timeout.
5593 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5596 6. Some OS don't have socklen_t but use size_t instead. This affects the
5597 fifth argument of getsockopt() amongst other things. This is now
5598 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5599 can be set for individual OS. I have set it for SunOS5, OSF1, and
5600 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5601 some earlier ones do not.
5603 7. Change 4.30/15 was not doing the test caselessly.
5605 8. The standard form for an IPv6 address literal was being rejected by address
5606 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5607 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5608 this, as well as the form without the "IPv6" on the front (but only when
5609 address literals are enabled, of course).
5611 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5613 10. Exim crashed if a message with an empty sender address specified by -f
5614 encountered a router with an errors_to setting. This could be provoked only
5615 by a command such as
5619 where an empty string was supplied; "<>" did not hit this bug.
5621 11. Installed PCRE release 4.5.
5623 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5624 remained set. It is now erased.
5626 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5627 times from message ids (which are base 36 rather than the normal 62).
5629 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5630 were being counted as actual protocol errors, and logged if the log
5631 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5632 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5633 and DATA following a set of rejected RCPTs do not count as protocol errors.
5634 In other words, Exim assumes they were pipelined, though this may not
5635 actually be the case. Of course, in all cases the client gets an
5636 appropriate error code.
5638 15. If a lookup fails in an ACL condition, a message about the failure may
5639 be available; it is used if testing the ACL cannot continue, because most
5640 such messages specify what the cause of the deferral is. However, some
5641 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5642 that caused an old message to be retained and used if a later statement
5643 caused a defer, replacing the real cause of the deferral.
5645 16. If an IP address had so many PTR records that the DNS lookup buffer
5646 was not large enough to hold them, Exim could crash while trying to process
5647 the truncated data. It now detects and logs this case.
5649 17. Further to 4.21/58, another change has been made: if (and only if) the
5650 first line of a message (the first header line) ends with CRLF, a bare LF
5651 in a subsequent header line has a space inserted after it, so as not to
5652 terminate the header.
5654 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5655 unnecessarily, used atoi() instead of strtol(), and didn't check the
5656 termination when getting file sizes from file names by regex.
5658 19. Completely re-implemented the support for maildirsize files, in the light
5659 of a number of problems with the previous contributed implementation
5660 (4.30/29). In particular:
5662 . If the quota is zero, the maildirsize file is maintained, but no quota is
5665 . If the maildir directory does not exist, it is created before any attempt
5666 to write a maildirsize file.
5668 . The quota value in the file is just a cache; if the quota is changed in
5669 the transport, the new value overrides.
5671 . A regular expression is available for excluding directories from the
5674 20. The autoreply transport checks the characters in options that define the
5675 message's headers; it allows continued headers, but it was checking with
5676 isspace() after an embedded newline instead of explicitly looking for a
5679 21. If all the "regular" hosts to which an address was routed had passed their
5680 expiry times, and had not reached their retry times, the address was
5681 bounced, even if fallback hosts were defined. Now Exim should go on to try
5684 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5685 equivalent code in the SMTP transport. Some hosts send humungous responses
5686 to HELO/EHLO, more than 1024 it seems.
5688 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5689 gives compiler warnings in some environments. I've now done it "properly",
5692 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5693 (because of problems with the built-in one) was declared to return uschar *
5694 instead of char *, causing compiler failure.
5696 25. Fixed a file descriptor leak when processing alias/forward files.
5698 26. Fixed a minor format string issue in dbfn.c.
5700 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5702 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5703 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5706 29. When a sender address is verified, it is cached, to save repeating the test
5707 when there is more than one recipient in a message. However, when the
5708 verification involves a callout, it is possible for different callout
5709 options to be set for different recipients. It is too complicated to keep
5710 track of this in the cache, so now Exim always runs a verification when a
5711 callout is required, relying on the callout cache for the optimization.
5712 The overhead is duplication of the address routing, but this should not be
5715 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5716 to be verified with callout=postmaster, and the main callout worked but the
5717 postmaster check failed, the verification correctly failed. However, if a
5718 subsequent RCPT command asked for sender verification *without* the
5719 postmaster check, incorrect caching caused this verification also to fail,
5722 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5723 it was not caching the DNS options (qualify_single, search_parents) that
5724 were used when the lookup failed. A subsequent lookup with different
5725 options therefore always gave the same answer, though there were cases
5726 where it should not have. (Example: a "domains = !$mx_any" option on a
5727 dnslookup router: the "domains" option is always processed without any
5728 widening, but the router might have qualify_single set.) Now Exim uses the
5729 cached value only when the same options are set.
5731 32. Added John Jetmore's "exipick" utility to the distribution.
5733 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5734 than a timeout (e.g. a certificate is required, and is not provided), an
5735 Exim server now closes the connection immediately. Previously it waited for
5736 the client to close - but if the client is SSL, it seems that they each
5737 wait for each other, leading to a delay before one of them times out.
5739 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5740 maintained 0.8.x compatibility because I don't think many are using it, and
5741 it is clearly obsolete.
5743 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5746 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5747 client certificate was expired. A simple patch fixes this, though I don't
5748 understand the full logic of why the verify callback is called multiple
5751 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5752 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5753 which causes problems with some clients (such as the Certicom SSL Plus
5754 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5755 disables the coutermeasure allowing Eudora to connect."
5757 38. Exim was not checking that a write() to a log file succeeded. This could
5758 lead to Bad Things if a log got too big, in particular if it hit a file
5759 size limit. Exim now panics and dies if it cannot write to a log file, just
5760 as it does if it cannot open a log file.
5762 39. Modified OS/Makefile-Linux so that it now contains
5764 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5766 The two -D definitions ensure that Exim is compiled with large file
5767 support, which makes it possible to handle log files that are bigger than
5770 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5771 instance) a domain was checked against a named list that involved a lookup,
5772 causing $domain_data to be set, then another domain was checked against the
5773 same list, then the first domain was re-checked, the value of $domain_data
5774 after the final check could be wrong. In particular, if the second check
5775 failed, it could be set empty. This bug probably also applied to
5778 41. The strip_trailing_dot option was not being applied to the address given
5779 with the -f command-line option.
