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.
10 HS/01 Handle trailing backslash gracefully. (CVE-2019-15846)
16 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917, OVE-20190718-0006)
22 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
23 buffer overrun for (non-chunking) other transports.
25 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
26 TLS1.3, means that a server rejecting a client certificate is not visible
27 to the client until the first read of encrypted data (typically the
28 response to EHLO). Add detection for that case and treat it as a failed
29 TLS connection attempt, so that the normal retry-in-clear can work (if
32 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
33 and/or domain. Found and fixed by Jason Betts.
35 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
36 configuration). If a CNAME target was not a wellformed name pattern, a
39 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
40 the OS reports them interleaved with other addresses.
47 JH/01 Remove code calling the customisable local_scan function, unless a new
48 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
50 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
51 non-signal-safe functions being used.
53 JH/03 Bug 2269: When presented with a received message having a stupidly large
54 number of DKIM-Signature headers, disable DKIM verification to avoid
55 a resource-consumption attack. The limit is set at twenty.
57 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
58 report of oldest_pass in ${authres } in consequence, and separate out
59 some descriptions of reasons for verification fail.
61 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
62 files in the spool were present and unlocked. A queue-runner could spot
63 them, resulting in a duplicate delivery. Fix that by doing the unlock
64 after the unlink. Investigation by Tim Stewart. Take the opportunity to
65 add more error-checking on spoolfile handling while that code is being
68 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
69 No known attacks, no CVE, this is defensive hardening.
71 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
72 a queue-runner could start a delivery while other operations were ongoing.
73 Cutthrough delivery was a common victim, resulting in duplicate delivery.
74 Found and investigated by Tim Stewart. Fix by using the open message data
75 file handle rather than opening another, and not locally closing it (which
76 releases a lock) for that case, while creating the temporary .eml format
77 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
79 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
80 $sender_verify_failure/$recipient_verify_failure to "random".
82 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
85 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
86 Previously this would segfault.
88 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
91 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
92 like zero, since the resolver should be doing this for us, But we need one
93 as a CNAME but no MX presence gets the CNAME returned; we need to check
94 that doesn't point to an MX to declare it "no MX returned" rather than
95 "error, loop". A new main option is added so the older capability of
96 following some limited number of chain links is maintained.
98 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
100 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
101 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
102 now-deprecated earlier definitions used only specified the range up to TLS
103 1.2 (in the older-version library docs).
105 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
107 JH/15 Rework TLS client-side context management. Stop using a global, and
108 explicitly pass a context around. This enables future use of TLS for
109 connections to service-daemons (eg. malware scanning) while a client smtp
110 connection is using TLS; with cutthrough connections this is quite likely.
112 JH/16 Fix ARC verification to do AS checks in reverse order.
114 JH/17 Support a "tls" option on the ${readsocket } expansion item.
116 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
117 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
118 Previously the "utf8" would be re-prepended for every additional message.
120 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
121 Previously thery were accepted, resulting in issues when attempting to
122 forward messages to a non-supporting MTA.
124 PP/02 Let -n work with printing macros too, not just options.
126 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
127 one parent address was copied, and bogus data was used at delivery-logging
128 time. Either a crash (after delivery) or bogus log data could result.
129 Discovery and analysis by Tim Stewart.
131 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
132 Previously if the string ended mid-character, we did not insert the
133 promised '?' replacement.
135 PP/04 Documentation: current string operators work on bytes, not codepoints.
137 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
138 should pack well giving a smaller memory footprint so better caching and
139 therefore performance. Group the declarations where this can't be done so
140 that the byte-sized flag variables are not interspersed among pointer
141 variables, giving a better chance of good packing by the compiler.
143 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
144 non-null, to avoid issues with sites running BATV. Previously reports were
145 sent with an empty envelope sender so looked like bounces.
147 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
148 The ignore_error flag wasn't being returned from the filter subprocess so
149 was not set for later routers. Investigation and fix by Matthias Kurz.
151 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
152 and a msg:complete for the whole, when a message is manually removed using
153 -Mrm. Developement by Matthias Kurz, hacked on by JH.
155 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
156 a "Gnu special" function, asprintf() in the DB utility binary builds; I
157 hope that is portable enough.
159 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
160 requiring a known-CA anchor certificate; make it now rely entirely on the
161 TLSA as an anchor. Checking the name on the leaf cert against the name
162 on the A-record for the host is still done for TA (but not for EE mode).
164 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
165 included in delivery lines for non-proxied connections, when compiled with
166 SUPPORT_SOCKS and running with proxy logging enabled.
168 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
169 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
170 move the existing event to fire before the normal logging of message
171 failure so that custom logging is bracketed by normal logging.
173 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
174 msg:fail:internal event. Developement by Matthias Kurz.
176 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
177 far too small for todays use of crypto signatures stored there. Go all
178 the way to the max DNS message size of 64kB, even though this might be
179 overmuch for IOT constrained device use.
181 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
182 copy a string over itself. The library routine is documented as not
183 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
185 JH/32 For main options check_spool_space and check_inode_space, where the
186 platform supports 64b integers, support more than the previous 2^31 kB
187 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
188 the previous G, M, k.
190 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
191 $authenticated_fail_id variable on authentication failure. Previously
194 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
195 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
196 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
197 GNUTLS_SEC_PARAM_MEDIUM.
199 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
200 side. Previously we would continue as if no SNI had been received.
202 JH/36 Harden the handling of string-lists. When a list consisted of a sole
203 "<" character, which should be a list-separator specification, we walked
204 off past the nul-terimation.
206 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
207 causes) even when the retry time is not yet met. Previously they were
208 not, meaning that when (say) an account was over-quota and temp-rejecting,
209 and multiple senders' messages were queued, only one sender would get
210 notified on each configured delay_warning cycle.
212 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
214 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
215 cipher-suites, an error can be left on the stack even for a succeeding
216 accept; this results in impossible error messages when a later operation
219 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
220 return error codes indicating retry. Under TLS1.3 this becomes required.
222 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
223 it only wrote the new authenticators, resulting in a lack of tracking of
224 peer changes of ESMTP extensions until the next cache flush.
226 JH/41 Fix the loop reading a message header line to check for integer overflow,
227 and more-often against header_maxsize. Previously a crafted message could
228 induce a crash of the recive process; now the message is cleanly rejected.
230 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
231 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
237 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
238 When redis_servers is set to a list of > 1 element, and the Redis servers
239 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
240 case of MOVED into a DEFER case instead, thus moving the query onto the
241 next server in the list. For a cluster of N elements, all N servers must
242 be defined in redis_servers.
244 GF/02 Catch and remove uninitialized value warning in exiqsumm
245 Check for existence of @ARGV before looking at $ARGV[0]
247 JH/01 Replace the store_release() internal interface with store_newblock(),
248 which internalises the check required to safely use the old one, plus
249 the allocate and data copy operations duplicated in both (!) of the
250 extant use locations.
252 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
253 modifier. This matches the restriction on the commandline.
255 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
256 Previously only the last row was returned.
258 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
259 we assumed that tags in the header were well-formed, and parsed the
260 element content after inspecting only the first char of the tag.
261 Assumptions at that stage could crash the receive process on malformed
264 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
265 While running the DKIM ACL we operate on the Permanent memory pool so that
266 variables created with "set" persist to the DATA ACL. Also (at any time)
267 DNS lookups that fail create cache records using the Permanent pool. But
268 expansions release any allocations made on the current pool - so a dnsdb
269 lookup expansion done in the DKIM ACL releases the memory used for the
270 DNS negative-cache, and bad things result. Solution is to switch to the
271 Main pool for expansions.
272 While we're in that code, add checks on the DNS cache during store_reset,
273 active in the testsuite.
274 Problem spotted, and debugging aided, by Wolfgang Breyha.
276 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
277 When none of the hosts presented to a transport match an already-open
278 connection, close it and proceed with the list. Previously we would
279 queue the message. Spotted by Lena with Yahoo, probably involving
282 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
283 Previously a spurious "250 OK id=" response was appended to the proper
286 JH/08 The "support for" informational output now, which built with Content
287 Scanning support, has a line for the malware scanner interfaces compiled
288 in. Interface can be individually included or not at build time.
290 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
291 by the template makefile "src/EDITME". The "STREAM" support for an older
292 ClamAV interface method is removed.
294 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
295 rows affected is given instead).
297 JH/11 The runtime Berkeley DB library version is now additionally output by
298 "exim -d -bV". Previously only the compile-time version was shown.
300 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
301 SMTP connection. Previously, when one had more receipients than the
302 first, an abortive onward connection was made. Move to full support for
303 multiple onward connections in sequence, handling cutthrough connection
304 for all multi-message initiating connections.
306 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
307 routers. Previously, a multi-recipient message would fail to match the
308 onward-connection opened for the first recipient, and cause its closure.
310 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
311 a timeout on read on a GnuTLS initiating connection, resulting in the
312 initiating connection being dropped. This mattered most when the callout
313 was marked defer_ok. Fix to keep the two timeout-detection methods
316 JH/15 Relax results from ACL control request to enable cutthrough, in
317 unsupported situations, from error to silently (except under debug)
318 ignoring. This covers use with PRDR, frozen messages, queue-only and
321 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
323 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
324 metadata, resulting in a crash in free().
326 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
327 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
328 Broken also in d185889f4, with init system revamp.
330 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
331 Previously we abruptly closed the connection after reading a malware-
332 found indication; now we go on to read the "scan ok" response line,
335 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
336 ACL. Previously, a crash would result.
338 JH/19 Speed up macro lookups during configuration file read, by skipping non-
339 macro text after a replacement (previously it was only once per line) and
340 by skipping builtin macros when searching for an uppercase lead character.
342 JH/20 DANE support moved from Experimental to mainline. The Makefile control
343 for the build is renamed.
345 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
346 was allocated for every new TLS startup, meaning one per message. Fix
347 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
349 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
350 reported the original. Fix to report (as far as possible) the ACL
351 result replacing the original.
353 JH/23 Fix memory leak during multi-message connections using STARTTLS under
354 OpenSSL. Certificate information is loaded for every new TLS startup,
355 and the resources needed to be freed.
357 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
359 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
362 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
363 DATA response info to the (existing) per-recipient response info for
364 the "C=" log element. It can have useful tracking info from the
365 destination system. Patch from Simon Arlott.
367 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
368 length value. Previously this would segfault.
370 HS/02 Support Avast multiline protoocol, this allows passing flags to
371 newer versions of the scanner.
373 JH/28 Ensure that variables possibly set during message acceptance are marked
374 dead before release of memory in the daemon loop. This stops complaints
375 about them when the debug_store option is enabled. Discovered specifically
376 for sender_rate_period, but applies to a whole set of variables.
377 Do the same for the queue-runner and queue-list loops, for variables set
378 from spool message files. Do the same for the SMTP per-message loop, for
379 certain variables indirectly set in ACL operations.
381 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
382 as a multi-recipient message from a mailinglist manager). The coding had
383 an arbitrary cutoff number of characters while checking for more input;
384 enforced by writing a NUL into the buffer. This corrupted long / fast
385 input. The problem was exposed more widely when more pipelineing of SMTP
386 responses was introduced, and one Exim system was feeding another.
387 The symptom is log complaints of SMTP syntax error (NUL chars) on the
388 receiving system, and refused recipients seen by the sending system
389 (propating to people being dropped from mailing lists).
390 Discovered and pinpointed by David Carter.
392 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
393 replaced by the ${authresults } expansion.
395 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
397 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
398 allows proper process termination in container environments.
400 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
401 Previously the "final dot" had a newline after it; ensure it is CR,LF.
403 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
404 and "err_perm", deprecated since 4.83 when the RFC-defined words
405 "temperror" and "permerror" were introduced.
407 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
408 transport-filters or DKIM-signing. The restriction was lost in the
409 consolidation of verify-callout and delivery SMTP handling.
410 Extend the restriction to also cover ARC-signing.
412 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
413 in defer=pass mode supply a 450 to the initiator. Previously the message
416 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
417 tls_require_ciphers is used as before.
419 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
420 "pass_unscanned". Only tmpfails from the scanner are written to
421 the paniclog, as they may require admin intervention (permission
422 denied, license issues). Other scanner errors (like decompression
423 bombs) do not cause a paniclog entry.
425 JH/36 Fix reinitialisation of DKIM logging variable between messages.
426 Previously it was possible to log spurious information in receive log
429 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
430 triggered odd behaviour from Outlook Express clients.
432 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
435 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
436 since the IETF WG has not yet settled on that versus the original
437 "bare" representation.
439 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
440 Previously the millisecond value corrupted the output.
441 Fix also for syslog_pid=no and log_selector +pid, for which the pid
442 corrupted the output.
448 JH/01 Rework error string handling in TLS interface so that the caller in
449 more cases is responsible for logging. This permits library-sourced
450 string to be attached to addresses during delivery, and collapses
451 pairs of long lines into single ones.
453 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
454 during configuration. Wildcards are allowed and expanded.
456 JH/02 Rework error string handling in DKIM to pass more info back to callers.
457 This permits better logging.
459 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
460 do not close it down and have the child transport start it up again on
461 the passed-on TCP connection. Instead, proxy the child (and any
462 subsequent ones) for TLS via a unix-domain socket channel. Logging is
463 affected: the continued delivery log lines do not have any DNSSEC, TLS
464 Certificate or OCSP information. TLS cipher information is still logged.
466 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
467 identical IP addresses on different listening ports. Will also affect
470 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
471 add noisy ifdef guards to special-case this sillyness.
472 Patch from Bernd Kuhls.
474 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
475 than 255 are no longer allowed.
477 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
478 Disable the session-cache too, which might reduce our load. Since we
479 currrectly use a new context for every connection, both as server and
480 client, there is no benefit for these.
481 GnuTLS appears to not support tickets server-side by default (we don't
482 call gnutls_session_ticket_enable_server()) but client side is enabled
483 by default on recent versions (3.1.3 +) unless the PFS priority string
486 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
487 <https://reproducible-builds.org/specs/source-date-epoch/>.
489 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
490 the check for any unsuccessful recipients did not notice the limit, and
491 erroneously found still-pending ones.
493 JH/08 Pipeline CHUNKING command and data together, on kernels that support
494 MSG_MORE. Only in-clear (not on TLS connections).
496 JH/09 Avoid using a temporary file during transport using dkim. Unless a
497 transport-filter is involved we can buffer the headers in memory for
498 creating the signature, and read the spool data file once for the
499 signature and again for transmission.
501 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
502 4.77 as the kernel support then wasn't solid, having issues in 64bit
503 mode. Now, it's been long enough. Add support for FreeBSD also.
505 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
506 case where the routing stage had gathered several addresses to send to
507 a host before calling the transport for the first, we previously failed
508 to close down TLS in the old transport process before passing the TCP
509 connection to the new process. The new one sent a STARTTLS command
510 which naturally failed, giving a failed delivery and bloating the retry
511 database. Investigation and fix prototype from Wolfgang Breyha.
513 JH/12 Fix check on SMTP command input synchronisation. Previously there were
514 false-negatives in the check that the sender had not preempted a response
515 or prompt from Exim (running as a server), due to that code's lack of
516 awareness of the SMTP input buffering.
518 PP/04 Add commandline_checks_require_admin option.
519 Exim drops privileges sanely, various checks such as -be aren't a
520 security problem, as long as you trust local users with access to their
521 own account. When invoked by services which pass untrusted data to
522 Exim, this might be an issue. Set this option in main configuration
523 AND make fixes to the calling application, such as using `--` to stop
526 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
527 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
528 responses to those, into a single TLS record each way (this usually means
529 a single packet). As a side issue, smtp_enforce_sync now works on TLS
532 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
533 affects you only if you're dancing at the edge of the param size limits.
534 If you are, and this message makes sense to you, then: raise the
535 configured limit or use OpenSSL 1.1. Nothing we can do for older
538 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
539 cmdline element to get the documented default one. Previously it was
542 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
543 are unsuable for non-message-specific reasons - eg. connection timeout,
544 banner-time rejection.
546 JH/16 Fix logging of delivery remote port, when specified by router, under
549 PP/06 Repair manualroute's ability to take options in any order, even if one
550 is the name of a transport.
553 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
555 JH/17 Change the list-building routines interface to use the expanding-string
556 triplet model, for better allocation and copying behaviour.
558 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
559 Previously it was constructed the first time a possibly-matching string
560 was met in the configuration file input during startup; now it is done
563 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
564 compatible one, to avoid the (poorly documented) possibility of a config
565 file in the working directory redirecting the DB files, possibly correpting
566 some existing file. CVE-2017-10140 assigned for BDB.
568 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
569 cache-hot. Previously, although the result was properly cached, the
570 initial verify call returned a defer.
572 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
573 the main verify for receipient in uncached-mode.
575 JH/22 Retire historical build files to an "unsupported" subdir. These are
576 defined as "ones for which we have no current evidence of testing".
578 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
579 if present. Previously it was ignored.
581 JH/24 Start using specified-initialisers in C structure init coding. This is
582 a C99 feature (it's 2017, so now considered safe).