5781 42. The code for reading a message's header from the spool was incrementing
5782 $received_count, but never initializing it. This meant that the value was
5783 incorrect (doubled) while delivering a message in the same process in which
5784 it was received. In the most common configuration of Exim, this never
5785 happens - a fresh exec is done - but it can happen when
5786 deliver_drop_privilege is set.
5788 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5789 it now includes up to 150 characters of the unexpected data in the log
5792 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5793 and building data strings. The size of both of these buffers was 10 000
5794 bytes - far larger than anybody would *ever* want, thought I. Needless to
5795 say, somebody hit the limit. I have increased the maximum line length to
5796 20 000 and the maximum data length of concatenated lines to 100 000. I have
5797 also fixed two bugs, because there was no checking on these buffers. Tsk,
5798 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5799 buffer is too small.
5801 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5802 lsearch lookups. Now it does.
5804 46. When parsing a route_list item in a manualroute router, a fixed-length
5805 buffer was used for the list of hosts. I made this 1024 bytes long,
5806 thinking that nobody would ever have a list of hosts that long. Wrong.
5807 Somebody had a whole pile of complicated expansion conditions, and the
5808 string was silently truncated, leading to an expansion error. It turns out
5809 that it is easier to change to an unlimited length (owing to other changes
5810 that have happened since this code was originally written) than to build
5811 structure for giving a limitation error. The length of the item that
5812 expands into the list of hosts is now unlimited.
5814 47. The lsearch lookup could not handle data where the length of text line was
5815 more than 4095 characters. Such lines were truncated, leading to shortened
5816 data being returned. It should now handle lines of any length.
5818 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5819 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5822 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5823 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5824 no longer happens in comment lines. A list of replacements is now placed
5825 at the head of all of the source files, except those whose only change is
5826 to replace PERL_COMMAND in the very first #! line.
5828 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5829 messages on the queue, with a bottom-up merge sort, using code contributed
5830 by Michael Haardt. This should make operations like -bp somewhat faster on
5831 large queues. It won't affect queue runners, except when queue_run_in_order
5834 51. Installed eximstats 1.31 in the distribution.
5836 52. Added support for SRV lookups to the dnslookup router.
5838 53. If an ACL referred to $message_body or $message_body_end, the value was not
5839 reset for any messages that followed in the same SMTP session.
5841 54. The store-handling optimization for building very long strings was not
5842 differentiating between the different store pools. I don't think this
5843 actually made any difference in practice, but I've tidied it.
5845 55. While running the routers to verify a sender address, $sender_address
5846 was still set to the sender address. This is wrong, because when routing to
5847 send a bounce to the sender, it would be empty. Therefore, I have changed
5848 it so that, while verifying a sender address, $sender_address is set to <>.
5849 (There is no change to what happens when verifying a recipient address.)
5851 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5852 target A or AAAA records (if not already returned) without resetting the
5853 qualify_single or search_parents options of the DNS resolver. These are
5854 inappropriate in this case because the targets of MX and SRV records must
5855 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5856 target that, when qualified, matched something in the local domain. These
5857 two options are now turned off when doing these lookups.
5859 57. It seems that at least some releases of Reiserfs (which does not have the
5860 concept of a fixed number of inodes) returns zero and not -1 for the
5861 number of available inodes. This interacted badly with check_spool_inodes,
5862 which assumed that -1 was the "no such thing" setting. What I have done is
5863 to check that the total number of inodes is greater than zero before doing
5864 the test of how many are available.
5866 58. When a "warn" ACL statement has a log_message modifier, the message is
5867 remembered, and not repeated. This is to avoid a lot of repetition when a
5868 message has many recipients that cause the same warning to be written.
5869 However, Exim was preserving the list of already written lines for an
5870 entire SMTP session, which doesn't seem right. The memory is now reset if a
5871 new message is started.
5873 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5874 debugging output unless log_rewrite was also set.
5876 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5877 of (int)(handle) when we know that handle contains (void *)(-1).
5879 61. The Exim daemon panic-logs an error return when it closes the incoming
5880 connection. However "connection reset by peer" seems to be common, and
5881 isn't really an error worthy of noting specially, so that particular error
5884 62. When Exim is trying to find all the local interfaces, it used to panic and
5885 die if the ioctl to get the interface flags failed. However, it seems that
5886 on at least one OS (Solaris 9) it is possible to have an interface that is
5887 included in the list of interfaces, but for which you get a failure error
5888 for this call. This happens when the interface is not "plumbed" into a
5889 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5890 failure of the "get flags" call assumes that the interface is down.
5892 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5893 makes life easier for people who are doing arithmetic on fields extracted
5894 from dates, where you often get leading zeros that should not be
5895 interpreted as octal.
5897 64. Added qualify_domain to the redirect router, to override the global
5900 65. If a pathologically long header line contained very many addresses (the
5901 report of this problem mentioned 10 000) and each of them was rewritten,
5902 Exim could use up a very large amount of memory. (It kept on making new
5903 copies of the header line as it rewrote, and never released the old ones.)
5904 At the expense of a bit more processing, the header rewriting function has
5905 been changed so that it no longer eats memory in this way.
5907 66. The generation of the Received: header has been moved from the time that a
5908 message starts to be received, to the time that it finishes. The timestamp
5909 in the Received: header should now be very close to that of the <= log
5910 line. There are two side-effects of this change:
5912 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5913 logged header lines no longer include the local Received: line, because
5914 it has not yet been created. The same applies to a copy of the message
5915 that is returned to a non-SMTP sender when a message is rejected.
5917 (b) When a filter file is tested using -bf, no additional Received: header
5918 is added to the test message. After some thought, I decided that this
5921 This change does not affect the value of $received_for. It is still set
5922 after address rewriting, but before local_scan() is called.
5924 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5926 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5927 gave an unhelpful panic error message, and a defer error. I have managed to
5928 change this behaviour so that it now rejects any supplied certificate,
5929 which seems right, as the list of acceptable certificates is empty.
5931 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5932 gave an unhelpful defer error. I have not managed to make this reject any
5933 supplied certificates, but the error message it gives is "no certificate
5934 supplied", which is not helpful.
5936 70. exigrep's output now also includes lines that are not associated with any
5937 message, but which match the given pattern. Implemented by a patch from
5938 Martin Sluka, which also tidied up the Perl a bit.
5940 71. Recipient callout verification, like sender verification, was using <> in
5941 the MAIL FROM command. This isn't really the right thing, since the actual
5942 sender may affect whether the remote host accepts the recipient or not. I
5943 have changed it to use the actual sender in the callout; this means that
5944 the cache record is now keyed on a recipient/sender pair, not just the
5945 recipient address. There doesn't seem to be a real danger of callout loops,
5946 since a callout by the remote host to check the sender would use <>.
5947 [SEE ABOVE: changed after hitting problems.]
5949 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5950 temporary errors. However, in the case of such a code being given after
5951 the end of a data transmission (i.e. after ".") Exim was failing to write
5952 a retry record for the message. (Yes, there was some broken host that was
5953 actually sending 8xx at this point.)
5955 73. An unknown lookup type in a host list could cause Exim to panic-die when
5956 the list was checked. (An example that provoked this was putting <; in the
5957 middle of a list instead of at the start.) If this happened during a DATA
5958 ACL check, a -D file could be left lying around. This kind of configuration
5959 error no longer causes Exim to die; instead it causes a defer error. The
5960 incident is still logged to the main and panic logs.
5962 74. Buglet left over from Exim 3 conversion. The message "too many messages
5963 in one connection" was written to the rejectlog but not the mainlog, except
5964 when address rewriting (yes!) was being logged.
5966 75. Added write_rejectlog option.
5968 76. When a system filter was run not as root (that is, when system_filter_user
5969 was set), the values of the $n variables were not being returned to the
5970 main process; thus, they were not subsequently available in the $sn
5973 77. Added +return_path_on_delivery log selector.
5975 78. A connection timeout was being treated differently from recipients deferred
5976 when testing hosts_max_try with a message that was older than the host's
5977 retry timeout. (The host should not be counted, thus allowing all hosts to
5978 be tried at least once before bouncing.) This may have been the cause of an
5979 occasionally reported bug whereby a message would remain on the queue
5980 longer than the retry timeout, but would be bounced if a delivery was
5981 forced. I say "may" because I never totally pinned down the problem;
5982 setting up timeout/retry tests is difficult. See also the next item.
5984 79. The ultimate address timeout was not being applied to errors that involved
5985 a combination of host plus message (for example, a timeout on a MAIL
5986 command). When an address resolved to a number of possible hosts, and they
5987 were not all tried for each delivery (e.g. because of hosts_max_try), a
5988 message could remain on the queue longer than the retry timeout.
5990 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5993 81. Fixed an obscure SMTP outgoing bug which required at least the following
5994 conditions: (a) there was another message waiting for the same server;
5995 (b) the server returned 5xx to all RCPT commands in the first message so
5996 that the message was not completed; (c) the server dropped the connection
5997 or gave a negative response to the RSET that Exim sends to abort the
5998 transaction. The observed case was a dropped connection after DATA that had
5999 been sent in pipelining mode. That is, the server had advertised PIPELINING
6000 but was not implementing it correctly. The effect of the bug was incorrect
6001 behaviour, such as trying another host, and this could lead to a crash.
6007 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6008 and daemon.c were passed as pointers to ints; they should have been
6009 pointers to socklen_t variables (which are typically unsigned ints).
6011 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6014 3. Fixed a really odd bug that affected only the testing scheme; patching a
6015 certain fixed string in the binary changed the value of another string that
6016 happened to be identical to the end of the original first string.
6018 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6019 name", it returns that address as the IP address. On some operating
6020 systems (e.g. Solaris), it also passes back the IP address string as the
6021 "host name". However, on others (e.g. Linux), it passes back an empty
6022 string. Exim wasn't checking for this, and was changing the host name to an
6023 empty string, assuming it had been canonicalized.
6025 5. Although rare, it is permitted to have more than one PTR record for a given
6026 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6027 all the names associated with an address, because they do in Solaris.
6028 However, it seems that they do not in Linux for data that comes from the
6029 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6030 I found this out when I moved to a new Linux workstation and tried to run
6031 the Exim test suite.
6033 To get round this problem I have changed the code so that it now does its
6034 own call to the DNS to look up PTR records when searching for a host name.
6035 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6036 addresses that are only in /etc/hosts are still found.
6038 This behaviour is, however, controlled by an option called host_lookup_
6039 order, which defaults to "bydns:byaddr". If people want to use the other
6040 order, or indeed, just use one or the other means of lookup, they can
6041 specify it in this variable.
6043 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6044 some operating systems, this comes back from gethostbyaddr() as an empty
6045 string, and this is what Exim used to test for. However, it seems that in
6046 other systems, "." is yielded. Exim now tests for this case too.
6048 7. The values of check_spool_space and check_log_space are now held internally
6049 as a number of kilobytes instead of an absolute number of bytes. If a
6050 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6051 kilobyte. This means that much larger values can be stored.
6053 8. Exim monitor: an attempt to get the action menu when not actually pointing
6054 at a message produces an empty menu entitled "No message selected". This
6055 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6056 no entries in it ("Shell widget menu has zero width and/or height"). So I
6057 have added a single, blank menu entry in this case.
6059 9. Added ${quote_local_part.
6061 10. MIME decoding is now applied to the contents of Subject: header lines when
6064 11. Now that a reference to $sender_host_address automatically causes a reverse
6065 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6066 host lookup before query-style lookups in lists that might use this
6067 variable. This has therefore been abolished, and the "net-" prefix is no
6068 longer necessary for query-style lookups.
6070 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6071 have been a typo for LFLAGS, so it has been changed.