584 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
585 if was a fixed-sized field and bitmask ops via macros; it is now more
588 PP/07 GitHub PR 56: Apply MariaDB build fix.
589 Patch provided by Jaroslav Škarvada.
591 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
592 during Coverity cleanups [4.87 JH/47]
593 Diagnosis and fix provided by Michael Fischer v. Mollard.
595 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
596 the right size to place the terminating semicolon on its own folded
597 line, the header hash was calculated to an incorrect value thanks to
598 the (relaxed) space the fold became.
600 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
601 and confused the parent.
603 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
604 which could crash as a result. This could lead to undeliverable messages.
606 JH/28 Logging: "next input sent too soon" now shows where input was truncated
609 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
610 matters on fast-turnover and PID-randomising systems, which were getting
611 out-of-order delivery.
613 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
614 a possibly-overlapping copy. The symptom was that "Remote host closed
615 connection in response to HELO" was logged instead of the actual 4xx
618 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
619 Previously only that bufferd was discarded, resulting in SYMTP command
622 JH/32 DKIM: when a message has multiple signatures matching an identity given
623 in dkim_verify_signers, run the dkim acl once for each. Previously only
624 one run was done. Bug 2189.
626 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
627 "panic the current process" to "deliberately defer". The panic log is
628 still written with the problem list name; the mail and reject logs now
629 get a temp-reject line for the message that was being handled, saying
630 something like "domains check lookup or other defer". The SMTP 451
631 message is still "Temporary local problem".
633 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
634 A crafted sequence of BDAT commands could result in in-use memory beeing
635 freed. CVE-2017-16943.
637 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
638 from SMTP input. Previously it was always done; now only done for DATA
639 and not BDAT commands. CVE-2017-16944.
641 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
642 to the message (such as an overlong header line). Previously this was
643 not done and we did not exit BDAT mode. Followon from the previous item
644 though a different problem.
650 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
651 than -2003 did; needs libidn2 in addition to libidn.
653 JH/02 The path option on a pipe transport is now expanded before use.
655 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
656 Patch provided by "Björn", documentation fix added too.
658 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
659 missing a wire-to-host endian conversion.
661 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
662 close after a BDAT command line could be taken as a following command,
663 giving a synch failure. Fix by only checking for synch immediately
664 before acknowledging the chunk.
666 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
667 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
668 macro. Patches provided by Josh Soref.
670 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
671 Previously we did not; the RFC seems ambiguous and VRFY is not listed
672 by IANA as a service extension. However, John Klensin suggests that we
675 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
676 the dkim code may be unix-mode line endings rather than smtp wire-format
677 CRLF, so prepend a CR to any bare LF.
679 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
680 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
682 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
683 the first were themselves being wrongly included in the feed into dkim
684 processing; with most chunk sizes in use this resulted in an incorrect
685 body hash calculated value.
687 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
688 DKIM signature block, for verification. Although advised against by
689 standards it is specifically not ruled illegal.
691 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
693 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
694 missing a body hash (the bh= tag).
696 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
697 It seems that HAProxy sends the Proxy Protocol information in clear and
698 only then does a TLS startup, so do the same.
700 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
701 TCP connections (such as for Spamd) unless the daemon successfully set
702 Fast Open mode on its listening sockets. This fixes breakage seen on
703 too-old kernels or those not configured for Fast Open, at the cost of
704 requiring both directions being enabled for TFO, and TFO never being used
705 by non-daemon-related Exim processes.
707 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
708 endings, at least on the first header line. Try to canonify any that get
709 past that check, despite the cost.
711 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
712 now limited to an arbitrary five deep, while parsing addresses with the
713 strip_excess_angle_brackets option enabled.
715 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
716 instead leave the unprompted TLS handshake in socket buffer for the
717 TLS library to consume.
719 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
721 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
723 JH/16 Drop variables when they go out of scope. Memory management drops a whole
724 region in one operation, for speed, and this leaves assigned pointers
725 dangling. Add checks run only under the testsuite which checks all
726 variables at a store-reset and panics on a dangling pointer; add code
727 explicitly nulling out all the variables discovered. Fixes one known
728 bug: a transport crash, where a dangling pointer for $sending_ip_address
729 originally assigned in a verify callout, is re-used.
731 PP/06 Drop '.' from @INC in various Perl scripts.
733 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
735 PP/08 Reduce a number of compilation warnings under clang; building with
736 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
737 should be warning-free.
739 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
741 HS/01 Fix portability problems introduced by PP/08 for platforms where
742 realloc(NULL) is not equivalent to malloc() [SunOS et al].
744 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
745 chunk. This allows us to accept broken chunked messages. We need a more
746 general solution here.
748 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
749 already-broken messages in the queue.
751 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
753 JH/19 Fix reference counting bug in routing-generated-address tracking.
759 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
760 supports it and a size is available (ie. the sending peer gave us one).
762 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
763 years of being deprecated). The replacements are the ACLs
764 acl_smtp_mime and acl_not_smtp_mime.
766 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
767 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
768 or even an in-clear connection were permitted. Now, if the host lookup
769 was dnssec and dane was requested then the host is only used if the
770 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
771 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
772 if one fails this test.
773 This means that a poorly-configured remote DNS will make it incommunicado;
774 but it protects against a DNS-interception attack on it.
776 JH/04 Bug 1810: make continued-use of an open smtp transport connection
777 non-noisy when a race steals the message being considered.
779 JH/05 If main configuration option tls_certificate is unset, generate a
780 self-signed certificate for inbound TLS connections.
782 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
783 in rewrites and routers.
785 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
786 and logged a warning sing 4.83; now they are a configuration file error.
788 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
789 (lacking @domain). Apply the same qualification processing as RCPT.
791 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
793 JH/10 Support ${sha256:} applied to a string (as well as the previous
796 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
797 a cutthrough deliver is pending, as we always want to make a connection.
798 This also avoids re-routing the message when later placing the cutthrough
799 connection after a verify cache hit.
800 Do not update it with the verify result either.
802 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
803 when routing results in more than one destination address.
805 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
806 signing (which inhibits the cutthrough capability). Previously only
807 the presence of an option was tested; now an expansion evaluating as
808 empty is permissible (obviously it should depend only on data available
809 when the cutthrough connection is made).
811 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
812 the relevant preceding SMTP command did not note the pipelining mode.
814 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
815 Previously they were not counted.
817 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
818 as one having no matching records. Previously we deferred the message
819 that needed the lookup.
821 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
822 distinguished as "(=".
824 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
825 for missing MX records. Previously it only worked for missing A records.
827 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
829 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
830 after the data-go-ahead and data-ack. Patch from Jason Betts.
832 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
833 even for a "none" policy. Patch from Tony Meyer.
835 JH/22 Fix continued use of a connection for further deliveries. If a port was
836 specified by a router, it must also match for the delivery to be
839 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
840 When one of the recipients of a message was unsuitable for the connection
841 (has no matching addresses), we lost track of needing to mark it
842 deferred. As a result mail would be lost.
844 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
846 JH/25 Decoding ACL controls is now done using a binary search; the source code
847 takes up less space and should be simpler to maintain. Merge the ACL
848 condition decode tables also, with similar effect.
850 JH/26 Fix problem with one_time used on a redirect router which returned the
851 parent address unchanged. A retry would see the parent address marked as
852 delivered, so not attempt the (identical) child. As a result mail would
855 JH/27 Fix a possible security hole, wherein a process operating with the Exim
856 UID can gain a root shell. Credit to http://www.halfdog.net/ for
857 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
860 JH/28 Enable {spool,log} filesystem space and inode checks as default.
861 Main config options check_{log,spool}_{inodes,space} are now
862 100 inodes, 10MB unless set otherwise in the configuration.
864 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
865 Previously it only applied to the main-section connection policy
868 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
870 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
871 by me. Added RFC7919 DH primes as an alternative.
873 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
874 are not in the system include path.
876 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
877 GnuTLS, when a session startup failed (eg because the client disconnected)
878 Exim did stdio operations after fclose. This was exposed by a recent
879 change which nulled out the file handle after the fclose.
881 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
882 signed directly by the cert-signing cert, rather than an intermediate
883 OCSP-signing cert. This is the model used by LetsEncrypt.
885 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
887 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
888 an incoming connection.
890 HS/02 Bug 1802: Do not half-close the connection after sending a request
893 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
894 fallback to "prime256v1".
896 JH/34 SECURITY: Use proper copy of DATA command in error message.
897 Could leak key material. Remotely exploitable. CVE-2016-9963.
903 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
904 and 3.4.4 - once the server is enabled to respond to an OCSP request
905 it does even when not requested, resulting in a stapling non-aware
906 client dropping the TLS connection.
908 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
909 support variable-length bit vectors. No functional change.
911 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
912 The I= interface field on outgoing lines is now after the H= remote
913 host field, same as incoming lines. There is a separate
914 outgoing_interface log selector which allows you to disable the
917 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
918 If not running log_selector +smtp_connection the mainlog would be held
919 open indefinitely after a "too many connections" event, including to a
920 deleted file after a log rotate. Leave the per net connection logging
921 leaving it open for efficiency as that will be quickly detected by the
922 check on the next write.
924 HS/01 Bug 1671: Fix post transport crash.
925 Processing the wait-<transport> messages could crash the delivery
926 process if the message IDs didn't exist for some reason. When
927 using 'split_spool_directory=yes' the construction of the spool
928 file name failed already, exposing the same netto behaviour.
930 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
931 mime_regex ACL conditions.
933 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
934 to DSN fail messages (bounces): remote IP, remote greeting, remote response
935 to HELO, local diagnostic string.
937 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
938 log line to debug. Even when configured with a tls authenticator many
939 client connections are expected to not authenticate in this way, so
940 an authenticate fail is not an error.
942 HS/02 Add the Exim version string to the process info. This way exiwhat
943 gives some more detail about the running daemon.
945 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
946 matter for fast-change records such as DNSBLs.
948 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
949 retry record, even if constant. There may be multiple transports with
950 different interface settings and the retry behaviour needs to be kept
953 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
955 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
957 JH/10 Bug 840: fix log_defer_output option of pipe transport
959 JH/11 Bug 830: use same host for all RCPTS of a message, even under
960 hosts_randomize. This matters a lot when combined with mua_wrapper.
962 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
963 ${quote_pgsql:<string>} operator.
965 JH/13 Bug 1708: avoid misaligned access in cached lookup.
967 JH/14 Change header file name for freeradius-client. Relevant if compiling
968 with Radius support; from the Gentoo tree and checked under Fedora.
970 JH/15 Bug 1712: Introduce $prdr_requested flag variable
972 JH/16 Bug 1714: Permit an empty string as expansion result for transport
973 option transport_filter, meaning no filtering.
975 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
977 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
978 defaults to "*" (all hosts). The variable is now available when not built
979 with TLS, default unset, mainly to enable keeping the testsuite sane.
980 If a server certificate is not supplied (via tls_certificate) an error is
981 logged, and clients will find TLS connections fail on startup. Presumably
982 they will retry in-clear.
983 Packagers of Exim are strongly encouraged to create a server certificate
984 at installation time.
986 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
987 with the $config_file variable.
989 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
990 in transport context, after the attempt, and per-recipient. The latter type
991 is per host attempted. The event data is the error message, and the errno
992 information encodes the lookup type (A vs. MX) used for the (first) host,
993 and the trailing two digits of the smtp 4xx response.
995 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
996 to write to mainlog (or rejectlog, paniclog) in the window between file
997 creation and permissions/ownership being changed. Particularly affects
998 installations where exicyclog is run as root, rather than exim user;
999 result is that the running daemon panics and dies.
1001 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1003 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1004 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1005 "pri" and "weight". Note that the previous implicit priority given by the
1006 list order is no longer honoured.
1008 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1009 for DKIM processing.
1011 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1012 by defining SUPPORT_SOCKS.
1014 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1015 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1016 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1017 variables are renamed to proxy_{local,external}_{address,port}.
1019 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1020 for a build by defining SUPPORT_I18N
1022 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1023 of the query string, and make ${quote_redis:} do that quoting.
1025 JH/29 Move Events support from Experimental to mainline, enabled by default
1026 and removable for a build by defining DISABLE_EVENT.
1028 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1030 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1031 cached by the daemon.
1033 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1034 by defining LOOKUP_REDIS. The libhiredis library is required.
1036 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1037 keys are given for lookup.
1039 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1040 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1041 only supported when built with TLS support. The PolarSSL SHA routines
1042 are still used when the TLS library is too old for convenient support.
1044 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1045 openssl_options), for security. OpenSSL forces this from version 1.1.0
1046 server-side so match that on older versions.
1048 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1049 allocation for $value could be released as the expansion processing
1050 concluded, but leaving the global pointer active for it.
1052 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1053 and to use the domains and local_parts ACL conditions.
1055 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1056 incorrectly not doubled on cutthrough transmission, hence seen as a
1057 body-termination at the receiving system - resulting in truncated mails.
1058 Commonly the sender saw a TCP-level error, and retransmitted the message
1059 via the normal store-and-forward channel. This could result in duplicates
1060 received - but deduplicating mailstores were liable to retain only the
1061 initial truncated version.
1063 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1065 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1067 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1068 we're in there, support oversigning also; bug 1309.
1070 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1072 HS/04 Add support for keep_environment and add_environment options.
1074 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1075 either intentional arithmetic overflow during PRNG, or testing config-
1078 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1079 delivery resulted in actual delivery. Cancel cutthrough before DATA
1082 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1083 rejects a recipient: pass the reject to the originator.
1085 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1086 Many were false-positives and ignorable, but it's worth fixing the
1089 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1090 for the new environment-manipulation done at startup. Move the routines
1091 from being local to tls.c to being global via the os.c file.
1093 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1094 an extract embedded as result-arg for a map, the first arg for extract
1095 is unavailable so we cannot tell if this is a numbered or keyed
1096 extraction. Accept either.
1102 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1105 JH/02 The smtp transport option "multi_domain" is now expanded.
1107 JH/03 The smtp transport now requests PRDR by default, if the server offers
1110 JH/04 Certificate name checking on server certificates, when exim is a client,
1111 is now done by default. The transport option tls_verify_cert_hostnames
1112 can be used to disable this per-host. The build option
1113 EXPERIMENTAL_CERTNAMES is withdrawn.
1115 JH/05 The value of the tls_verify_certificates smtp transport and main options
1116 default to the word "system" to access the system default CA bundle.
1117 For GnuTLS, only version 3.0.20 or later.
1119 JH/06 Verification of the server certificate for a TLS connection is now tried
1120 (but not required) by default. The verification status is now logged by
1121 default, for both outbound TLS and client-certificate supplying inbound
1124 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1127 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1128 Status Notification (bounce) messages are now MIME format per RFC 3464.
1129 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1130 under the control of the dsn_advertise_hosts option, and routers may
1131 have a dsn_lasthop option.
1133 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1134 default, modifiable by a malware= option. The list separator for
1135 the options can now be changed in the usual way. Bug 68.
1137 JH/10 The smtp_receive_timeout main option is now expanded before use.
1139 JH/11 The incoming_interface log option now also enables logging of the
1140 local interface on delivery outgoing connections.
1142 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1143 if the interface and destination host and port all match.
1145 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1148 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1149 Patch from Andrew Lewis.
1151 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1152 now supports optional time-restrictions, weighting, and priority
1153 modifiers per server. Patch originally by <rommer@active.by>.
1155 JH/16 The spamd_address main option now supports a mixed list of local
1156 and remote servers. Remote servers can be IPv6 addresses, and
1157 specify a port-range.
1159 JH/17 Bug 68: The spamd_address main option now supports an optional
1160 timeout value per server.
1162 JH/18 Bug 1581: Router and transport options headers_add/remove can
1163 now have the list separator specified.
1165 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1168 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1171 JH/21 Support for the A6 type of dns record is withdrawn.
1173 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1174 rather than the verbs used.
1176 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1177 from 255 to 1024 chars.
1179 JH/24 Verification callouts now attempt to use TLS by default.
1181 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1182 are generic router options now. The defaults didn't change.
1184 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1185 Original patch from Alexander Shikoff, worked over by JH.
1187 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1188 files if ZCAT_COMMAND is not executable.
1190 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1192 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1194 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1195 Normally benign, it bites when the pair was led to by a CNAME;
1196 modern usage is to not canonicalize the domain to a CNAME target
1197 (and we were inconsistent anyway for A-only vs AAAA+A).
1199 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1201 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1202 when evaluating $sender_host_dnssec.
1204 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1205 $sender_helo_dnssec variable.
1207 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1209 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1211 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1213 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1214 documented as working, but never had. Support all but $spam_report.
1216 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1217 added for tls authenticator.
1219 HS/03 Add perl_taintmode main config option
1225 TL/01 When running the test suite, the README says that variables such as
1226 no_msglog_check are global and can be placed anywhere in a specific
1227 test's script, however it was observed that placement needed to be near
1228 the beginning for it to behave that way. Changed the runtest perl
1229 script to read through the entire script once to detect and set these
1230 variables, reset to the beginning of the script, and then run through
1231 the script parsing/test process like normal.
1233 TL/02 The BSD's have an arc4random API. One of the functions to induce
1234 adding randomness was arc4random_stir(), but it has been removed in
1235 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1236 function when detected.
1238 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1239 cause callback expansion.
1241 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1242 syntax errors in an expansion can be treated as a string instead of
1243 logging or causing an error, due to the internal use of bool_lax
1244 instead of bool when processing it.