6073 13. The install script calls Exim with "-C /dev/null" in order to find the
6074 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6075 to be output. However, since Exim outputs its version number before the
6076 error, it didn't break the script. It just looked ugly. I fixed this by
6077 always allowing "-C /dev/null" if the caller is root.
6079 14. Ignore overlarge ACL variable number when reading spool file - insurance
6080 against a later release with more variables having written the file.
6082 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6083 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6084 this, as well as the form without the "IPv6" on the front.
6086 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6087 OS/Makefile-Darwin file.
6089 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6090 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6092 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6093 variable. However, some OS use ints for inodes. I've added cast to long int
6094 to get rid of the compiler warning.
6096 19. I had forgotten to lock out "/../" in configuration file names when
6097 ALT_CONFIG_PREFIX was set.
6099 20. Routers used for verification do not need to specify transports. However,
6100 if such a router generated a host list, and callout was configured, Exim
6101 crashed, because it could not find a port number from the (non-existent)
6102 transport. It now assumes port 25 in this circumstance.
6104 21. Added the -t option to exigrep.
6106 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6107 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6108 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6109 them is set, LOOKUP_LSEARCH is forced.
6111 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6113 24. Added sender and host information to the "rejected by local_scan()" log
6114 line; previously there was no indication of these.
6116 25. Added .include_if_exists.
6118 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6119 Linux. It turns out that not all file systems support this. Apparently some
6120 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6121 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6122 not supported on the file descriptor, is now ignored when Exim is trying to
6123 sync a directory. This applies only to Linux.
6125 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6127 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6128 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6131 29. Added support for maildirsize files from supplied patch (modified a bit).
6133 30. The use of :fail: followed by an empty string could lead Exim to respond to
6134 sender verification failures with (e.g.):
6136 550 Verification failed for <xxx>
6137 550 Sender verify failed
6139 where the first response line was missing the '-' that indicates it is not
6140 the final line of the response.
6142 31. The loop for finding the name of the user that called Exim had a hardwired
6143 limit of 10; it now uses the value of finduser_retries, which is used for
6144 all other user lookups.
6146 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6149 33. Exim was neglecting to zero errno before one call of strtol() when
6150 expanding a string and expecting an integer value. On some systems this
6151 resulted in spurious "integer overflow" errors. Also, it was casting the
6152 result into an int without checking.
6154 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6155 did not work. The code looks as if it has *never* worked, though it appears
6156 to have been documented since at least release 1.62. I have made it work.
6158 35. The "timeout_DNS" error in retry rules, also documented since at least
6159 1.62, also never worked. As it isn't clear exactly what this means, and
6160 clearly it isn't a major issue, I have abolished the feature by treating it
6161 as "timeout", and writing a warning to the main and panic logs.
6163 36. The display of retry rules for -brt wasn't always showing the error code
6166 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6167 timeout_connect_A, timeout_connect_MX.
6169 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6170 to the empty sender.
6172 39. The daemon was not analysing the content of -oX till after it had closed
6173 stderr and disconnected from the controlling terminal. This meant that any
6174 syntax errors were only noted on the panic log, and the return code from
6175 the command was 0. By re-arranging the code a little, I've made the
6176 decoding happen first, so such errors now appear on stderr, and the return
6177 code is 1. However, the actual setting up of the sockets still happens in
6178 the disconnected process, so errors there are still only recorded on the
6181 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6182 connections (as happens on some IP stacks) was logged at start up time as
6183 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6184 it from "IPv6 and IPv4", which means that two separate sockets are being
6187 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6188 says whether AF_INET or AF_INET6 was passed as an argument.
6190 42. Exiwhat output was messed up when time zones were included in log
6193 43. Exiwhat now gives more information about the daemon's listening ports,
6194 and whether -tls-on-connect was used.
6196 44. The "port" option of the smtp transport is now expanded.
6198 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6199 silently ignored. Now an error message is written to the main and panic
6202 46. There's a new ACL modifier called "logwrite" which writes to a log file
6203 as soon as it is encountered.
6205 47. Added $local_user_uid and $local_user_gid at routing time.
6207 48. Exim crashed when trying to verify a sender address that was being
6210 49. Exim was recognizing only a space character after ".include". It now also
6211 recognizes a tab character.
6213 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6214 extracting the relevant information from the specification. The man page no
6215 longer contains scrambled data for the -d option, and I've added a section
6216 at the front about calling Exim under different names.
6218 51. Added "extra_headers" argument to the "mail" command in filter files.
6220 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6223 53. Installed eximstats 1.29.
6225 54. Added transport_filter_timeout as a generic transport option.
6227 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6228 Cc: header lines. This was required by RFC 822, but it not required by RFC
6231 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6232 incoming messages that did not have them. Now it does so only if the
6233 message originates locally, that is, if there is no associated remote host
6234 address. When Resent- header lines are present, this applies to the Resent-
6235 lines rather than the non-Resent- lines.
6237 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6238 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6240 58. Messages for configuration errors now include the name of the main
6241 configuration file - useful now that there may be more than one file in a
6242 list (.included file names were always shown).
6244 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6245 for those rare installations that do not start the daemon as root or run it
6246 setuid root. I've cut out the call to initgroups() if the daemon is not
6249 60. The Exim user and group can now be bound into the binary as text strings
6250 that are looked up at the start of Exim's processing.
6252 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6254 62. Added $mailstore_basename variable.
6256 63. Installed patch to sieve.c from Michael Haardt.
6258 64. When Exim failed to open the panic log after failing to open the main log,
6259 the original message it was trying to log was written to stderr and debug
6260 output, but if they were not available (the usual case in production), it
6261 was lost. Now it is written to syslog before the two lines that record the
6262 failures to open the logs.
6264 65. Users' Exim filters run in subprocesses under the user's uid. It is
6265 possible for a "deliver" command or an alias in a "personal" command to
6266 provoke an address rewrite. If logging of address rewriting is configured,
6267 this fails because the process is not running as root or exim. There may be
6268 a better way of dealing with this, but for the moment (because 4.30 needs
6269 to be released), I have disabled address rewrite logging when running a
6270 filter in a non-root, non-exim process.
6276 1. The buildconfig auxiliary program wasn't quoting the value set for
6277 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6278 not defined. This bug was masked in 4.22 by the effect that was fixed in
6281 2. Some messages that were rejected after a message id was allocated were
6282 shown as "incomplete" by exigrep. It no longer does this for messages that
6283 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6285 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6286 have allow_domain_literals set, the ID did not get logged in the <= line.
6287 Domain literals are now always recognized in Message-ID: header lines.
6289 4. The first argument for a ${extract expansion item is the key name or field
6290 number. Leading and trailing spaces in this item were not being ignored,
6291 causing some misleading effects.
6293 5. When deliver_drop_privilege was set, single queue runner processes started
6294 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6295 same command in the process it spins off) were not dropping privilege.
6297 6. When the daemon running as "exim" started a queue runner, it always
6298 re-executed Exim in the spun-off process. This is a waste of effort when
6299 deliver_drop_privilege is set. The new process now just calls the
6300 queue-runner function directly.
6306 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6309 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6310 ignored. Though the use of -oP was forcing the writing of a pid file, it
6311 was always written to the default place.
6313 3. If the message "no IP address found for host xxxx" is generated during
6314 incoming verification, it is now followed by identification of the incoming
6315 connection (so you can more easily find what provoked it).
6317 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6319 5. Added some features to "harden" Exim a bit more against certain attacks:
6321 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6322 be put in Local/Makefile. This is like the never_users runtime option,
6323 but it cannot be overridden. The default setting is "root".
6325 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6326 prefix string with which any file named in a -C command line option
6329 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6330 is retained for -C and -D only if the caller of Exim is root. Without
6331 it, the exim user may also use -C and -D and retain privilege.
6333 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6334 command line option is disabled.
6336 6. Macro names set by the -D option must start with an upper case letter, just
6337 like macro names defined in the configuration file.
6339 7. Added "dereference=" facility to LDAP.
6341 8. Two instances of the typo "uknown" in the source files are fixed.
6343 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6344 the Configure-Makefile script screwed up while processing it.
6346 10. Incorporated PCRE 4.4.
6348 11. The SMTP synchronization check was not operating right at the start of an
6349 SMTP session. For example, it could not catch a HELO sent before the client
6350 waited for the greeting. There is now a check for outstanding input at the
6351 point when the greeting is written. Because of the duplex, asynchronous
6352 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6353 way, but not yet received, when the check is performed.
6355 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6356 on TCP/IP sockets, because this apparently causes some broken clients to
6359 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6360 unchanged) from the Cygwin maintainer.
6362 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6363 was supported instead of testing for mbx. Effectively a typo.
6365 15. The spa authenticator server code was not checking that the input it
6366 received was valid base64.
6368 16. The debug output line for the "set" modifier in ACLs was not showing the
6369 name of the variable that was being set.
6371 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6373 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6374 lookup on its own. Something else was needed to trigger the lookup. For
6375 example, a match in host_lookup or the need for a host name in a host list.
6376 Now, if $sender_host_name is referenced and the host name has not yet been
6377 looked up, a lookup is performed. If the lookup fails, the variable remains
6378 empty, and $host_lookup_failed is set to "1".
6380 19. Added "eqi" as a case-independent comparison operator.
6382 20. The saslauthd authentication condition could segfault if neither service
6383 nor realm was specified.
6385 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6386 error message that was logged was misleading, and incoming SMTP
6387 connections were dropped. The message is now more accurate, and temporary
6388 errors are given to SMTP connections.
6390 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6391 (except RSET) with a 421 error, until QUIT is received. However, it was
6392 failing to send a response to QUIT.
6394 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6395 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6396 ACL is now run after the helo_try_verify_hosts code.
6398 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6399 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6400 case-independent, so other case variants are also recognized). Apparently
6401 some systems use these upper case variants.
6403 25. If more than two messages were waiting for the same host, and a transport
6404 filter was specified for the transport, Exim sent two messages over the
6405 same TCP/IP connection, and then failed with "socket operation on non-
6406 socket" when it tried to send the third.
6408 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6410 27. The extern definition of crypt16() in expand.c was not being excluded when
6411 the OS had its own crypt16() function.
6413 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6414 as a preferred synonym for return_size_limit, both as an option and as an
6417 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6419 30. Changed the default configuration ACL to relax the local part checking rule
6420 for addresses that are not in any local domains. For these addresses,
6421 slashes and pipe symbols are allowed within local parts, but the sequence
6422 /../ is explicitly forbidden.
6424 31. SPA server authentication was not clearing the challenge buffer before
6427 32. log_message in a "warn" ACL statement was writing to the reject log as
6428 well as to the main log, which contradicts the documentation and doesn't
6429 seem right (because no rejection is happening). So I have stopped it.
6431 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6432 However, I am unable to do any testing of this.
6434 34. Fixed an infelicity in the appendfile transport. When checking directories
6435 for a mailbox, to see if any needed to be created, it was accidentally
6436 using path names with one or more superfluous leading slashes; tracing
6437 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6439 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6440 contents are added to the log line that is written for every discarded
6441 recipient. (Previously a log_message setting was ignored.)
6443 36. The ${quote: operator now quotes the string if it is empty.
6445 37. The install script runs exim in order to find its version number. If for
6446 some reason other than non-existence or emptiness, which it checks, it
6447 could not run './exim', it was installing it with an empty version number,
6448 i.e. as "exim-". This error state is now caught, and the installation is
6451 38. An argument was missing from the function that creates an error message
6452 when Exim fails to connect to the socket for saslauthd authentication.
6453 This could cause Exim to crash, or give a corrupted message.
6455 39. Added isip, isip4, and isip6 to ${if conditions.
6457 40. The ACL variables $acl_xx are now saved with the message, and can be
6458 accessed later in routers, transports, and filters.