1246 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1247 server certificates when making smtp deliveries.
1249 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1251 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1253 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1255 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1256 Merged patch from Sebastian Wiedenroth.
1258 JH/05 Fix results-pipe from transport process. Several recipients, combined
1259 with certificate use, exposed issues where response data items split
1260 over buffer boundaries were not parsed properly. This eventually
1261 resulted in duplicates being sent. This issue only became common enough
1262 to notice due to the introduction of connection certificate information,
1263 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1265 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1266 size buffer was used, resulting in syntax errors when an expansion
1269 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1270 version 3.3.6 or later.
1272 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1273 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1274 both become "event_action", the variables become $event_name, $event_data
1275 and $event_defer_errno. There is a new variable $verify_mode, usable in
1276 routers, transports and related events. The tls:cert event is now also
1277 raised for inbound connections, if the main configuration event_action
1280 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1281 early OCSP support, but no stapling (appears to be less than 1.0.0).
1283 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1284 server certificate names available under the smtp transport option
1285 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1288 JH/10 Time-related extraction expansions from certificates now use the main
1289 option "timezone" setting for output formatting, and are consistent
1290 between OpenSSL and GnuTLS compilations. Bug 1541.
1292 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1293 encoded parameter in the incoming message. Bug 1558.
1295 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1296 include certificate info, eximon was claiming there were spoolfile
1299 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1301 JH/14 Log delivery-related information more consistently, using the sequence
1302 "H=<name> [<ip>]" wherever possible.
1304 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1305 are problematic for Debian distribution, omit them from the release
1308 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1310 JH/16 Fix string representation of time values on 64bit time_t architectures.
1313 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1314 output list separator was used.
1319 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1320 checkers that were complaining about end of non-void function with no
1323 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1324 This was a regression introduced in 4.83 by another bugfix.
1326 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1328 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1329 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1335 TF/01 Correctly close the server side of TLS when forking for delivery.
1337 When a message was received over SMTP with TLS, Exim failed to clear up
1338 the incoming connection properly after forking off the child process to
1339 deliver the message. In some situations the subsequent outgoing
1340 delivery connection happened to have the same fd number as the incoming
1341 connection previously had. Exim would try to use TLS and fail, logging
1342 a "Bad file descriptor" error.
1344 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1345 utilities have not been installed.
1347 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1348 temporary space as the ACL may create new global variables.
1350 TL/01 LDAP support uses per connection or global context settings, depending
1351 upon the detected version of the libraries at build time.
1353 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1354 to extract and use the src ip:port in logging and expansions as if it
1355 were a direct connection from the outside internet. PPv2 support was
1356 updated based on HAProxy spec change in May 2014.
1358 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1360 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1361 Properly escape header and check for NULL return.
1363 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1366 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1368 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1369 characters in header names, implemented as a verify condition.
1370 Contributed by Michael Fischer v. Mollard.
1372 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1373 results permerror and temperror. Previous values are deprecated but
1374 still accepted. In a future release, err_perm and err_temp will be
1375 completely removed, which will be a backward incompatibility if the
1376 ACL tests for either of these two old results. Patch contributed by
1377 user bes-internal on the mailing list.
1379 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1381 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1382 selectors, in both main and reject logs.
1384 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1387 JH/07 Add malware type "sock" for talking to simple daemon.
1389 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1391 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1392 routers/transports under cutthrough routing.
1394 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1395 numbers. Touch up "bool" conditional to keep the same definition.
1397 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1399 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1401 JH/12 Expand items in router/transport headers_add or headers_remove lists
1402 individually rather than the list as a whole. Bug 1452.
1404 Required for reasonable handling of multiple headers_ options when
1405 they may be empty; requires that headers_remove items with embedded
1406 colons must have them doubled (or the list-separator changed).
1408 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1409 view the policy declared in the DMARC record. Currently, $dmarc_status
1410 is a combined value of both the record presence and the result of the
1413 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1415 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1416 dnslookup router and the smtp transport (applying to the forward
1419 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1420 of ldap servers used for a specific lookup. Patch provided by Heiko
1423 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1424 New variable $lookup_dnssec_authenticated for observability.
1426 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1427 Patch submitted by Lars Timman.
1429 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1431 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1432 Requires trusted mode and valid format message id, aborts otherwise.
1433 Patch contributed by Heiko Schlichting.
1435 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1436 certextract with support for various fields. Bug 1358.
1438 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1439 is requested by default, modifiable by smtp transport option
1442 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1443 operate on certificate variables to give certificate fingerprints
1444 Also new ${sha256:cert_variable}.
1446 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1448 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1449 Christian Aistleitner.
1451 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1453 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1454 file. Patch from Wolfgang Breyha.
1456 JH/25 Expand the coverage of the delivery $host and $host_address to
1457 client authenticators run in verify callout. Bug 1476.
1459 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1460 align with daemon_smtp_ports. Bug 72.
1462 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1463 support and error reporting did not work properly.
1465 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1466 and is readable. Patch from Andrew Colin Kissa.
1468 TL/14 Enhance documentation of ${run expansion and how it parses the
1469 commandline after expansion, particularly in the case when an
1470 unquoted variable expansion results in an empty value.
1472 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1474 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1475 by renaming away from it. Addresses GH issue 15, reported by
1478 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1480 TL/15 SECURITY: prevent double expansion in math comparison functions
1481 (can expand unsanitized data). Not remotely exploitable.
1488 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1490 PP/02 Make -n do something, by making it not do something.
1491 When combined with -bP, the name of an option is not output.
1493 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1496 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1497 $sender_host_name and config options to manage this, and basic check
1500 PP/05 DSCP support for outbound connections and control modifier for inbound.
1502 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1503 (Only plugin which currently uses this is kerberos4, which nobody should
1504 be using, but we should make it available and other future plugins might
1505 conceivably use it, even though it would break NAT; stuff *should* be
1506 using channel bindings instead).
1508 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1509 name; added for Sendmail compatibility; requires admin caller.
1510 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1511 just ignore it); requires trusted caller.
1512 Also parse but ignore: -Ac -Am -X<logfile>
1515 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1517 TL/02 Add +smtp_confirmation as a default logging option.
1519 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1520 Patch by Magnus Holmgren from 2007-02-20.
1522 TL/04 Bugzilla 1281 - Spec typo.
1523 Bugzilla 1283 - Spec typo.
1524 Bugzilla 1290 - Spec grammar fixes.
1526 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1528 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1530 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1531 mailing list by by Dmitry Isaikin.
1533 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1535 JH/02 Support "G" suffix to numbers in ${if comparisons.
1537 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1539 NM/01 Bugzilla 1197 - Spec typo
1540 Bugzilla 1196 - Spec examples corrections
1542 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1544 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1545 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1548 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1549 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1551 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1552 "acl {{name}{arg}...}", and optional args on acl condition
1555 JH/05 Permit multiple router/transport headers_add/remove lines.
1557 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1559 JH/07 Avoid using a waiting database for a single-message-only transport.
1560 Performance patch from Paul Fisher. Bugzilla 1262.
1562 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1565 JH/09 Add $headers_added variable, with content from use of ACL modifier
1566 add_header (but not yet added to the message). Bugzilla 199.
1568 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1569 Pulled from Bugzilla 817 by Wolfgang Breyha.
1571 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1573 (nb: this is the same fix as in Exim 4.80.1)
1575 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1578 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1581 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1583 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1584 advertise SMTP AUTH mechanism to us, instead of a generic
1585 protocol violation error. Also, make Exim more robust to bad
1586 data from the Dovecot auth socket.
1588 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1590 When a queue runner is handling a message, Exim first routes the
1591 recipient addresses, during which it prunes them based on the retry
1592 hints database. After that it attempts to deliver the message to
1593 any remaining recipients. It then updates the hints database using
1596 So if a recipient address works intermittently, it can get repeatedly
1597 deferred at routing time. The retry hints record remains fresh so the
1598 address never reaches the final cutoff time.
1600 This is a fairly common occurrence when a user is bumping up against
1601 their storage quota. Exim had some logic in its local delivery code
1602 to deal with this. However it did not apply to per-recipient defers
1603 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1605 This change adds a proper retry rule check during routing so that the
1606 final cutoff time is checked against the message's age. We only do
1607 this check if there is an address retry record and there is not a
1608 domain retry record; this implies that previous attempts to handle
1609 the address had the retry_use_local_parts option turned on. We use
1610 this as an approximation for the destination being like a local
1611 delivery, as in LMTP.
1613 I suspect this new check makes the old local delivery cutoff check
1614 redundant, but I have not verified this so I left the code in place.
1616 TF/02 Correct gecos expansion when From: is a prefix of the username.
1618 Test 0254 submits a message to Exim with the header
1622 When I ran the test suite under the user fanf2, Exim expanded
1623 the header to contain my full name, whereas it should have added
1624 a Resent-Sender: header. It erroneously treats any prefix of the
1625 username as equal to the username.
1627 This change corrects that bug.
1629 GF/01 DCC debug and logging tidyup
1630 Error conditions log to paniclog rather than rejectlog.
1631 Debug lines prefixed by "DCC: " to remove any ambiguity.
1633 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1635 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1636 Bug spotted by Jeremy Harris; was flawed since initial commit.
1637 Would have resulted in OCSP responses post-SNI triggering an Exim
1638 NULL dereference and crash.
1640 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1642 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1643 Bug detection, analysis and fix by Samuel Thibault.
1644 Bugzilla 1331, Debian bug #698092.
1646 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1648 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1649 Server implementation by Todd Lyons, client by JH.
1650 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1651 config variable "prdr_enable" controls whether the server
1652 advertises the facility. If the client requests PRDR a new
1653 acl_data_smtp_prdr ACL is called once for each recipient, after
1654 the body content is received and before the acl_smtp_data ACL.
1655 The client is controlled by both of: a hosts_try_prdr option
1656 on the smtp transport, and the server advertisement.
1657 Default client logging of deliveries and rejections involving
1658 PRDR are flagged with the string "PRDR".
1660 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1661 fclose(). Diagnosis by Todd Lyons.
1663 PP/17 Update configure.default to handle IPv6 localhost better.
1664 Patch by Alain Williams (plus minor tweaks).
1667 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1668 This is now consistent with GnuTLS, and is now documented: the
1669 previous undocumented portable approach to treating the option as
1670 unset was to force an expansion failure. That still works, and
1671 an empty string is now equivalent.
1673 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1674 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1675 not performing validation itself.
1677 PP/20 Added force_command boolean option to pipe transport.
1678 Patch from Nick Koston, of cPanel Inc.
1680 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1683 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1685 PP/21 Fix eximon continuous updating with timestamped log-files.
1686 Broken in a format-string cleanup in 4.80, missed when I repaired the
1687 other false fix of the same issue.
1688 Report and fix from Heiko Schlichting.
1691 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1692 Report from Prashanth Katuri.
1694 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1695 It's SecureTransport, so affects any MacOS clients which use the
1696 system-integrated TLS libraries, including email clients.
1698 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1699 using a MIME ACL for non-SMTP local injection.
1700 Report and assistance in diagnosis by Warren Baker.
1702 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1704 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1706 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1707 last id that failed so it may be referenced in subsequent ACL's.
1709 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1712 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1713 ldap library initialization, allowing self-signed CA's to be
1714 used. Also properly sets require_cert option later in code by
1715 using NULL (global ldap config) instead of ldap handle (per
1716 session). Bug diagnosis and testing by alxgomz.
1718 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1719 the src/util/ subdirectory.
1721 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1722 renamed to Transport Post Delivery Action by Jeremy Harris, as
1725 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1726 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1727 redis_servers = needs to be configured which will be used by the redis
1728 lookup. Patch from Warren Baker, of The Packet Hub.
1730 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1732 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1733 hostname or reverse DNS when processing a host list. Used suggestions
1734 from multiple comments on this bug.
1736 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1738 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1739 Missed a few lines, added it to make the runtest require no keyboard
1742 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1743 contains upper case chars. Make router use caseful_local_part.
1745 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1746 support when GnuTLS has been built with p11-kit.
1752 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1754 This, or similar/improved, will also be change PP/11 of 4.82.
1760 PP/01 Handle short writes when writing local log-files.
1761 In practice, only affects FreeBSD (8 onwards).
1762 Bugzilla 1053, with thanks to Dmitry Isaikin.
1764 NM/01 Bugzilla 949 - Documentation tweak
1766 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1769 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1771 PP/02 Implemented gsasl authenticator.
1773 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1775 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1776 `pkg-config foo` for cflags/libs.
1778 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1779 with rest of GSASL and with heimdal_gssapi.
1781 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1782 `pkg-config foo` for cflags/libs for the TLS implementation.
1784 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1785 properties get this fed in as external SSF. A number of robustness
1786 and debugging improvements to the cyrus_sasl authenticator.
1788 PP/08 cyrus_sasl server now expands the server_realm option.
1790 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1791 Patch by Jeremy Harris.
1793 PP/10 Added dbmjz lookup type.
1795 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1797 PP/12 MAIL args handles TAB as well as SP, for better interop with
1798 non-compliant senders.
1799 Analysis and variant patch by Todd Lyons.
1801 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1802 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1803 Patch from Dirk Mueller <dmueller@suse.com>
1805 PP/13 tls_peerdn now print-escaped for spool files.
1806 Observed some $tls_peerdn in wild which contained \n, which resulted
1807 in spool file corruption.
1809 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1810 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1811 or write after TLS renegotiation, which otherwise led to messages
1814 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1815 as a tracking header (ie: a signed header comes before the signature).
1816 Patch from Wolfgang Breyha.
1818 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1819 comma-sep list; embedded commas doubled.
1821 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1823 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1825 Report and patch from Dmitry Banschikov.
1827 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1828 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1829 failures. We appear to now support TLS1.1+ with Exim.
1831 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1832 lets Exim select keys and certificates based upon TLS SNI from client.
1833 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1834 before an outbound SMTP session. New log_selector, +tls_sni.
1836 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1837 NULL dereference. Report and patch from Alun Jones.
1839 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1840 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1841 Not seeing resolver debug output on NetBSD, but suspect this is a
1842 resolver implementation change.
1844 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1845 Left warnings. Added "eximon gdb" invocation mode.
1847 PP/21 Defaulting "accept_8bitmime" to true, not false.
1849 PP/22 Added -bw for inetd wait mode support.
1851 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1852 locate the relevant includes and libraries. Made this the default.
1854 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1855 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1857 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1858 This may cause build issues on older platforms.
1860 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1861 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1862 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1863 Added SNI support via GnuTLS too.
1864 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1866 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1868 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1869 Applied second patch from Janne, implementing suggestion to default
1870 multiple-strings-in-record handling to match SPF spec.
1872 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1874 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1875 read-only, out of scope).
1876 Patch from Wolfgang Breyha, report from Stuart Northfield.
1878 PP/29 Fix three issues highlighted by clang analyser static analysis.
1879 Only crash-plausible issue would require the Cambridge-specific
1880 iplookup router and a misconfiguration.
1881 Report from Marcin Mirosław.
1883 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1885 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1886 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1887 As part of this, removing so much warning spew let me fix some minor
1888 real issues in debug logging.
1890 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1891 assignment on my part. Fixed.
1893 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1894 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1895 Janne Snabb (who went above and beyond: thank you).
1897 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1898 string otherwise requires a connection and a bunch more work and it's
1899 relatively easy to get wrong. Should also expose TLS library linkage
1902 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1903 64-bit ${eval} (JH/03).
1905 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1906 GNU libc to support some of the 64-bit stuff, should not lead to
1907 conflicts. Defined before os.h is pulled in, so if a given platform
1908 needs to override this, it can.
1910 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1911 protection layer was required, which is not implemented.
1912 Bugzilla 1254, patch from Wolfgang Breyha.
1914 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1915 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1916 tls_dhparam take prime identifiers. Also unbreak combination of
1917 OpenSSL+DH_params+TLSSNI.
1919 PP/39 Disable SSLv2 by default in OpenSSL support.
1925 PP/01 Solaris build fix for Oracle's LDAP libraries.
1926 Bugzilla 1109, patch from Stephen Usher.
1928 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1930 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1933 TF/02 Fix a couple more cases where we did not log the error message
1934 when unlink() failed. See also change 4.74-TF/03.
1936 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1937 lock up or crash if it happened to be inside a call to libc when it
1938 got a SIGUSR1 from exiwhat.
1940 The SIGUSR1 handler appends the current process status to the process
1941 log which is later printed by exiwhat. It used to use the general
1942 purpose logging code to do this, but several functions it calls are
1943 not safe for signals.
1945 The new output code in the SIGUSR1 handler is specific to the process
1946 log, and simple enough that it's easy to inspect for signal safety.
1947 Removing some special cases also simplifies the general logging code.
1948 Removing the spurious timestamps from the process log simplifies
1951 TF/04 Improved ratelimit ACL condition.
1953 The /noupdate option has been deprecated in favour of /readonly which
1954 has clearer semantics. The /leaky, /strict, and /readonly update modes
1955 are mutually exclusive. The update mode is no longer included in the
1956 database key; it just determines when the database is updated. (This
1957 means that when you upgrade Exim will forget old rate measurements.)