6460 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6461 strings in the file are not string-expanded.
6463 42. If a MAIL command specified a SIZE value that was too large to fit into an
6464 int variable, the check against message_size_limit failed. Such values are
6465 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6466 day this will have to be increased, but I don't think I want to be around
6467 when emails are that large.
6474 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6475 iconv() is not standard in FreeBSD.
6477 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6478 IPv6 enabled. The observed symptom was a segmentation fault on return from
6479 the function os_common_find_running_interfaces() in src/os.c.
6481 3. In the check_special_case() function in daemon.c I had used "errno" as an
6482 argument name, which causes warnings on some systems. This was basically a
6483 typo, since it was named "eno" in the comments!
6485 4. The code that waits for the clock to tick (at a resolution of some fraction
6486 of a second) so as to ensure message-id uniqueness was always waiting for
6487 at least one whole tick, when it could have waited for less. [This is
6488 almost certainly not relevant at current processor speeds, where it is
6489 unlikely to ever wait at all. But we try to future-proof.]
6491 5. The function that sleeps for a time interval that includes fractions of a
6492 second contained a race. It did not block SIGALRM between setting the
6493 timer, and suspending (a couple of lines later). If the interval was short
6494 and the sigsuspend() was delayed until after it had expired, the suspension
6495 never ended. On busy systems this could lead to processes getting stuck for
6498 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6499 process, before it forks any delivery processes. The open lookup caching
6500 mechanism meant that the open file or database connection was passed into
6501 the delivery process. The problem was that delivery processes always tidy
6502 up cached lookup data. This could cause a problem for the next delivery
6503 process started by the queue runner, because the external queue runner
6504 process does not know about the closure. So the next delivery process
6505 still has data in the lookup cache. In the case of a file lookup, there was
6506 no problem because closing a file descriptor in a subprocess doesn't affect
6507 the parent. However, if the lookup was caching a connection to a database,
6508 the connection was closed, and the second delivery process was likely to
6509 see errors such as "PGSQL: query failed: server closed the connection
6510 unexpectedly". The problem has been fixed by closing all cached lookups
6511 in a queue runner before running a delivery process.
6513 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6514 seem to have the "const" qualifier which it has on other OS. I've
6517 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6518 *of the same type* (client or server) with the same public name that an
6519 error should be diagnosed.
6521 9. When Exim looked up a host name for an IP address, but failed to find the
6522 original IP address when looking up the host name (a safety check), it
6523 output the message "<ip address> does not match any IP for NULL", which was
6524 confusing, to say the least. The bug was that the host name should have
6525 appeared instead of "NULL".
6527 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6528 user that is built into the binary, and the -C or -D options is used, root
6529 privilege is dropped before the configuration file is read. In addition,
6530 logging is switched to stderr instead of the normal log files. If the
6531 configuration then re-defines the Exim user, the unprivileged environment
6532 is probably not what is expected, so Exim logs a panic warning message (but
6535 However, if deliver_drop_privilege is set, the unprivileged state may well
6536 be exactly what is intended, so the warning has been cut out in that case,
6537 and Exim is allowed to try to write to its normal log files.
6543 1. smtp_return_error_details was not giving details for temporary sender
6544 or receiver verification errors.
6546 2. Diagnose a configuration error if two authenticators have the same public
6549 3. Exim used not to create the message log file for a message until the first
6550 delivery attempt. This could be confusing when incoming messages were held
6551 for policy or load reasons. The message log file is now created at the time
6552 the message is received, and an initial "Received" line is written to it.
6554 4. The automatically generated man page for command line options had a minor
6555 bug that caused no ill effects; however, a more serious problem was that
6556 the procedure for building the man page automatically didn't always
6557 operate. Consequently, release 4.20 contains an out-of-date version. This
6558 shouldn't happen again.
6560 5. When building Exim with embedded Perl support, the script that builds the
6561 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6562 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6564 6. The freeze_tell option was not being used for messages that were frozen on
6565 arrival, either by an ACL or by local_scan().
6567 7. Added the smtp_incomplete_transaction log selector.
6569 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6570 was accepting AUTH without a new EHLO.
6572 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6573 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6576 10. Logging of TCP/IP connections (when configured) now happens in the main
6577 daemon process instead of the child process, so that the TCP/IP connection
6578 count is more accurate (but it can never be perfect).
6580 11. The use of "drop" in a nested ACL was not being handled correctly in the
6581 outer ACL. Now, if condition failure induced by the nested "drop" causes
6582 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6583 or "require"), the connection is dropped.
6585 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6586 that yield "discard" can now be used with an "accept" or a "discard" verb,
6587 but an error is generated for any others (because I can't see a useful way
6588 to define what should happen).
6590 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6591 are now processed in the same way as lines in the Exim configuration file.
6592 In particular, continuation lines are supported.
6594 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6596 15. Added -ti meaning -t -i.
6598 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6599 domains, and warn by logging if others are found.
6601 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6602 returned from GIFCONF when Exim is trying to find the list of interfaces on
6603 a host. The code in os.c has been modified to copy each ifreq to an aligned
6604 structure in all cases.
6606 Also, in some cases, the returned ifreq's were being copied to a 'struct
6607 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6608 means the last couple of bytes of an IPv6 address could be chopped if the
6609 ifreq contained only a normal sockaddr (14 bytes storage).
6611 18. Named domain lists were not supported in the hosts_treat_as_local option.
6612 An entry such as +xxxx was not recognized, and was treated as a literal
6615 19. Ensure that header lines added by a DATA ACL are included in the reject log
6616 if the ACL subsequently rejects the message.
6618 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6619 MD5 (which is deprecated).
6621 21. When testing a filter file using -bf, Exim was writing a message when it
6622 took the sender from a "From " line in the message, but it was not doing so
6623 when it took $return_path from a Return-Path: header line. It now does.
6625 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6626 with a valid header line field name (a series of printing characters
6627 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6629 23. Changed "disc" in the source to "disk" to conform to the documentation and
6630 the book and for uniformity.
6632 24. Ignore Sendmail's -Ooption=value command line item.
6634 25. When execve() failed while trying to run a command in a pipe transport,
6635 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6636 could be confused with a return value of 69 from the command itself. This
6637 has been changed to 127, the value the shell returns if it is asked to run
6638 a non-existent command. The wording for the related log line suggests a
6639 non-existent command as the problem.
6641 26. If received_header_text expands to an empty string, do not add a Received:
6642 header line to the message. (Well, it adds a token one on the spool, but
6643 marks it "old" so that it doesn't get used or transmitted.)
6645 27. Installed eximstats 1.28 (addition of -nt option).
6647 28. There was no check for failure on the call to getsockname() in the daemon
6648 code. This can fail if there is a shortage of resources on the system, with
6649 ENOMEM, for example. A temporary error is now given on failure.
6651 29. Contrary to the C standard, it seems that in some environments, the
6652 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6653 program. Exim now does this explicitly; it affects the formatting of
6654 timestamps using strftime().
6656 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6657 complaints. I've now initialized all the variables, to avoid this.
6659 32. Header lines added by a system filter were not being "seen" during
6660 transport-time rewrites.
6662 33. The info_callback() function passed to OpenSSL is set up with type void
6663 (*)(SSL *, int, int), as described somewhere. However, when calling the
6664 function (actually a macro) that sets it up, the type void(*)() is
6665 expected. I've put in a cast to prevent warnings from picky compilers.
6667 34. If a DNS black list lookup found a CNAME record, but there were no A
6668 records associated with the domain it pointed at, Exim crashed.
6670 35. If a DNS black list lookup returned more than one A record, Exim ignored
6671 all but the first. It now scans all returned addresses if a particular IP
6672 value is being sought. In this situation, the contents of the
6673 $dnslist_value variable are a list of all the addresses, separated by a
6676 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6677 to accept a host name and all its aliases if the forward lookup for any of
6678 them yielded the IP address of the incoming connection. Now it accepts only
6679 those names whose forward lookup yields the correct IP address. Any other
6680 names are discarded. This closes a loophole whereby a rogue DNS
6681 administrator could create reverse DNS records to break through a
6682 wildcarded host restriction in an ACL.
6684 37. If a user filter or a system filter that ran in a subprocess used any of
6685 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6686 the wrong values were passed to the pipe command ($thisaddress had the
6687 value of $0, $0 had the value of $1, etc). This bug was introduced by
6688 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6690 38. Improved the line breaking for long SMTP error messages from ACLs.
6691 Previously, if there was no break point between 40 and 75 characters, Exim
6692 left the rest of the message alone. Two changes have been made: (a) I've
6693 reduced the minimum length to 35 characters; (b) if it can't find a break
6694 point between 35 and 75 characters, it looks ahead and uses the first one
6695 that it finds. This may give the occasional overlong line, but at least the
6696 remaining text gets split now.
6698 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6699 file descriptors might be low, and that setting 1000 would always raise it.
6700 It turns out that in some environments, the limit is already over 1000 and
6701 that lowering it causes trouble. So now Exim takes care not to decrease it.
6703 40. When delivering a message, the value of $return_path is set to $sender_
6704 address at the start of routing (routers may change the value). By an
6705 oversight, this default was not being set up when an address was tested by
6706 -bt or -bv, which affected the outcome if any router or filter referred to
6709 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6710 log only when "message" or "log_message" are set. However, if one of the
6711 conditions was an address verification, or a call to a nested ACL, the
6712 messages generated by the underlying test were being passed through. This
6713 no longer happens. The underlying message is available in $acl_verify_
6714 message for both "message" and "log_message" expansions, so it can be
6715 passed through if needed.
6717 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6718 new expansion $bh_ to give the encoded byte string without charset
6719 translation. Translation happens only if iconv() is available; HAVE_ICONV
6720 indicates this at build time. HEADERS_CHARSET gives the charset to
6721 translate to; headers_charset can change it in the configuration, and
6722 "headers charset" can change it in an individual filter file.
6724 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6725 that creates RFC 2047 encoded "words" labels them as that charset instead
6726 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6727 expansion operator; (ii) when Exim creates a From: line for a local
6728 message; (iii) when a header line is rewritten to include a "phrase" part.
6730 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6731 buggy, causing it to skip the first lines of messages whose message ID
6732 ended in 'D'. This would not have bitten before Exim release 4.14, because
6733 message IDs were unlikely to end in 'D' before then. The effect was to have
6734 incorrect size information for certain domains.
6736 45. #include "config.h" was missing at the start of the crypt16.c module. This
6737 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6740 46. If there was a timeout during a "random" callout check, Exim treated it as
6741 a failure of the random address, and carried on sending RSET and the real
6742 address. If the delay was just some slowness somewhere, the response to the
6743 original RCPT would be taken as a response to RSET and so on, causing
6744 mayhem of various kinds.
6746 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6747 when I implemented it. It didn't allow for the fact that some option values
6748 may legitimately be negative (e.g. size_addition), and it didn't even do
6749 the right test for positive values.
6751 48. Domain names in DNS records are case-independent. Exim always looks them up
6752 in lower case. Some resolvers return domain names in exactly the case they
6753 appear in the zone file, that is, they may contain uppercase letters. Not
6754 all resolvers do this - some return always lower case. Exim was treating a
6755 change of case by a resolver as a change of domain, similar to a widening
6756 of a domain abbreviation. This triggered its re-routing code and so it was
6757 trying to route what was effectively the same domain again. This normally
6758 caused routing to fail (because the router wouldn't handle the domain
6759 twice). Now Exim checks for this case specially, and just changes the
6760 casing of the domain that it ultimately uses when it transmits the message
6763 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6766 50. If a filter generated a file delivery with a non-absolute name (possible if
6767 no home directory exists for the router), the forbid_file option was not
6770 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6771 the existing equality matching.
6773 52. Exim was using ints instead of ino_t variables in some places where it was
6774 dealing with inode numbers.