1959 Exim now checks that the per_* options are used with an update mode that
1960 makes sense for the current ACL. For example, when Exim is processing a
1961 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1962 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1963 must specify per_mail/readonly. If you omit the update mode it defaults to
1964 /leaky where that makes sense (as before) or /readonly where required.
1966 The /noupdate option is now undocumented but still supported for
1967 backwards compatibility. It is equivalent to /readonly except that in
1968 ACLs where /readonly is required you may specify /leaky/noupdate or
1969 /strict/noupdate which are treated the same as /readonly.
1971 A useful new feature is the /count= option. This is a generalization
1972 of the per_byte option, so that you can measure the throughput of other
1973 aggregate values. For example, the per_byte option is now equivalent
1974 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1976 The per_rcpt option has been generalized using the /count= mechanism
1977 (though it's more complicated than the per_byte equivalence). When it is
1978 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1979 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1980 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1981 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1982 Note that using per_rcpt with a non-readonly update mode in more than
1983 one ACL will cause the recipients to be double-counted. (The per_mail
1984 and per_byte options don't have this problem.)
1986 The handling of very low rates has changed slightly. If the computed rate
1987 is less than the event's count (usually one) then this event is the first
1988 after a long gap. In this case the rate is set to the same as this event's
1989 count, so that the first message of a spam run is counted properly.
1991 The major new feature is a mechanism for counting the rate of unique
1992 events. The new per_addr option counts the number of different
1993 recipients that someone has sent messages to in the last time period. It
1994 behaves like per_rcpt if all the recipient addresses are different, but
1995 duplicate recipient addresses do not increase the measured rate. Like
1996 the /count= option this is a general mechanism, so the per_addr option
1997 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1998 example, measure the rate that a client uses different sender addresses
1999 with the options per_mail/unique=$sender_address. There are further
2000 details in the main documentation.
2002 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2004 TF/06 Removed a few PCRE remnants.
2006 TF/07 Automatically extract Exim's version number from tags in the git
2007 repository when doing development or release builds.
2009 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2010 Bugzilla 879. Patch from Paul Fisher.
2012 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2013 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2016 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2018 PP/05 Variable $av_failed, true if the AV scanner deferred.
2019 Bugzilla 1078. Patch from John Horne.
2021 PP/06 Stop make process more reliably on build failure.
2022 Bugzilla 1087. Patch from Heiko Schlittermann.
2024 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2025 Bugzilla 1089. Patch from Heiko Schlittermann.
2027 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2028 Bugzilla 1131. Patch from Holger Weiß.
2030 PP/09 Handle IPv6 addresses with SPF.
2031 Bugzilla 860. Patch from Wolfgang Breyha.
2033 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2035 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2038 PP/11 match_* no longer expand right-hand-side by default.
2039 New compile-time build option, EXPAND_LISTMATCH_RHS.
2040 New expansion conditions, "inlist", "inlisti".
2042 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2044 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2046 PP/14 fix log_write() format string regression from TF/03.
2047 Bugzilla 1152. Patch from Dmitry Isaikin.
2053 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2055 PP/02 Harmonised TLS library version reporting; only show if debugging.
2056 Layout now matches that introduced for other libraries in 4.74 PP/03.
2058 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2060 PP/04 New "dns_use_edns0" global option.
2062 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2065 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2066 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2068 TK/01 Updated PolarSSL code to 0.14.2.
2069 Bugzilla 1097. Patch from Andreas Metzler.
2071 PP/07 Catch divide-by-zero in ${eval:...}.
2072 Fixes bugzilla 1102.
2074 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2077 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2078 format-string attack -- SECURITY: remote arbitrary code execution.
2080 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2081 time unintentionally subject to list matching rules, letting the header
2082 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2083 arbitrary string expansion). This allowed for information disclosure.
2085 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2086 INT_MIN/-1 -- value coerced to INT_MAX.
2092 NM/01 Workaround for PCRE version dependency in version reporting
2095 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2096 This fixes portability to compilers other than gcc, notably
2097 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2099 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2100 makefiles for portability to HP-UX and POSIX correctness.
2102 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2103 Also via indented variable definition in the Makefile.
2104 (Debugging by Oliver Heesakkers).
2106 PP/02 Restore caching of spamd results with expanded spamd_address.
2107 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2109 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2110 Improves build reliability. Fix from: Frank Elsner
2112 NM/02 Fix wide character breakage in the rfc2047 coding
2113 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2115 NM/03 Allow underscore in dnslist lookups
2116 Fixes bug 1026. Patch from Graeme Fowler
2118 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2119 Code patches from Adam Ciarcinski of NetBSD.
2121 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2124 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2125 is logged, to avoid truncation. Patch from John Horne.
2127 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2128 Patch from Jakob Hirsch.
2130 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2131 SQL string expansion failure details.
2132 Patch from Andrey Oktyabrski.
2134 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2135 Patch from Simon Arlott.
2137 PP/09 New lookups functionality failed to compile on old gcc which rejects
2138 extern declarations in function scope.
2139 Patch from Oliver Fleischmann
2141 PP/10 Use sig_atomic_t for flags set from signal handlers.
2142 Check getgroups() return and improve debugging.
2143 Fixed developed for diagnosis in bug 927 (which turned out to be
2146 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2147 Patch from Mark Zealey.
2149 PP/12 Bugzilla 1056: Improved spamd server selection.
2150 Patch from Mark Zealey.
2152 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2153 Based on patch from Heiko Schlittermann.
2155 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2156 Patch from Uwe Doering, sign-off by Michael Haardt.
2158 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2159 variable declaration deep within a block. Bug and patch from
2162 PP/15 lookups-Makefile IRIX compatibility coercion.
2164 PP/16 Make DISABLE_DKIM build knob functional.
2166 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2167 Patch by Simon Arlott
2169 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2170 variable argument macros. Our copy now differs from upstream.
2176 TF/01 Failure to get a lock on a hints database can have serious
2177 consequences so log it to the panic log.
2179 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2180 controlled using the smtp_confirmation log selector.
2182 TF/03 Include the error message when we fail to unlink a spool file.
2184 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2185 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2186 for maintaining out-of-tree patches for some time.
2188 PP/01 Bugzilla 139: Documentation and portability issues.
2189 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2190 Handle per-OS dynamic-module compilation flags.
2192 PP/02 Let /dev/null have normal permissions.
2193 The 4.73 fixes were a little too stringent and complained about the
2194 permissions on /dev/null. Exempt it from some checks.
2195 Reported by Andreas M. Kirchwitz.
2197 PP/03 Report version information for many libraries, including
2198 Exim version information for dynamically loaded libraries. Created
2199 version.h, now support a version extension string for distributors
2200 who patch heavily. Dynamic module ABI change.
2202 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2203 privilege escalation vulnerability whereby the Exim run-time user
2204 can cause root to append content of the attacker's choosing to
2207 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2210 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2211 If dropping privileges for untrusted macros, we disabled normal logging
2212 on the basis that it would fail; for the Exim run-time user, this is not
2213 the case, and it resulted in successful deliveries going unlogged.
2214 Fixed. Reported by Andreas Metzler.
2220 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2221 only prepend for the Resent-* case. Fixes regression introduced in
2222 Exim 4.70 by NM/22 for Bugzilla 607.
2224 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2225 increasing numbers of administrators be bitten by this.
2227 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2229 PP/03 Bugzilla 994: added openssl_options main configuration option.
2231 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2233 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2235 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2237 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2238 without a peer certificate, leading to a segfault because of an
2239 assumption that peers always have certificates. Be a little more
2240 paranoid. Problem reported by Martin Tscholak.
2242 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2243 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2244 NB: ClamAV planning to remove STREAM in "middle of 2010".
2245 CL also introduces -bmalware, various -d+acl logging additions and
2246 more caution in buffer sizes.
2248 PP/09 Implemented reverse_ip expansion operator.
2250 PP/10 Bugzilla 937: provide a "debug" ACL control.
2252 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2254 PP/12 Bugzilla 973: Implement --version.
2256 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2258 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2260 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2262 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2263 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2264 ignore trailing whitespace.
2266 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2268 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2271 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2272 Notification from Dr Andrew Aitchison.
2274 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2275 ExtendedDetectionInfo response format.
2276 Notification from John Horne.
2278 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2281 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2282 XSL and documented dependency on system catalogs, with examples of how
2285 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2288 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2289 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2290 configuration file which is writeable by the Exim user or group.
2292 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2293 of configuration files to cover files specified with the -C option if
2294 they are going to be used with root privileges, not just the default
2297 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2298 option (effectively making it always true).
2300 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2301 files to be used while preserving root privileges.
2303 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2304 that rogue child processes cannot use them.
2306 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2307 run-time user, instead of root.
2309 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2310 Exim run-time user without dropping privileges.
2312 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2313 result string, instead of calling string_vformat() twice with the same
2316 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2317 for other users. Others should always drop root privileges if they use
2318 -C on the command line, even for a whitelisted configure file.
2320 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2322 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2328 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2329 $data_path, and $header_path variables; fixed documentation bugs and
2332 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2333 exipick to access non-standard spools, including the "frozen" queue
2336 NM/01 Bugzilla 965: Support mysql stored procedures.
2337 Patch from Alain Williams
2339 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2341 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2342 Patch from Andreas Metzler
2344 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2345 Patch from Kirill Miazine
2347 NM/05 Bugzilla 671: Added umask to procmail example.
2349 JJ/03 installed exipick 20100323.0, fixing doc bug
2351 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2352 directory. Notification and patch from Dan Rosenberg.
2354 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2356 TK/02 Improve log output when DKIM signing operation fails.
2358 MH/01 Treat the transport option dkim_domain as a colon separated
2359 list, not as a single string, and sign the message with each element,
2360 omitting multiple occurences of the same signer.
2362 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2363 Bugzilla 985, 986. Patch by Simon Arlott
2365 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2366 Patch by Simon Arlott
2368 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2369 MBX locking. Notification from Dan Rosenberg.
2375 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2377 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2379 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2381 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2383 NM/04 Bugzilla 915: Flag broken perl installation during build.
2389 TK/01 Added patch by Johannes Berg that expands the main option
2390 "spamd_address" if it starts with a dollar sign.
2392 TK/02 Write list of recipients to X-Envelope-Sender header when building
2393 the mbox-format spool file for content scanning (suggested by Jakob
2396 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2397 (http://www.dcc-servers.net/) support via dccifd. Activated by
2398 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2400 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2401 by Mark Daniel Reidel <mr@df.eu>.
2403 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2404 When building exim an external PCRE library is now needed -
2405 PCRE is a system library on the majority of modern systems.
2406 See entry on PCRE_LIBS in EDITME file.
2408 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2409 conversation. Added nologin parameter to request.
2410 Patch contributed by Kirill Miazine.
2412 TF/01 Do not log submission mode rewrites if they do not change the address.
2414 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2416 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2417 log files in place. Contributed by Roberto Lima.
2419 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2421 TF/03 Bugzilla 615: When checking the local_parts router precondition
2422 after a local_part_suffix or local_part_prefix option, Exim now
2423 does not use the address's named list lookup cache, since this
2424 contains cached lookups for the whole local part.
2426 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2427 Robert Millan. Documentation is in experimental-spec.txt.
2429 TF/04 Bugzilla 668: Fix parallel build (make -j).
2431 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2433 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2434 Patch provided by Jan Srzednicki.
2436 TF/05 Leading white space used to be stripped from $spam_report which
2437 wrecked the formatting. Now it is preserved.
2439 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2440 that they are available at delivery time.
2442 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2444 TF/08 TLS error reporting now respects the incoming_interface and
2445 incoming_port log selectors.
2447 TF/09 Produce a more useful error message if an SMTP transport's hosts
2448 setting expands to an empty string.
2450 NM/06 Bugzilla 744: EXPN did not work under TLS.
2451 Patch provided by Phil Pennock.
2453 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2454 Patch provided by Richard Godbee.
2456 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2457 acl_smtp_notquit, added index entry.
2459 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2460 Patch provided by Eugene Bujak.
2462 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2463 accept(). Patch provided by Maxim Dounin.
2465 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2466 Patch provided by Phil Pennock.
2468 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2470 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2471 Patch provided by Brad "anomie" Jorsch.
2473 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2474 Patch provided by Dean Brooks.
2476 TK/05 Add native DKIM support (does not depend on external libraries).
2478 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2479 Patch provided by Graeme Fowler.
2481 NM/16 Bugzilla 851: Documentation example syntax fix.
2483 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2485 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2488 NM/19 Bugzilla 745: TLS version reporting.
2489 Patch provided by Phil Pennock.
2491 NM/20 Bugzilla 167: bool: condition support.
2492 Patch provided by Phil Pennock.
2494 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2495 clients. Patch provided by Phil Pennock.
2497 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2498 Patch provided by Brad "anomie" Jorsch.
2500 NM/23 Bugzilla 687: Fix misparses in eximstats.
2501 Patch provided by Heiko Schlittermann.
2503 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2504 Patch provided by Heiko Schlittermann.
2506 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2507 plus update to original patch.
2509 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2511 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2512 Patch provided by David Brownlee.
2514 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2516 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2518 NM/30 Bugzilla 888: TLS documentation bugfixes.
2520 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2522 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2523 Unlike the original bugzilla I have changed all shell scripts in src tree.
2525 NM/33 Bugzilla 898: Transport filter timeout fix.
2526 Patch by Todd Rinaldo.
2528 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2529 Patch by Serge Demonchaux.
2531 NM/35 Bugzilla 39: Base64 decode bug fixes.
2532 Patch by Jakob Hirsch.
2534 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2536 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2538 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2540 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2546 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2547 ALT-N's libdkim that I have put here:
2548 http://duncanthrax.net/exim-experimental/
2550 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2551 were called 'true' and it seems that C99 defines that as a reserved
2552 keyword to be used with 'bool' variable types. That means you could
2553 not include C99-style headers which use bools without triggering
2554 build errors in sieve.c.
2556 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2557 as mailq or other aliases. Changed the --help handling significantly
2558 to do whats expected. exim_usage() emits usage/help information.
2560 SC/01 Added the -bylocaldomain option to eximstats.
2562 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2564 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2566 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2572 PH/01 Another patch from the Sieve maintainer.
2574 PH/02 When an IPv6 address is converted to a string for single-key lookup
2575 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2576 dots are used instead of colons so that keys in lsearch files need not
2577 contain colons. This was done some time before quoting was made available
2578 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2579 (notated using the quote facility) so as to distinguish them from IPv4
2580 keys. This meant that lookups for IP addresses in host lists did not work
2581 for iplsearch lookups.
2583 This has been fixed by arranging for IPv6 addresses to be expressed with
2584 colons if the lookup type is iplsearch. This is not incompatible, because
2585 previously such lookups could never work.
2587 The situation is now rather anomalous, since one *can* have colons in
2588 ordinary lsearch keys. However, making the change in all cases is
2589 incompatible and would probably break a number of configurations.
2591 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2594 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2595 conversion specification without a maximum field width, thereby enabling
2596 a rogue spamd server to cause a buffer overflow. While nobody in their
2597 right mind would setup Exim to query an untrusted spamd server, an
2598 attacker that gains access to a server running spamd could potentially
2599 exploit this vulnerability to run arbitrary code as the Exim user.
2601 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2602 $primary_hostname instead of what libspf2 thinks the hosts name is.
2604 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2605 a directory entry by the name of the lookup key. Previously, if a
2606 symlink pointed to a non-existing file or a file in a directory that
2607 Exim lacked permissions to read, a lookup for a key matching that
2608 symlink would fail. Now it is enough that a matching directory entry
2609 exists, symlink or not. (Bugzilla 503.)
2611 PH/03 The body_linecount and body_zerocount variables are now exported in the
2614 PH/04 Added the $dnslist_matched variable.
2616 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2617 This means they are set thereafter only if the connection becomes
2620 PH/06 Added the client_condition to authenticators so that some can be skipped
2621 by clients under certain conditions.
2623 PH/07 The error message for a badly-placed control=no_multiline_responses left
2624 "_responses" off the end of the name.
2626 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2628 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2629 (without spaces) instead of just copying the configuration text.
2631 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2633 PH/11 Added $max_received_linelength.
2635 PH/12 Added +ignore_defer and +include_defer to host lists.
2637 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2638 way in which PCRE > 7.0 is built.
2640 PH/14 Implemented queue_only_load_latch.
2642 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2643 MAIL command. The effect was to mangle the value on 64-bit systems.
2645 PH/16 Another patch from the Sieve maintainer.
2647 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2649 PH/18 If a system quota error occurred while trying to create the file for
2650 a maildir delivery, the message "Mailbox is full" was not appended to the
2651 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2652 only to a quota excession during the actual writing of the file.
2654 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2655 characters?) which causes problems in log lines. The DN values are now
2656 passed through string_printing() before being added to log lines.
2658 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2659 and InterBase are left for another time.)
2661 PH/21 Added message_body_newlines option.
2663 PH/22 Guard against possible overflow in moan_check_errorcopy().
2665 PH/23 POSIX allows open() to be a macro; guard against that.
2667 PH/24 If the recipient of an error message contained an @ in the local part
2668 (suitably quoted, of course), incorrect values were put in $domain and
2669 $local_part during the evaluation of errors_copy.
2675 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2676 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2679 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2680 issue a MAIL command.
2682 PH/02 In an ACL statement such as
2684 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2686 if a client was not listed at all, or was listed with a value other than
2687 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2688 the condition was not true (as it should be), so access was not denied.
2689 The bug was that the ! inversion was incorrectly passed on to the second
2690 item. This has been fixed.
2692 PH/03 Added additional dnslists conditions == and =& which are different from
2693 = and & when the dns lookup returns more than one IP address.
2695 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2696 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2698 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2699 FSYNC, which compiles an option called disable_fsync that allows for
2700 bypassing fsync(). The documentation is heavily laced with warnings.
2702 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2704 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2705 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2706 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2707 including adding "make clean"; (3) Added -fPIC when compiling the test
2708 dynamically loaded module, to get rid of a warning.
2710 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2711 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2712 = 0s. The bug is otherwise harmless.
2714 PH/07 There was a bug in the dovecot authenticator such that the value of
2715 $auth1 could be overwritten, and so not correctly preserved, after a
2716 successful authentication. This usually meant that the value preserved by
2717 the server_setid option was incorrect.
2719 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2721 PH/09 Installed PCRE release 7.0.
2723 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2724 run for batched SMTP input. It is now run at the start of every message
2725 in the batch. While fixing this I discovered that the process information
2726 (output by running exiwhat) was not always getting set for -bs and -bS
2727 input. This is fixed, and it now also says "batched" for BSMTP.
2729 PH/11 Added control=no_pipelining.
2731 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2732 patch, slightly modified), and move the expansion of helo_data till after
2733 the connection is made in the smtp transport (so it can use these
2736 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2738 PH/14 Added log_selector = +pid.
2740 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2742 PH/16 Add ${if forany and ${if forall.
2744 PH/17 Added dsn_from option to vary the From: line in DSNs.
2746 PH/18 Flush SMTP output before performing a callout, unless control =
2747 no_callout_flush is set.
2749 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2750 was true (the default) a successful delivery failed to delete the retry
2751 item, thus causing premature timeout of the address. The bug is now
2754 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2756 PH/21 Long custom messages for fakedefer and fakereject are now split up
2757 into multiline responses in the same way that messages for "deny" and
2758 other ACL rejections are.
2760 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2761 with slight modification.
2763 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2764 draft, changing the syntax and factoring some duplicate code".
2766 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2767 for deliveries of the second and subsequent messages over the same SMTP
2770 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2771 ${reduce, with only minor "tidies".
2773 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2775 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2776 expansion side effects.
2778 PH/27 When a message times out after an over-quota error from an Exim-imposed
2779 quota, the bounce message says "mailbox is full". This message was not
2780 being given when it was a system quota that was exceeded. It now should
2783 MH/03 Made $recipients available in local_scan(). local_scan() already has
2784 better access to the recipient list through recipients_list[], but
2785 $recipients can be useful in postmaster-provided expansion strings.
2787 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2788 in the case of a MAIL command with additional options following the
2789 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2790 were accidentally chopped off.
2792 PH/29 SMTP synchronization checks are implemented when a command is read -
2793 there is a check that no more input is waiting when there shouldn't be
2794 any. However, for some commands, a delay in an ACL can mean that it is
2795 some time before the response is written. In this time, more input might
2796 arrive, invalidly. So now there are extra checks after an ACL has run for
2797 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2798 pipelining has not been advertised.
2800 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2802 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2803 returned in the SMTP response when smtp_return_error_details was false.
2804 This has been fixed.
2806 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2807 instead of the C I/O that was originally supplied, because problems were
2808 reported on Solaris.
2810 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2811 Exim which did not show up earlier: it was assuming that a call to
2812 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2813 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2814 was a macro that became an assignment, so it seemed to work. This has
2815 changed to a proper function call with a void return, hence the compile
2816 error. Exim's code has been fixed.
2818 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2821 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2823 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2825 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2826 because it turns out to be too big or there is a timeout) there is no
2827 "Completed" line in the log. When some messages of this type were
2828 selected by exigrep, they were listed as "not completed". Others were
2829 picked up by some special patterns. I have improved the selection
2830 criteria to be more general.
2832 PH/38 The host_find_failed option in the manualroute router can now be set
2833 to "ignore", to completely ignore a host whose IP address cannot be
2834 found. If all hosts are ignored, the behaviour is controlled by the new
2835 host_all_ignored option.
2837 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2838 homing or because of multiple MX records with /mx) generated more than
2839 one IP address, and the following item turned out to be the local host,
2840 all the secondary addresses of the first item were incorrectly removed
2841 from the list, along with the local host and any following hosts (which
2842 is what is supposed to happen).
2844 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2845 whoever called Exim into the -H file. In the case of the daemon it was
2846 behaving confusingly. When first started, it used values for whoever
2847 started the daemon, but after a SIGHUP it used the Exim user (because it
2848 calls itself on a restart). I have changed the code so that it now always
2851 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2852 message are rejected with the same error (e.g. no authentication or bad
2853 sender address), and a DATA command is nevertheless sent (as can happen
2854 with PIPELINING or a stupid MUA), the error message that was given to the
2855 RCPT commands is included in the rejection of the DATA command. This is
2856 intended to be helpful for MUAs that show only the final error to their
2859 PH/42 Another patch from the Sieve maintainer.
2861 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2862 Eximstats - Fixed some broken HTML links and added missing column headers
2864 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2865 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2867 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2869 PH/43 Yet another patch from the Sieve maintainer.
2871 PH/44 I found a way to check for a TCP/IP connection going away before sending
2872 the response to the final '.' that terminates a message, but only in the
2873 case where the client has not sent further data following the '.'
2874 (unfortunately, this is allowed). However, in many cases there won't be
2875 any further data because there won't be any more messages to send. A call
2876 to select() can be used: if it shows that the input is "ready", there is
2877 either input waiting, or the socket has been closed. An attempt to read
2878 the next input character can distinguish the two cases. Previously, Exim
2879 would have sent an OK response which the client would never have see.
2880 This could lead to message repetition. This fix should cure that, at
2881 least in a lot of common cases.
2883 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2884 advertised in response to EHLO.
2890 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2891 fixed by 4.65/MH/01 (is this a record?) are fixed:
2893 (i) An empty string was always treated as zero by the numeric comparison
2894 operators. This behaviour has been restored.
2896 (ii) It is documented that the numeric comparison operators always treat
2897 their arguments as decimal numbers. This was broken in that numbers
2898 starting with 0 were being interpreted as octal.
2900 While fixing these problems I realized that there was another issue that
2901 hadn't been noticed. Values of message_size_limit (both the global option
2902 and the transport option) were treated as octal if they started with 0.
2903 The documentation was vague. These values are now always treated as
2904 decimal, and I will make that clear in the documentation.
2910 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2911 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2914 MH/01 Don't check that the operands of numeric comparison operators are
2915 integers when their expansion is in "skipping" mode (fixes bug
2916 introduced by 4.64-PH/07).
2918 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2919 child addresses, Exim now panics and dies. Previously, because the count
2920 is held in a short int, deliveries were likely to be lost. As such a
2921 large number of recipients for a single message is ridiculous
2922 (performance will be very, very poor), I have chosen to impose a limit
2923 rather than extend the field.
2929 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2930 leftover -K file (the existence of which was triggered by #402).
2931 While we were at it, introduced process PID as part of the -K
2932 filename. This should rule out race conditions when creating
2935 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2936 processing considerably. Previous code took too long for large mails,
2937 triggering a timeout which in turn triggers #401.
2939 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2940 in the DK code in transports.c. sendfile() is not really portable,
2941 hence the _LINUX specificness.
2943 TF/01 In the add_headers option to the mail command in an Exim filter,
2944 there was a bug that Exim would claim a syntax error in any
2945 header after the first one which had an odd number of characters
2948 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2949 callout verification, Exim cached a "reject" for the entire domain. This
2950 is correct for most verifications, but it is not correct for a recipient
2951 verification with use_sender or use_postmaster set, because in that case
2952 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2953 case of MAIL FROM:<> rejection from other early rejections (e.g.
2954 rejection of HELO). When verifying a recipient using a non-null MAIL
2955 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2956 Whatever the result of the callout, the value of the domain cache is
2957 left unchanged (for any other kind of callout, getting as far as trying
2958 RCPT means that the domain itself is ok).
2960 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2963 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2964 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2965 session, the daemon ignores ECONNECT errors and logs others; it now
2966 ignores EPIPE as well.
2968 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2969 (quoted-printable decoding).
2971 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2972 later the small subsequent patch to fix an introduced bug.
2974 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2976 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2978 PH/08 An error is now given if message_size_limit is specified negative.
2980 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2981 to be given (somewhat) arbitrary names.
2983 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2986 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2987 miscellaneous code fixes
2989 PH/10 Added the log_reject_target ACL modifier to specify where to log
2992 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2993 hostname. This is wrong, because it relates to the incoming message (and
2994 probably the interface on which it is arriving) and not to the outgoing
2995 callout (which could be using a different interface). This has been
2996 changed to use the value of the helo_data option from the smtp transport
2997 instead - this is what is used when a message is actually being sent. If
2998 there is no remote transport (possible with a router that sets up host
2999 addresses), $smtp_active_hostname is used.
3001 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3002 tweaks were necessary in order to get it to work (see also 21 below):
3003 (a) The code assumed that strncpy() returns a negative number on buffer
3004 overflow, which isn't the case. Replaced with Exim's string_format()
3006 (b) There were several signed/unsigned issues. I just did the minimum
3007 hacking in of casts. There is scope for a larger refactoring.
3008 (c) The code used strcasecmp() which is not a standard C function.
3009 Replaced with Exim's strcmpic() function.
3010 (d) The code set only $1; it now sets $auth1 as well.
3011 (e) A simple test gave the error "authentication client didn't specify
3012 service in request". It would seem that Dovecot has changed its
3013 interface. Fortunately there's a specification; I followed it and
3014 changed what the client sends and it appears to be working now.
3016 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3019 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3020 address A is aliased to B and C, where B exists and C does not. Without
3021 -v the output is "A verified" because verification stops after a
3022 successful redirection if more than one address is generated. However,
3023 with -v the child addresses are also verified. Exim was outputting "A
3024 failed to verify" and then showing the successful verification for C,
3025 with its parentage. It now outputs "B failed to verify", showing B's
3026 parentage before showing the successful verification of C.
3028 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3029 look up a TXT record in a specific list after matching in a combined
3032 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3033 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3034 they consult the DNS. I had assumed they would set it the way they
3035 wanted; and indeed my experiments on Linux seem to show that in some
3036 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3037 To be on the safe side, however, I have now made the interface to
3038 host_find_byname() similar to host_find_bydns(), with an argument
3039 containing the DNS resolver options. The host_find_byname() function now
3040 sets these options at its start, just as host_find_bydns() does. The smtp
3041 transport options dns_qualify_single and dns_search_parents are passed to
3042 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3043 of host_find_byname() use the default settings of RES_DEFNAMES
3044 (qualify_single) but not RES_DNSRCH (search_parents).
3046 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3047 spool_read_header() do less string testing, by means of a preliminary
3048 switch on the second character of optional "-foo" lines. (This is
3049 overdue, caused by the large number of possibilities that now exist.
3050 Originally there were few.) While I was there, I also converted the
3051 str(n)cmp tests so they don't re-test the leading "-" and the first
3052 character, in the hope this might squeeze out yet more improvement.
3054 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3055 flag allowing group syntax was set by the header_syntax check but not
3056 turned off, possible causing trouble later; (2) The flag was not being
3057 set at all for the header_verify test, causing "group"-style headers to
3058 be rejected. I have now set it in this case, and also caused header_
3059 verify to ignore an empty address taken from a group. While doing this, I
3060 came across some other cases where the code for allowing group syntax
3061 while scanning a header line wasn't quite right (mostly, not resetting
3062 the flag correctly in the right place). These bugs could have caused
3063 trouble for malformed header lines. I hope it is now all correct.
3065 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3066 with the "reply" argument non-NULL. The code, however (which originally
3067 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3068 but it didn't always do it. This confused somebody who was copying the
3069 code for some other use. I have removed all the tests.
3071 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3072 feature that was used to support insecure browsers during the U.S. crypto
3073 embargo. It requires special client support, and Exim is probably the
3074 only MTA that supported it -- and would never use it because real RSA is
3075 always available. This code has been removed, because it had the bad
3076 effect of slowing Exim down by computing (never used) parameters for the
3077 RSA_EXPORT functionality.
3079 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3080 authenticator to fail if there's a tab character in the incoming data
3081 (there should never be unless someone is messing about, as it's supposed
3082 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3083 if the connection is using TLS or if the remote IP is the same as the
3084 local IP, and the "valid-client-cert option" if a client certificate has
3087 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3088 authenticators. This can be used for authorization after authentication
3089 succeeds. (In the case of plaintext, it servers for both authentication
3092 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3093 if any retry times were supplied.
3095 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3096 connection, where there is no client IP address to check. In this
3097 situation, the verify now always succeeds.
3099 PH/25 Applied John Jetmore's -Mset patch.
3101 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3103 PH/27 In a string expansion for a processed (not raw) header when multiple
3104 headers of the same name were present, leading whitespace was being
3105 removed from all of them, but trailing whitespace was being removed only
3106 from the last one. Now trailing whitespace is removed from each header
3107 before concatenation. Completely empty headers in a concatenation (as
3108 before) are ignored.
3110 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3111 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3113 PH/29 [Removed. This was a change that I later backed out, and forgot to
3114 correct the ChangeLog entry (that I had efficiently created) before
3115 committing the later change.]
3117 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3118 address errors (4xx response to RCPT) over the same connection as other
3119 messages routed to the same hosts. Such deliveries are always "forced",
3120 so retry times are not inspected. This resulted in far too many retries
3121 for the affected addresses. The effect occurred only when there were more
3122 hosts than the hosts_max_try setting in the smtp transport when it had
3123 the 4xx errors. Those hosts that it had tried were not added to the list
3124 of hosts for which the message was waiting, so if all were tried, there
3125 was no problem. Two fixes have been applied:
3127 (i) If there are any address or message errors in an SMTP delivery, none
3128 of the hosts (tried or untried) are now added to the list of hosts
3129 for which the message is waiting, so the message should not be a
3130 candidate for sending over the same connection that was used for a
3131 successful delivery of some other message. This seems entirely
3132 reasonable: after all the message is NOT "waiting for some host".
3133 This is so "obvious" that I'm not sure why it wasn't done
3134 previously. Hope I haven't missed anything, but it can't do any
3135 harm, as the worst effect is to miss an optimization.
3137 (ii) If, despite (i), such a delivery is accidentally attempted, the
3138 routing retry time is respected, so at least it doesn't keep
3139 hammering the server.
3141 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3142 in ${readsocket because some servers need this prod.
3144 PH/32 Added some extra debug output when updating a wait-xxx database.
3146 PH/33 The hint "could be header name not terminated by colon", which has been
3147 given for certain expansion errors for a long time, was not being given
3148 for the ${if def:h_colon_omitted{... case.
3150 PH/34 The spec says: "With one important exception, whenever a domain list is
3151 being scanned, $domain contains the subject domain." There was at least
3152 one case where this was not true.
3154 PH/35 The error "getsockname() failed: connection reset by peer" was being
3155 written to the panic log as well as the main log, but it isn't really
3156 panic-worthy as it just means the connection died rather early on. I have
3157 removed the panic log writing for the ECONNRESET error when getsockname()
3160 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3161 runs only) independently of the message's sender address. This meant
3162 that, if the 4xx error was in fact related to the sender, a different
3163 message to the same recipient with a different sender could confuse
3164 things. In particular, this can happen when sending to a greylisting
3165 server, but other circumstances could also provoke similar problems.
3166 I have changed the default so that the retry time for these errors is now
3167 based a combination of the sender and recipient addresses. This change
3168 can be overridden by setting address_retry_include_sender=false in the
3171 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3172 remote server are returned as part of bounce messages. This was not
3173 happening for LMTP over a pipe (the lmtp transport), but now it is the
3174 same for both kinds of LMTP.
3176 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3177 in header lines that were in fact CNAMEs. This is no longer the case.
3179 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3180 and queue runs started by the daemon processed all messages. This has
3181 been fixed so that -R and -S can now usefully be given with -q<time>.
3183 PH/40 Import PCRE release 6.7 (fixes some bugs).
3185 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3187 PH/42 Give an error if -q is specified more than once.
3189 PH/43 Renamed the variables $interface_address and $interface_port as
3190 $received_ip_address and $received_port, to make it clear that these
3191 values apply to message reception, and not to the outgoing interface when
3192 a message is delivered. (The old names remain recognized, of course.)
3194 PH/44 There was no timeout on the connect() call when using a Unix domain
3195 socket in the ${readsocket expansion. There now is.
3197 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3198 be meaningful with "accept".
3200 SC/01 Eximstats V1.43
3201 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3203 SC/02 Eximstats V1.44
3204 Use a glob alias rather than an array ref in the generated
3205 parser. This improves both readability and performance.
3207 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3208 Collect SpamAssassin and rejection statistics.
3209 Don't display local sender or destination tables unless
3210 there is data to show.
3211 Added average volumes into the top table text output.
3213 SC/04 Eximstats V1.46
3214 Collect data on the number of addresses (recipients)
3215 as well as the number of messages.
3217 SC/05 Eximstats V1.47
3218 Added 'Message too big' to the list of mail rejection
3219 reasons (thanks to Marco Gaiarin).
3221 SC/06 Eximstats V1.48
3222 Mainlog lines which have GMT offsets and are too short to
3223 have a flag are now skipped.
3225 SC/07 Eximstats V1.49 (Alain Williams)
3226 Added the -emptyok flag.
3228 SC/08 Eximstats V1.50
3229 Fixes for obtaining the IP address from reject messages.
3231 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3232 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3233 whitespace changes from 4.64-PH/27
3235 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3238 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3239 are found, allow negative numbers in numeric criteria)
3241 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3243 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3244 to match changes made in 4.64-PH/43
3246 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3248 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3249 "domain" is actually an IP address, and force a failure. This locks out
3250 those revolvers/nameservers that support "A-for-A" lookups, in
3251 contravention of the specifications.
3253 PH/48 When a host name was looked up from an IP address, and the subsequent
3254 forward lookup of the name timed out, the host name was left in
3255 $sender_host_name, contrary to the specification.
3257 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3258 restricted to single-key lookups, Exim was not diagnosing an error if
3259 * or *@ was used with a query-style lookup.
3261 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3263 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3264 long ago, but noone interested enough thought of it. Let's just say that
3265 the "1.1" means that there are some new functions that weren't there at
3266 some point in the past.
3268 PH/51 Error processing for expansion failure of helo_data from an smtp
3269 transport during callout processing was broken.
3271 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3272 tested/used via the -bh/-bhc/-bs options.
3274 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3275 bug, fixed in subsequent PCRE releases).
3277 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3278 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3280 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3286 SC/01 Use a glob alias rather than an array ref in eximstats generated
3287 parser. This improves both readability and performance.
3289 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3290 Don't display local sender or destination tables in eximstats unless
3291 there is data to show.
3292 Added average volumes into the eximstats top table text output.
3294 SC/03 Collect data on the number of addresses (recipients) as well
3295 as the number of messages in eximstats.
3297 TF/01 Correct an error in the documentation for the redirect router. Exim
3298 does (usually) call initgroups() when daemonizing.
3300 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3301 with consistent privilege compared to when running as a daemon.
3303 TF/03 Note in the spec that $authenticated_id is not set for local
3304 submissions from trusted users.
3306 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3307 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3309 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3310 by adding some example configuration directives to the default
3311 configuration file. A little bit of work is required to uncomment the
3312 directives and define how usernames and passwords are checked, but
3313 there is now a framework to start from.
3315 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3316 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3317 without this. I don't know how relevant this is to other LDAP libraries.
3319 PH/02 Add the verb name to the "unknown ACL verb" error.
3321 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3323 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3325 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3326 directory not expanded when it should be if an expanded home directory
3327 was set for the address (which is overridden by the transport).
3329 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3332 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3333 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3334 because it is too late at that time, and has no effect.
3336 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3337 security issue with \' (bugzilla #107). I could not use the
3338 PQescapeStringConn() function, because it needs a PGconn value as one of
3341 PH/08 When testing addresses using -bt, indicate those final addresses that
3342 are duplicates that would not cause an additional delivery. At least one
3343 person was confused, thinking that -bt output corresponded to deliveries.
3344 (Suppressing duplicates isn't a good idea as you lose the information
3345 about possibly different redirections that led to the duplicates.)
3347 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3348 systems where poll() doesn't work, in particular OS X.
3350 PH/10 Added more information to debugging output for retry time not reached.
3352 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3353 operations in malware.c.
3355 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3358 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3359 syslog_duplication set false, log lines that would normally be written
3360 both the the main log and to the reject log were not written to syslog at
3363 PH/14 In the default configuration, change the use of "message" in ACL warn
3364 statements to "add_header".
3366 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3367 not followed by a command (e.g. "seen endif").
3369 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3370 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3373 PH/17 Added extra conditions to the default value of delay_warning_condition
3377 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3378 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3379 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3382 The Auto-Submitted: and various List- headers are standardised, whereas I
3383 don't think Precedence: ever was.
3385 PH/18 Refactored debugging code in route_finduser() to show more information,
3386 in particular, the error code if getpwnam() issues one.
3388 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3389 This is apparently needed in addition to the PH/07 change above to avoid
3390 any possible encoding problems.
3392 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3393 but not after initializing Perl.
3395 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3396 output them only if debugging. By default they are written stderr,
3397 apparently, which is not desirable.
3399 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3402 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3405 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3407 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3408 authenticated or an ident call has been made. Suppress the default
3409 values for $authenticated_id and $authenticated_sender (but permit -oMai
3410 and -oMas) when testing with -bh.
3412 PH/24 Re-jigged the order of the tests in the default configuration so that the
3413 tests for valid domains and recipients precede the DNS black list and CSA
3414 tests, on the grounds that those ones are more expensive.
3416 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3417 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3418 This bug exists in every version of Exim that I still have, right back to
3421 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3422 However, an attempt to turn on case-sensitivity in a regex key by
3423 including (?-i) didn't work because the subject string was already
3424 lowercased, and the effects were non-intuitive. It turns out that a
3425 one-line patch can be used to allow (?-i) to work as expected.
3431 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3432 other effects) broke the use of negated acl sub-conditions.
3434 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3437 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3438 "Deny" causes Exim to reject the incoming connection with a 554 error.
3439 Unfortunately, if there is a major crisis, such as a disk failure,
3440 tcp-wrappers gives "deny", whereas what one would like would be some
3441 kind of temporary error. A kludge has been added to help with this.
3442 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3443 554 error is used if errno is still zero or contains ENOENT (which occurs
3444 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3447 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3449 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3450 errors. Otherwise a message that provokes a temporary error (when other
3451 messages do not) can cause a whole host to time out.
3453 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3454 addresses were routed directly to files or pipes from a redirect router.
3455 File deliveries just didn't batch; pipe deliveries might have suffered
3458 PH/06 A failure to get a lock for a hints database would erroneously always say
3459 "Failed to get write lock", even when it was really a read lock.
3461 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3462 of 0600. This has been changed to use the value of the lockfile_mode
3463 option (which defaults to 0600).
3465 PH/08 Applied small patch from the Sieve maintainer.
3467 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3468 folder from quota calculations, a direct delivery into this folder messed
3469 up the contents of the maildirsize file. This was because the regex was
3470 used only to exclude .Trash (or whatever) when the size of the mailbox
3471 was calculated. There was no check that a delivery was happening into an
3472 excluded directory. This bug has been fixed by ignoring all quota
3473 processing for deliveries into excluded directories.
3475 PH/10 Added the maildirfolder_create_regex option to appendfile.
3481 PH/01 The code for finding all the local interface addresses on a FreeBSD
3482 system running IPv6 was broken. This may well have applied to all BSD
3483 systems, as well as to others that have similar system calls. The broken
3484 code found IPv4 interfaces correctly, but gave incorrect values for the
3485 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3486 that it would not match correctly against @[] and not recognize the IPv6
3489 PH/02 The ipliteral router was not recognizing addresses of the form user@
3490 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3492 PH/03 Added disable_ipv6.
3494 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3495 decoded form, because it is most often used to construct To: headers
3496 lines in autoreplies, and the decoded form may well be syntactically
3497 invalid. However, $reply_address has leading white space removed, and all
3498 newlines turned into spaces so that the autoreply transport does not
3501 PH/05 If group was specified without a user on a router, and no group or user
3502 was specified on a transport, the group from the router was ignored.
3504 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3505 for visible compile-time settings that can be used to change these
3506 numbers, for those that want even more. Backwards compatibility with old
3507 spool files has been maintained. However, going back to a previous Exim
3508 release will lost any variables that are in spool files.
3510 PH/07 Two small changes when running in the test harness: increase delay when
3511 passing a TCP/IP connection to a new process, in case the original
3512 process has to generate a bounce, and remove special handling of
3513 127.0.0.2 (sic), which is no longer necessary.
3515 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3516 be the same on different OS.
3518 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3521 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3522 whether --show-vars was specified or not
3524 JJ/02 exipick: Added support for new ACL variable spool format introduced
3527 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3528 syntactically invalid From: or Reply-to: line, and a filter used this to
3529 generate an autoreply, and therefore failed to obtain an address for the
3530 autoreply, Exim could try to deliver to a non-existent relative file
3531 name, causing unrelated and misleading errors. What now happens is that
3532 it logs this as a hard delivery error, but does not attempt to create a
3535 PH/11 The exinext utility has a -C option for testing purposes, but although
3536 the given file was scanned by exinext itself; it wasn't being passed on
3537 when Exim was called.
3539 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3540 an end-of-file indication when reading a command response.
3542 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3543 compiled. In many other places in Exim, IPv6 addresses are always
3544 recognized, so I have changed this. It also means that IPv4 domain
3545 literals of the form [IPV4:n.n.n.n] are now always recognized.
3547 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3548 used if the router is not running as root, for example, when verifying at
3549 ACL time, or when using -bh. The debugging output from this situation was
3550 non-existent - all you got was a failure to exec. I have made two
3553 (a) Failures to set uid/gid, the current directory, or a process leader
3554 in a subprocess such as that created by queryprogram now generate
3555 suitable debugging output when -d is set.
3557 (b) The queryprogram router detects when it is not running as root,
3558 outputs suitable debugging information if -d is set, and then runs
3559 the subprocess without attempting to change uid/gid.
3561 PH/15 Minor change to Makefile for building test_host (undocumented testing
3564 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3565 additional section of a DNS packet that returns MX or SRV records.
3566 Instead, it always explicitly searches for A/AAAA records. This avoids
3567 major problems that occur when a DNS server includes only records of one
3568 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3569 fixed another bug: if SRV records were looked up and the corresponding
3570 address records were *not* found in the additional section, the port
3571 values from the SRV records were lost.
3573 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3574 using the correct key (the original address) when searching the retry
3575 rules in order to find which one to use for generating the retry hint.
3577 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3578 adding the default one. Similarly, if it contains a Reply-To: header, the
3579 errors_reply_to option, if set, is not used.
3581 PH/19 When calculating a retry time, Exim used to measure the "time since
3582 failure" by looking at the "first failed" field in the retry record. Now
3583 it does not use this if it is later than than the arrival time of the
3584 message. Instead it uses the arrival time. This makes for better
3585 behaviour in cases where some deliveries succeed, thus re-setting the
3586 "first failed" field. An example is a quota failure for a huge message
3587 when small messages continue to be delivered. Without this change, the
3588 "time since failure" will always be short, possible causing more frequent
3589 delivery attempts for the huge message than are intended.
3590 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3592 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3593 $1, $2, $3) because the numerical variables can be reset during some
3594 expansion items (e.g. "match"), thereby losing the authentication data.
3596 PH/21 Make -bV show the size of off_t variables so that the test suite can
3597 decide whether to run tests for quotas > 2G.
3599 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3600 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3601 filecount value is greater than 2G or if a quota value is greater than 2G
3602 on a system where the size of off_t is not greater than 4, a panic error
3605 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3606 never match. The debug and -bh output now contains an explicit error
3607 message indicating a malformed IPv4 address or mask.
3609 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3610 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3611 PH/23 above applies.
3613 PH/25 Do not write to syslog when running in the test harness. The only
3614 occasion when this arises is a failure to open the main or panic logs
3615 (for which there is an explicit test).
3617 PH/26 Added the /no_tell option to "control=freeze".
3619 PH/27 If a host name lookup failed very early in a connection, for example, if
3620 the IP address matched host_lookup and the reverse lookup yielded a name
3621 that did not have a forward lookup, an error message of the form "no IP
3622 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3623 could be logged. Now it outputs the IP address instead of "NULL".
3625 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3626 allows the sender and the authenticated sender to be set when
3627 submitting a message from within Exim. Since child_open_exim() is
3628 documented for local_scan(), the new function should be too.
3630 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3631 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3632 results in an empty string is now treated as unset.
3634 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3636 PH/31 Added sender_verify_fail logging option.
3638 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3639 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3640 tidied the source and removed it altogether.
3642 PH/33 When a queue run was abandoned because the load average was too high, a
3643 log line was always written; now it is written only if the queue_run log
3644 selector is set. In addition, the log line for abandonment now contains
3645 information about the queue run such as the pid. This is always present
3646 in "start" and "stop" lines but was omitted from the "abandon" line.
3648 PH/34 Omit spaces between a header name and the colon in the error message that
3649 is given when verify = headers_syntax fails (if there are lots of them,
3650 the message gets confusing).
3652 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3653 names, as there are now some PTR records that contain slashes. This check
3654 is only to protect against broken name servers that fall over on strange
3655 characters, so the fact that it applies to all lookups doesn't matter.
3657 PH/36 Now that the new test suite is complete, we can remove some of the
3658 special code in Exim that was needed for the old test suite. For example,
3659 sorting DNS records because real resolvers return them in an arbitrary
3660 order. The new test suite's fake resolver always returns records in the
3663 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3664 bounces) except when queue_only is set, to avoid logging races between
3665 the different processes.
3667 PH/38 Panic-die if .include specifies a non-absolute path.
3669 PH/39 A tweak to the "H" retry rule from its user.
3671 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3672 a label. They prevented compilation on older perls.
3674 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3675 a warning to be raised on newish perls.
3677 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3678 on queue. Changes to match documented behaviour of showing count of
3679 messages matching specified criteria.
3681 PH/40 Changed the default ident timeout from 30s to 5s.
3683 PH/41 Added support for the use of login_cap features, on those BSD systems
3684 that have them, for controlling the resources used by pipe deliveries.
3686 PH/42 The content-scanning code uses fopen() to create files in which to put
3687 message data. Previously it was not paying any attention to the mode of
3688 the files. Exim runs with umask(0) because the rest of the code creates
3689 files with open(), and sets the required mode explicitly. Thus, these
3690 files were ending up world-writeable. This was not a big issue, because,
3691 being within the spool directory, they were not world-accessible. I have
3692 created a function called modefopen, which takes an additional mode
3693 argument. It sets umask(777), creates the file, chmods it to the required
3694 mode, then resets the umask. All the relevant calls to fopen() in the
3695 content scanning code have been changed to use this function.
3697 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3698 to 24 hours. This avoids potential overflow problems when processing G
3699 and H retry rules. I suspect nobody ever tinkers with this value.
3701 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3703 PH/45 When the plaintext authenticator is running as a client, the server's
3704 challenges are checked to ensure they are valid base64 strings. By
3705 default, the authentication attempt is cancelled if an invalid string is
3706 received. Setting client_ignore_invalid_base64 true ignores these errors.
3707 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3708 they are received. Thus, the responses can be made to depend on the
3709 challenges. If an invalid string is ignored, an empty string is placed in
3712 PH/46 Messages that are created by the autoreply transport now contains a
3713 References: header, in accordance with RFCs 2822 and 3834.
3715 PH/47 Added authenticated_sender_force to the smtp transport.
3717 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3719 PH/49 Installed latest patch from the Sieve maintainer.
3721 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3722 was also set, the appendfile transport was unnecessarily scanning a
3723 directory of message files (e.g. for maildir delivery) to find the count
3724 of files (along with the size), even though it did not need this
3725 information. It now does the scan only if it needs to find either the
3726 size of the count of files.
3728 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3730 PH/52 Two bugs concerned with error handling when the smtp transport is
3733 (i) Exim was not creating retry information for temporary errors given
3734 for individual recipients after the DATA command when the smtp transport
3735 was used in LMTP mode. This meant that they could be retried too
3736 frequently, and not timed out correctly.
3738 (ii) Exim was setting the flag that allows error details to be returned
3739 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3740 recipients that were returned after the DATA command.
3742 PH/53 This is related to PH/52, but is more general: for any failing address,
3743 when detailed error information was permitted to be returned to the
3744 sender, but the error was temporary, then after the final timeout, only
3745 "retry timeout exceeded" was returned. Now it returns the full error as
3746 well as "retry timeout exceeded".
3748 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3749 do this, and (what is worse) MTAs that accept it.
3751 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3752 will now be deprecated.
3754 PH/56 New os.c-cygwin from the Cygwin maintainer.
3756 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3757 formats (previously only available in exim formats via -bpr, -bpru,
3758 and -bpra. Now also available in native and exiqgrep formats)
3760 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3761 with very large, slow to parse queues
3763 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3765 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3767 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3768 responses to authentication challenges, though it was showing the
3769 challenges; (ii) I've removed the CR characters from the debug output for
3772 PH/58 Allow for the insertion of a newline as well as a space when a string
3773 is turned into more than one encoded-word during RFC 2047 encoding. The
3774 Sieve code now uses this.
3776 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3777 data_4xx, lost_connection, tls_required.
3779 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3780 message was being sent as an SMTP response.
3782 PH/61 Add -l and -k options to exicyclog.
3784 PH/62 When verifying, if an address was redirected to one new address, so that
3785 verification continued, and the new address failed or deferred after
3786 having set something in $address_data, the value of $address_data was not
3787 passed back to the ACL. This was different to the case when no
3788 redirection occurred. The value is now passed back in both cases.
3790 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3791 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3792 use login_cap.h, so on its own it isn't the distinguishing feature. The
3793 new name refers directly to the setclassresources() function.
3795 PH/65 Added configuration files for NetBSD3.
3797 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3799 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3800 is preferred over IPv4.
3802 PH/68 The bounce_return_message and bounce_return_body options were not being
3803 honoured for bounces generated during the reception of non-SMTP messages.
3804 In particular, this applied to messages rejected by the ACL. This bug has
3805 been fixed. However, if bounce_return_message is true and bounce_return_
3806 body is false, the headers that are returned for a non-SMTP message
3807 include only those that have been read before the error was detected.
3808 (In the case of an ACL rejection, they have all been read.)
3810 PH/69 The HTML version of the specification is now built in a directory called
3811 spec_html instead of spec.html, because the latter looks like a path with
3812 a MIME-type, and this confuses some software.
3814 PH/70 Catch two compiler warnings in sieve.c.
3816 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3817 function verify_get_ident() calls ip_connect() to connect a socket, but
3818 if the "connect()" function timed out, ip_connect() used to close the
3819 socket. However, verify_get_ident() also closes the socket later, and in
3820 between Exim writes to the log, which may get opened at this point. When
3821 the socket was closed in ip_connect(), the log could get the same file
3822 descriptor number as the socket. This naturally causes chaos. The fix is
3823 not to close the socket in ip_connect(); the socket should be closed by
3824 the function that creates it. There was only one place in the code where
3825 this was missing, in the iplookup router, which I don't think anybody now
3826 uses, but I've fixed it anyway.
3828 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3829 well as to direct DNS lookups. Otherwise the handling of names in host
3830 lists is inconsistent and therefore confusing.
3836 PH/01 Two changes to the default runtime configuration:
3838 (1) Move the checks for relay_from_hosts and authenticated clients from
3839 after to before the (commented out) DNS black list checks.
3841 (2) Add control=submission to the relay_from_hosts and authenticated
3842 clients checks, on the grounds that messages accepted by these
3843 statements are most likely to be submissions.
3845 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3847 (1) Generate an error if the third argument for the ${prvs expansion is
3850 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3853 (3) Reset the variables that are obtained from the first argument of
3854 ${prvscheck and used in the second argument before leaving the code,
3855 because their memory is reclaimed, so using them afterwards may do
3858 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3859 one (it's much easier than Tom thought :-).
3861 (5) Because of (4), we can now allow for the use of $prvscheck_result
3862 inside the third argument.
3864 PH/03 For some reason, the default setting of PATH when running a command from
3865 a pipe transport was just "/usr/bin". I have changed it to
3868 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3869 anything to be listed in the output from -bV.
3871 PH/05 When a filter generated an autoreply, the entire To: header line was
3872 quoted in the delivery log line, like this:
3874 => >A.N.Other <ano@some.domain> <original@ddress> ...
3876 This has been changed so that it extracts the operative address. There
3877 may be more than one such address. If so, they are comma-separated, like
3880 => >ano@some.domain,ona@other.domain <original@ddress> ...
3882 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3883 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3884 not being looked up in the rDNS to get a host name, Exim was showing the
3885 IP address twice in Received: lines, even though the IP addresses were
3886 identical. For example:
3888 Received: from [1.2.3.4] (helo=[1.2.3.4])
3890 However, if the real host name was known, it was omitting the HELO data
3891 if it matched the actual IP address. This has been tidied up so that it
3892 doesn't show the same IP address twice.
3894 PH/07 When both +timestamp and +memory debugging was on, the value given by
3895 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3896 called by the debug printing, thereby overwriting the timestamp buffer.
3897 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3899 PH/08 When the original message was included in an autoreply transport, it
3900 always said "this is a copy of the message, including all the headers",
3901 even if body_only or headers_only was set. It now gives an appropriate
3904 PH/09 Applied a patch from the Sieve maintainer which:
3906 o fixes some comments
3907 o adds the (disabled) notify extension core
3908 o adds some debug output for the result of if/elsif tests
3909 o points to the current vacation draft in the documentation
3910 and documents the missing references header update
3914 o fixes a bug in processing the envelope test (when testing
3915 multiple envelope elements, the last element determined the
3918 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3919 Electronic Mail") by including:
3921 Auto-submitted: auto-generated
3923 in the messages that it generates (bounce messages and others, such as
3924 warnings). In the case of bounce messages for non-SMTP messages, there was
3925 also a typo: it was using "Auto_submitted" (underscore instead of
3926 hyphen). Since every message generated by Exim is necessarily in response
3927 to another message, thes have all been changed to:
3929 Auto-Submitted: auto-replied
3931 in accordance with these statements in the RFC:
3933 The auto-replied keyword:
3935 - SHOULD be used on messages sent in direct response to another
3936 message by an automatic process,
3938 - MUST NOT be used on manually-generated messages,
3940 - MAY be used on Delivery Status Notifications (DSNs) and Message
3941 Disposition Notifications (MDNs),
3943 - MUST NOT be used on messages generated by automatic or periodic
3944 processes, except for messages which are automatic responses to
3947 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3948 to the default Received: header definition.
3950 PH/12 Added log selector acl_warn_skipped (default on).
3952 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3953 variables because (a) they are in the wrong storage pool and (b) even if
3954 they were copied, it wouldn't work properly because of the caching.
3956 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3957 checking when decoding. Apparently there are clients that generate
3958 overlong encoded strings. Why am I not surprised?
3960 PH/15 If the first argument of "${if match_address" was not empty, but did not
3961 contain an "@" character, Exim crashed. Now it writes a panic log message
3962 and treats the condition as false.
3964 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3966 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3967 extension "envelope-auth". The code is finished and in agreement with
3968 other implementations, but there is no documentation so far and in fact,
3969 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3970 not changing the active code.
3972 Print executed "if" and "elsif" statements when debugging is used. This
3973 helps a great deal to understand what a filter does.
3975 Document more things not specified clearly in RFC3028. I had all this
3976 sorted out, when out of a sudden new issues came to my mind. Oops."
3978 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3981 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3982 canonical form (as documented). However, after a host name lookup from
3983 the IP address, check_host() was doing a simple string comparison with
3984 addresses acquired from the DNS when checking that the found name did
3985 have the original IP as one of its addresses. Since any found IPv6
3986 addresses are likely to be in abbreviated form, the comparison could
3987 fail. Luckily, there already exists a function for doing the comparison
3988 by converting both addresses to binary, so now that is used instead of
3989 the text comparison.
3991 PH/20 There was another similar case to PH/19, when a complete host name was
3992 given in a host list; looking up its IP address could give an abbreviated
3993 form, whereas the current host's name might or might not be abbreviated.
3994 The same fix has been applied.
4000 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4001 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4004 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4005 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4007 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4009 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4010 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4011 Also, for "Auto-Submitted", treat anything other than "no" as
4012 non-personal, in accordance with RFC 3834. (Previously it treated
4013 anything starting "auto-" as non-personal.)
4015 TF/01 The control=submission/name=... option had a problem with syntax
4016 errors if the name included a slash character. The /name= option
4017 now slurps the rest of the string, so it can include any characters
4018 but it must come last in the list of options (after /sender_retain
4021 PH/05 Some modifications to the interface to the fake nameserver for the new
4029 TK/01 Added the "success_on_redirect" address verification option. See
4030 NewStuff for rationale and an example.
4032 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4034 PH/02 Patch to exigrep to allow it to work on syslog lines.
4036 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4037 fread() to skip over the body file's header line, because in Cygwin the
4038 header line is locked and is inaccessible.
4040 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4041 co-exist for some time) to make it clear that it is the Exim ID that is
4042 referenced, not the Message-ID: header line.
4044 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4045 string_format() function, because snprintf() does not exist on all
4048 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4049 ${stat: expansion item.
4051 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4052 protocol synchronization error", to keep the pedants happy.
4054 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4055 well as for IRIX systems, when gcc is being used. See the host.c source
4058 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4060 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4063 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4064 transport and to the smtp transport in LMTP mode.
4066 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4068 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4069 run a filter in a subprocess. This could lead to confusion in subsequent
4070 lookups in the parent process. There should also be a search_tidyup() at
4071 the end of the subprocess.
4073 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4074 only if the host matched helo_try_verify_hosts, which caused the
4075 verification to occur when the EHLO/HELO command was issued. The ACL just
4076 tested the remembered result. Now, if a previous verification attempt has
4077 not happened, "verify = helo" does it there and then.
4079 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4081 TK/03 Fix log output including CR from clamd.
4083 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4084 exist provoked a memory error which could cause a segfault.
4086 PH/15 Installed PCRE 6.2
4088 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4090 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4091 of the problem. Specifically, suggested +O2 rather than +O1 for the
4094 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4096 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4097 record, the port was not correctly logged when the outgoing_port log
4098 selector was set (it logged the transort's default port).
4100 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4101 fallback_hosts, and "hosts" in the smtp transport.
4103 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4104 host errors such as "Connection refused".
4106 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4107 authentication with radiusclient 0.4.9:
4109 - Error returned from rc_read_config was caught wrongly
4110 - Username/password not passed on to radius server due to wrong length.
4112 The presumption is that some radiusclient API changes for 4.51/PH/17
4113 were not taken care of correctly. The code is still untested by me (my
4114 Linux distribution still has 0.3.2 of radiusclient), but it was
4115 contributed by a Radius user.
4117 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4118 expanding the "port" option of the smtp transport.
4120 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4121 while reading a MIME header. Thanks to Tom Hughes for a patch.
4123 PH/24 Include config.h inside local_scan.h so that configuration settings are
4126 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4127 that in an ACL for RCPT (for example), you can examine exactly what was
4130 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4131 commands, but it was not correctly comparing the address with the actual
4132 client host address. Thus, it would show the EHLO address in Received:
4133 header lines when this was not necessary.
4135 PH/27 Added the % operator to ${eval:}.
4137 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4138 it should be ignoring failures (because with -C, for example, it has lost
4139 privilege). It wasn't ignoring creation failures other than "already
4142 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4145 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4146 because an input line was too long, either on its own, or by virtue of
4147 too many continuations, the temporary file was not being removed, and the
4148 return code was incorrect.
4150 PH/31 Missing "BOOL" in function definition in filtertest.c.
4152 PH/32 Applied Sieve patches from the maintainer.
4154 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4156 PH/33 Added "verify = not_blind".
4158 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4159 Local/Makefile (with some defaults set). These are used in built scripts
4160 such as exicyclog, but they have never been used in the exim_install
4161 script (though there are many overriding facilities there). I have
4162 arranged that the exim_install script now takes note of these two
4165 PH/35 Installed configuration files for Dragonfly.
4167 PH/36 When a locally submitted message by a trusted user did not contain a
4168 From: header, and the sender address was obtained from -f or from an SMTP
4169 MAIL command, and the trusted user did not use -F to supply a sender
4170 name, $originator_name was incorrectly used when constructing a From:
4171 header. Furthermore, $originator_name was used for submission mode
4172 messages from external hosts without From: headers in a similar way,
4173 which is clearly wrong.
4175 PH/37 Added control=suppress_local_fixups.
4177 PH/38 When log_selector = +received_sender was set, and the addition of the
4178 sender made the log line's construction buffer exactly full, or one byte
4179 less than full, an overflow happened when the terminating "\n" was
4182 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4183 when the result of a list match is failure because a DNS lookup failed.
4185 PH/40 RM_COMMAND is now used in the building process.
4187 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4188 the "build-* directories that it finds.
4190 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4191 address in a domain literal was a prefix of an interface address.
4193 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4194 when verifying a sender address, unless rewrite_headers is false.
4196 PH/44 Wrote a long comment about why errors_to addresses are verified as
4197 recipients, not senders.
4199 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4200 the ratelimit ACL was added.
4202 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4204 PH/46 Added extra information about PostgreSQL errors to the error string.
4206 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4207 avoiding the need to install special zones in a real server. This is
4208 backwards compatible; if it can't find the fake resolver, it drops back.
4209 Thus, both old and new test suites can be run.
4211 TF/02 Added util/ratelimit.pl
4213 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4214 clock is set back in time.
4216 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4217 Candler <B.Candler@pobox.com>.
4219 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4220 OK for addresses that are the result of redirections.
4222 PH/48 A number of further additions for the benefit of the new test suite,
4223 including a fake gethostbyname() that interfaces to the fake DNS resolver
4226 TF/06 The fix for widen_domains has also been applied to qualify_single and
4227 search_parents which are the other dnslookup options that can cause
4230 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4233 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4235 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4236 service" if it encountered an MX record with an empty target hostname.
4237 The message is now "an MX or SRV record indicated no SMTP service".
4239 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4240 if the DNS of the sending site is misconfigured. This is quite a
4241 common situation. This change restores the behaviour of treating a
4242 helo verification defer as a failure.
4244 PH/51 If self=fail was set on a router, the bounce message did not include the
4245 actual error message.
4251 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4253 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4254 command itself ended in error, the underlying message about the transport
4255 filter timeout was being overwritten with the pipe command error. Now the
4256 underlying error message should be appended to the second error message.
4258 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4260 PH/02 Reduce the amount of output that "make" produces by default. Full output
4261 can still be requested.
4263 PH/03 The warning log line about a condition test deferring for a "warn" verb
4264 was being output only once per connection, rather than after each
4265 occurrence (because it was using the same function as for successful
4266 "warn" verbs). This seems wrong, so I have changed it.
4268 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4269 it should not have, which might have caused a crash in the right
4270 circumstances, but probably never did.
4272 PH/04 Installed a modified version of Tony Finch's patch to make submission
4273 mode fix the return path as well as the Sender: header line, and to
4274 add a /name= option so that you can make the user's friendly name appear
4277 TF/03 Added the control = fakedefer ACL modifier.
4279 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4280 Mark Lowes for thorough testing.
4282 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4284 TK/03 Merged latest SRS patch from Miles Wilton.
4286 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4287 with the definition in sysexits.h (which is #included earlier).
4288 Fortunately, Exim does not actually use EX_OK. The code used to try to
4289 preserve the sysexits.h value, by assuming that macro definitions were
4290 scanned for macro replacements. I have been disabused of this notion,
4291 so now the code just undefines EX_OK before #including unistd.h.
4293 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4294 in the smtp transport. When a block could not be written in a single
4295 write() function, the timeout was being re-applied to each part-write.
4296 This seems wrong - if the receiver was accepting one byte at a time it
4297 would take for ever. The timeout is now adjusted when this happens. It
4298 doesn't have to be particularly precise.
4300 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4301 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4303 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4304 without a domain if the check to <postmaster@domain> fails.
4306 SC/01 Eximstats: added -xls and the ability to specify output files
4307 (patch written by Frank Heydlauf).
4309 SC/02 Eximstats: use FileHandles for outputting results.
4311 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4313 SC/04 Eximstats: fixed display of large numbers with -nvr option
4315 SC/05 Eximstats: fixed merging of reports with empty tables.
4317 SC/06 Eximstats: added the -include_original_destination flag
4319 SC/07 Eximstats: removed tabs and trailing whitespace.
4321 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4323 TK/06 MBOX spool code: Add real "From " MBOX separator line
4324 so the .eml file is really in mbox format (even though
4325 most programs do not really care). Patch from Alex Miller.
4327 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4328 The latter is generated from $received_to and is only set if the
4329 message has one envelope recipient. SA can use these headers,
4330 obviously out-of-the-box. Patch from Alex Miller.
4332 PH/08 The ${def test on a variable was returning false if the variable's
4333 value was "0", contrary to what the specification has always said!
4334 The result should be true unless the variable is empty.
4336 PH/09 The syntax error of a character other than { following "${if
4337 def:variable_name" (after optional whitespace) was not being diagnosed.
4338 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4339 accidental colon was present, for example, could give incorrect results.
4341 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4342 result is used (not including appendfile, where other changes are about
4345 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4346 This involved changing a lot of size variables from int to off_t. It
4347 should work with maildirs and everything.
4349 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4350 spamd dying while we are connected to it.
4352 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4355 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4356 function for BDB 4.3.
4358 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4360 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4361 the sources for PCRE have been re-arranged and more files are now
4364 PH/15 The code I had for printing potentially long long variables in PH/11
4365 above was not the best (it lost precision). The length of off_t variables
4366 is now inspected at build time, and an appropriate printing format (%ld
4367 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4368 to be "long long int" or "long int". This is needed for the internal
4369 formatting function string_vformat().
4371 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4372 the configuration file to be ":syslog", then the script "guesses" where
4373 the logs files are, rather than using the compiled in default. In our
4374 case the guess is not the same as the compiled default, so the script
4375 suddenly stopped working when I started to use syslog. The patch checks
4376 to see if log_file_path is "". If so, it attempts to read it from exim
4377 with no configuration file to get the compiled in version, before it
4378 falls back to the previous guessing code."
4380 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4381 implementing BATV in an Exim configuration. See NewStuff for the gory
4384 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4385 Makefile that are specific to HP-UX.
4387 PH/18 If the "use_postmaster" option was set for a recipient callout together
4388 with the "random" option, the postmaster address was used as the MAIL
4389 FROM address for the random test, but not for the subsequent recipient
4390 test. It is now used for both.
4392 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4393 patch removes a few documentation additions to RFC 3028, because the
4394 latest draft now contains them. It adds the new en;ascii-case comparator
4395 and a new error check for 8bit text in MIME parts. Comparator and
4396 require names are now matched exactly. I enabled the subaddress
4397 extension, but it is not well tested yet (read: it works for me)."
4399 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4400 rework some of the code of TK/09 above to avoid the hardwired use of
4401 "%lld" and "long long". Replaced the call to snprintf() with a call to
4404 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4405 records point to non-existent hosts", "retry timeout exceeded", and
4406 "retry time not reached for any host after a long failure period".
4408 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4409 experimental DomainKeys support:
4411 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4412 (2) On an error such as an illegally used "control", the wrong name for
4413 the control was given.
4415 These problems did NOT occur unless DomainKeys support was compiled.
4417 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4419 PH/24 Added ${if match_ip condition.
4421 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4422 databases so that it will be absolutely obvious if a crash occurs in the
4423 DB library. This is a regular occurrence (often caused by mis-matched
4426 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4427 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4428 picked up on a user's system that detects such things. There doesn't seem
4429 to be a gcc warning option for this - only an attribute that has to be
4430 put on the function's prototype. It seems that in Fedora Core 4 they have
4431 set this on a number of new functions. No doubt there will be more in due
4434 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4435 specify a transport. However, if an address that was verified by such a
4436 router was the subject of a callout, Exim crashed because it tried to
4437 read the rcpt_include_affixes from the non-existent transport. Now it
4438 just assumes that the setting of that option is false. This bug was
4439 introduced by 4.51/PH/31.
4441 PH/28 Changed -d+all to exclude +memory, because that information is very
4442 rarely of interest, but it makes the output a lot bigger. People tend to
4443 do -d+all out of habit.
4445 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4446 code in os-type was giving problems when libc.so lives in lib64, like on
4449 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4450 aren't the modern standard, and it seems that some systems' include files
4451 don't always have them. Exim was already checking for some of the newer
4452 ones like T_AAAA, and defining it itself. I've added checks for all the
4453 record types that Exim uses.
4455 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4456 not automatically generating a new one, as it is supposed to. This
4457 prevented TLS from working. If the file did exist, but contained invalid
4458 data, a new version was generated, as expected. It was only the case of a
4459 non-existent file that was broken.
4461 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4462 with a change in libdomainkeys > 0.64.
4464 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4465 from DNS. If the selector record carries the flag, it now has
4466 precedence over the domain-wide flag.
4468 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4470 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4471 the use of a port name that isn't defined in /etc/services) occurred, the
4472 message was deferred as in a normal delivery, and thus remained on the
4473 spool, instead of being failed because of the mua_wrapper setting. This
4474 is now fixed, and I tidied up some of the mua_wrapper messages at the
4477 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4478 the messages in a hash of arrays rather than using individual hashes.
4479 This is a bit cleaner and results in dramatic memory savings, albeit
4480 at a slight CPU cost.
4482 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4483 as requested by Marc Sherman.
4485 SC/10 Eximstats: added histograms for user specified patterns as requested
4488 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4490 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4491 fopen() in the content-scanning modules that did not already have it.
4497 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4498 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4500 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4502 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4504 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4505 to test Sieve filters that use "vacation".
4507 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4508 that changes the way the GnuTLS parameters are stored in the cache file.
4509 The new format can be generated externally. For backward compatibility,
4510 if the data in the cache doesn't make sense, Exim assumes it has read an
4511 old-format file, and it generates new data and writes a new file. This
4512 means that you can't go back to an older release without removing the
4515 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4516 work if there are any delivery delays because "one_time" forces the
4517 parent to be marked "delivered", so its unseen clone is never tried
4518 again. For this reason, Exim now forbids the simultaneous setting of
4521 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4522 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4523 entry if you want to know the details. The fix, however, neglected to
4524 consider the case when local delivery batching is involved. The test for
4525 "previously delivered" was not happening when checking to see if an
4526 address could be batched with a previous (undelivered) one; under
4527 certain circumstances this could lead to multiple deliveries to the same
4530 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4531 in its include files, and this causes problems building Exim.
4533 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4534 header_syntax) but Exim was just ignoring anything given after a slash.
4535 In particular, this caused confusion with an attempt to use "verify =
4536 reverse_host_lookup/defer_ok". An error is now given when options are
4537 supplied for verify items that do not have them. (Maybe reverse_host_
4538 lookup should have a defer_ok option, but that's a different point.)
4540 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4541 defined by RFC 821) to 2048, because there were problems with some AUTH
4542 commands, and RFC 1869 says the size should be increased for extended
4543 SMTP commands that take arguments.
4545 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4548 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4549 "unknown" error; now it says that the functionality isn't in the binary.
4551 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4552 an address' error message when a string expansion fails (syntax or
4553 whatever). Otherwise the password may appear in the log. Following change
4554 PH/42 below, there is no longer a chance of it appearing in a bounce
4557 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4559 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4560 of its addresses was ever tried. (Bugzilla bug #2.)
4562 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4563 the result incorrectly in the debug output. (It correctly added a newline
4564 to what was transported.)
4566 TF/01 Added $received_time.
4568 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4569 commented out examples of how to interface to a virus scanner and to
4570 SpamAssassin. Also added commented examples of av_scanner and
4571 spamd_address settings.
4573 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4574 and controls are allowed in which ACLs. There were a couple of minor
4575 errors. Some of the entries in the conditions table (which is a table of
4576 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4577 negation of where the condition IS allowed.
4579 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4581 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4582 header file does not have a version number, so I've had to invent a new
4583 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4584 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4585 radiusclient), but it was contributed by a Radius user.
4587 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4588 files or directories, for OpenSSL.
4590 PH/19 When an Exim process that is running as root has to create an Exim log
4591 file, it does so in a subprocess that runs as exim:exim so as to get the
4592 ownership right at creation (otherwise, other Exim processes might see
4593 the file with the wrong ownership). There was no test for failure of this
4594 fork() call, which would lead to the process getting stuck as it waited
4595 for a non-existent subprocess. Forks do occasionally fail when resources
4596 run out. I reviewed all the other calls to fork(); they all seem to check
4599 PH/20 When checking for unexpected SMTP input at connect time (before writing
4600 the banner), Exim was not dealing correctly with a non-positive return
4601 from the read() function. If the client had disconnected by this time,
4602 the result was a log entry for a synchronization error with an empty
4603 string after "input=" when read() returned zero. If read() returned -1
4604 (an event I could not check), uninitialized data bytes were printed.
4605 There were reports of junk text (parts of files, etc) appearing after
4608 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4610 PH/22 Added support for macro redefinition, and (re)definition in between
4611 driver and ACL definitions.
4613 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4614 forgetting to use the resulting value; it was using the unexpanded value.
4616 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4617 hadn't been configured. The fix is from Juergen Kreileder, who
4618 understands it better than I do:
4620 "Here's what I see happening with three configured cyrus_sasl
4621 authenticators configured (plain, login, cram-md5):
4623 On startup auth_cyrus_sasl_init() gets called for each of these.
4624 This means three calls to sasl_listmech() without a specified mech_list.
4625 => SASL tests which mechs of all available mechs actually work
4626 => three warnings about OTP not working
4627 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4629 With the patch, sasl_listmech() also gets called three times. But now
4630 SASL's mech_list option is set to the server_mech specified in the the
4631 authenticator. Or in other words, the answer from sasl_listmech()
4632 gets limited to just the mech you're testing for (which is different
4634 => the return list contains just 'plain' or 'login', 'cram-md5' or
4635 nothing depending on the value of ob->server_mech.
4637 I've just tested the patch: Authentication still works fine,
4638 unavailable mechs specified in the exim configuration are still
4639 caught, and the auth.log warnings about OTP are gone."
4641 PH/25 When debugging is enabled, the contents of the command line are added
4642 to the debugging output, even when log_selector=+arguments is not
4645 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4646 answer is "GNU", and only if the return is "GNU/something" is the answer
4649 PH/27 $acl_verify_message is now set immediately after the failure of a
4650 verification in an ACL, and so is available in subsequent modifiers. In
4651 particular, the message can be preserved by coding like this:
4653 warn !verify = sender
4654 set acl_m0 = $acl_verify_message
4656 Previously, $acl_verify_message was set only while expanding "message"
4657 and "log_message" when a very denied access.
4659 PH/28 Modified OS/os.c-Linux with
4661 -#ifndef OS_LOAD_AVERAGE
4662 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4664 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4665 nomenclature these days.)
4667 PH/29 Installed patch from the Sieve maintainer that adds the options
4668 sieve_useraddress and sieve_subaddress to the redirect router.
4670 PH/30 In these circumstances:
4671 . Two addresses routed to the same list of hosts;
4672 . First host does not offer TLS;
4673 . First host accepts first address;
4674 . First host gives temporary error to second address;
4675 . Second host offers TLS and a TLS session is established;
4676 . Second host accepts second address.
4677 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4678 and peerdn, if requested) that were in fact used only for the second
4681 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4682 attention to any local part prefix or suffix that was matched by the
4683 router that accepted the address. It now behaves in the same way as it
4684 does for delivery: the affixes are removed from the local part unless
4685 rcpt_include_affixes is set on the transport.
4687 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4688 timeout during the DATA phase of an incoming message.
4690 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4691 applied a patch sanctioned by the Sieve maintainer.
4693 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4694 the uid or gid is negative. A case of a negative gid caused this to be
4695 noticed. The fix allows for either to be negative.
4697 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4698 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4701 PH/36 The change PH/12 above was broken. Fixed it.
4703 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4704 the grounds that routing the same address twice would always produce the
4705 same answer. This might have been true once, but it is certainly no
4706 longer true now. Routing a child address may depend on the previous
4707 routing that produced that child. Some complicated redirection strategies
4708 went wrong when messages had multiple recipients, and made Exim's
4709 behaviour dependent on the order in which the addresses were given.
4711 I have moved the duplicate checking until after the routing is complete.
4712 Exim scans the addresses that are assigned to local and remote
4713 transports, and removes any duplicates. This means that more work will be
4714 done, as duplicates will always all be routed, but duplicates are
4715 presumably rare, so I don't expect this is of any significance.
4717 For deliveries to pipes, files, and autoreplies, the duplicate checking
4718 still happens during the routing process, since they are not going to be
4721 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4722 It corrects a timeout issue with spamd. This is Ian's comment: "The
4723 background is that sometimes spamd either never reads data from a
4724 connection it has accepted, or it never writes response data. The exiscan
4725 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4726 blindly assumes that writes won't block so it may never time out."
4728 PH/39 Allow G after quota size as well as K and M.
4730 PH/40 The value set for $authenticated_id in an authenticator may not contain
4731 binary zeroes or newlines because the value is written to log lines and
4732 to spool files. There was no check on this. Now the value is run through
4733 the string_printing() function so that such characters are converted to
4734 printable escape sequences.
4736 PH/41 $message_linecount is a new variable that contains the total number of
4737 lines in the message. Compare $body_linecount, which is the count for the
4740 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4741 bounce and delay warning messages, except in certain special cases, which
4744 (a) An SMTP error message from a remote host;
4745 (b) A message specified in a :fail: redirection;
4746 (c) A message specified in a "fail" command in a system filter;
4747 (d) A message specified in a FAIL return from the queryprogram router;
4748 (e) A message specified by the cannot_route_message router option.
4750 In these cases only, Exim does include the error details in bounce and
4751 warning messages. There are also a few cases where bland messages such
4752 as "unrouteable address" or "local delivery error" are given.
4754 PH/43 $value is now also set for the "else" part of a ${run expansion.
4756 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4757 being worked on, but at least Exim now implements the latest version to
4760 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4761 process to complete was treated as a delivery failure, a timeout while
4762 writing the message to the pipe was logged, but erroneously treated as a
4763 successful delivery. Such timeouts include transport filter timeouts. For
4764 consistency with the overall process timeout, these timeouts are now
4765 treated as errors, giving rise to delivery failures by default. However,
4766 there is now a new Boolean option for the pipe transport called
4767 timeout_defer, which, if set TRUE, converts the failures into defers for
4768 both kinds of timeout. A transport filter timeout is now identified in
4771 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4772 systems where "make" and "gmake" are different, calling "gmake" at top
4773 level broke things. I've arranged for the value of $(MAKE) to be passed
4774 from the Makefile to this script so that it can call the same version of
4778 A note about Exim versions 4.44 and 4.50
4779 ----------------------------------------
4781 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4782 changes of various kinds. As a consequence, a big documentation update was
4783 needed. This delayed the release for rather longer than seemed good, especially
4784 in the light of a couple of (minor) security issues. Therefore, the changes
4785 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4786 release. So 4.44 and 4.50 are in effect two different branches that both start
4789 I have left the 4.50 change log unchanged; it contains all the changes since
4790 4.43. The change log for 4.44 is below; many of its items are identical to
4791 those for 4.50. This seems to be the most sensible way to preserve the
4792 historical information.
4798 1. Minor wording change to the doc/README.SIEVE file.
4800 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4801 computation of the current number of files was incorrect.
4803 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4804 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4807 4. Give more explanation in the error message when the command for a transport
4808 filter fails to execute.
4810 5. There are several places where Exim runs a non-Exim command in a
4811 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4812 was being done only for the command run by the queryprogram router. It is
4813 now done for all such subprocesses. The other cases are: ${run, transport
4814 filters, and the commands run by the lmtp and pipe transports.
4816 6. Added CONFIGURE_GROUP build-time option.
4818 7. Some older OS have a limit of 256 on the maximum number of file
4819 descriptors. Exim was using setrlimit() to set 1000 as a large value
4820 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4821 systems. I've change it so that if it can't get 1000, it tries for 256.
4823 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4824 was an oversight, and furthermore, ever since the addition of extra
4825 controls (e.g. 4.43/32), the checks on when to allow different forms of
4826 "control" were broken. There should now be diagnostics for all cases when a
4827 control that does not make sense is encountered.
4829 9. Added the /retain_sender option to "control=submission".
4831 10. $recipients is now available in the predata ACL (oversight).
4833 11. Tidy the search cache before the fork to do a delivery from a message
4834 received from the command line. Otherwise the child will trigger a lookup
4835 failure and thereby defer the delivery if it tries to use (for example) a
4836 cached ldap connection that the parent has called unbind on.
4838 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4839 of $address_data from the recipient verification was clobbered by the
4840 sender verification.
4842 13. The value of address_data from a sender verification is now available in
4843 $sender_address_data in subsequent conditions in the ACL statement.
4845 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4847 15. Added a new option "connect=<time>" to callout options, to set a different
4850 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4851 was its contents. (It was OK if the option was not defined at all.)
4853 17. A "Completed" log line is now written for messages that are removed from
4854 the spool by the -Mrm option.
4856 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4857 information about exactly what failed.
4859 19. Added -dd to debug only the daemon process.
4861 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4862 handles timeouts, both on the server side and network timeouts. Renamed the
4863 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4865 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4866 It is now set to "smtps".
4868 22. $host_address is now set to the target address during the checking of
4869 ignore_target_hosts.
4871 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4872 being passed; this would have caused $sender_host_name to have been used if
4873 matching the list had actually called for a host name (not very likely,
4874 since this list is usually IP addresses). A host name is now passed as
4877 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4878 code that specifies a non-restarting handler (typically sigaction() in
4879 modern systems) in an attempt to fix a rare and obscure crash bug.
4881 25. Narrowed the window for a race in the daemon that could cause it to ignore
4882 SIGCHLD signals. This is not a major problem, because they are used only to
4883 wake it up if nothing else does.
4885 26. A malformed maildirsize file could cause Exim to calculate negative values
4886 for the mailbox size or file count. Odd effects could occur as a result.
4887 The maildirsize information is now recalculated if the size or filecount
4890 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4891 support for a long time. Removed HAVE_SYS_VFS_H.
4893 28. Installed the latest version of exipick from John Jetmore.
4895 29. In an address list, if the pattern was not a regular expression, an empty
4896 subject address (from a bounce message) matched only if the pattern was an
4897 empty string. Non-empty patterns were not even tested. This was the wrong
4898 because it is perfectly reasonable to use an empty address as part of a
4899 database query. An empty address is now tested by patterns that are
4900 lookups. However, all the other forms of pattern expect the subject to
4901 contain a local part and a domain, and therefore, for them, an empty
4902 address still always fails if the pattern is not itself empty.
4904 30. Exim went into a mad DNS loop when attempting to do a callout where the
4905 host was specified on an smtp transport, and looking it up yielded more
4906 than one IP address.
4908 31. Re-factored the code for checking spool and log partition space into a
4909 function that finds that data and another that does the check. The former
4910 is then used to implement four new variables: $spool_space, $log_space,
4911 $spool_inodes, and $log_inodes.
4913 32. The RFC2047 encoding function was originally intended for short strings
4914 such as real names; it was not keeping to the 75-character limit for
4915 encoded words that the RFC imposes. It now respects the limit, and
4916 generates multiple encoded words if necessary. To be on the safe side, I
4917 have increased the buffer size for the ${rfc2047: expansion operator from
4920 33. It is now permitted to omit both strings after an "if" condition; if the
4921 condition is true, the result is "true". As before, when the second string
4922 is omitted, a false condition yields an empty string. This makes it less
4923 cumbersome to write custom ACL and router conditions.
4925 34. Failure to deliver a bounce message always caused it to be frozen, even if
4926 there was an errors_to setting on the