6776 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6777 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6778 called TMPDIR, and if it finds it is different, it changes its value.
6780 54. The smtp_printf() function is now made available to local_scan() so
6781 additional output lines can be written before returning. There is also an
6782 smtp_fflush() function to enable the detection of a dropped connection.
6783 The variables smtp_input and smtp_batched_input are exported to
6786 55. Changed the default runtime configuration: the message "Unknown user"
6787 has been removed from the ACL, and instead placed on the localuser router,
6788 using the cannot_route_message feature. This means that any verification
6789 failures that generate their own messages won't get overridden. Similarly,
6790 the "Unrouteable address" message that was in the ACL for unverifiable
6791 relay addresses has also been removed.
6793 56. Added hosts_avoid_esmtp to the smtp transport.
6795 57. The exicyclog script was not checking for the esoteric option
6796 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6797 will work only if exicyclog is run under the appropriate euid.
6799 58. Following a discussion on the list, the rules by which Exim recognises line
6800 endings on incoming messages have been changed. The -dropcr and drop_cr
6801 options are now no-ops, retained only for backwards compatibility. The
6802 following line terminators are recognized: LF CRLF CR. However, special
6803 processing applies to CR:
6805 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6806 nor a local message in the state where . is a terminator.
6808 (ii) If a bare CR is encountered in a header line, an extra space is added
6809 after the line terminator so as not to end the header. The reasoning
6810 behind this is that bare CRs in header lines are most likely either
6811 to be mistakes, or people trying to play silly games.
6813 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6814 was being incorrectly given as 18 bytes larger than it should have been.
6815 This is a VOB (very old bug).
6817 60. This may never have affected anything current, but just in case it has:
6818 When the local host is found other than at the start of a list of hosts,
6819 the local host, those with the same MX, and any that follow, are discarded.
6820 When the list in question was part of a longer list of hosts, the following
6821 hosts (not currently being processed) were also being discarded. This no
6822 longer happens. I'm not sure if this situation could ever has previously
6825 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6828 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6829 header. This is something that is recommended in a new Internet Draft, and
6830 is something that is documented as being done by Sendmail. There are two
6831 possible values. For messages generated by the autoreply transport, Exim
6834 Auto-Submitted: auto-replied
6836 whereas for all other generated messages (e.g. bounces) it adds
6838 Auto-Submitted: auto-generated
6840 63. The "personal" condition in filters now includes a test for the
6841 Auto-Submitted: header. If it contains the string "auto-" the message it
6842 not considered personal.
6844 64. Added rcpt_include_affixes as a generic transport option.
6846 65. Added queue_only_override (default true).
6848 66. Added the syslog_duplication option.
6850 67. If what should have been the first header line of a message consisted of
6851 a space followed by a colon, Exim was mis-interpreting it as a header line.
6852 It isn't of course - it is syntactically invalid and should therefore be
6853 treated as the start of the message body. The misbehaviour could have
6854 caused a number of strange effects, including loss of data in subsequent
6855 header lines, and spool format errors.
6857 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6858 client host had authenticated. This control can now be exercised by an ACL
6859 for more flexibility.
6861 69. By default, callouts do not happen when testing with -bh. There is now a
6862 variant, -bhc, which does actually run the callout code, including
6863 consulting and updating the callout cache.
6865 70. Added support for saslauthd authentication, courtesy of Alexander
6868 71. If statvfs() failed on the spool or log directories while checking their
6869 size for availability, Exim confusingly gave the error "space shortage".
6870 Furthermore, in debugging mode it crashed with a floating point exception.
6871 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6872 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6873 serious problem, Exim now writes to the main and panic logs when this
6874 happens, with details of the failure. It then refuses to accept the
6875 incoming message, giving the message "spool directory problem" or "log
6876 directory problem" with a 421 code for SMTP messages.
6878 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6879 0, 1, and 2 exist, because some OS complain for execs without them (see
6880 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6881 descriptors. However, the code omitted to check that the open succeeded,
6882 causing mysterious errors if for some reason the permissions on /dev/null
6883 got screwed. Now Exim writes a message to the main and panic logs, and
6884 bombs out if it can't open /dev/null.
6886 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6887 interact so that it is all more flexible. It is supposed to remain
6888 backwards compatible. Also added extra_local_interfaces.
6890 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6891 to bomb out with an assertion failure - to the client this appears as a
6892 connection drop. This problem occurs in the part of the code that was taken
6893 from the Samba project. Fortunately, the assertion is in a very simple
6894 function, so I have fixed this by reproducing the function inline in the
6895 one place where it is called, and arranging for authentication to fail
6896 instead of killing the process with assert().
6898 75. The SPA client code was not working when the server requested OEM rather
6899 than Unicode encoding.
6901 76. Added code to make require_files with a specific uid setting more usable in
6902 the case where statting the file as root fails - usually a non-root-mounted
6903 NFS file system. When this happens and the failure is EACCES, Exim now
6904 forks a subprocess and does the per-uid checking as the relevant uid.
6906 77. Added process_log_path.
6908 78. If log_file_path was not explicitly set, a setting of check_log_space or
6909 check_log_inodes was ignored.
6911 79. If a space check for the spool or log partitions fails, the incident is now
6912 logged. Of course, in the latter case the data may get lost...
6914 80. Added the %p formatting code to string_format() so that it can be used to
6915 print addresses in debug_print(). Adjusted all the address printing in the
6916 debugging in store.c to use %p rather than %d.
6918 81. There was a concern that a line of code in smtp_in.c could overflow a
6919 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6920 initially expressed, the concern was not well-founded, because trailing
6921 spaces are removed early. However, if the trailing spaces were followed by
6922 a NULL, they did not get removed, so the overflow was possible. Two fixes
6925 (a) I re-wrote the offending code in a cleaner fashion.
6926 (b) If an incoming SMTP command contains a NULL character, it is rejected
6929 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6930 runs initgroups(), so that if the Exim user is in any additional groups,
6931 they will be used during message reception.
6937 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob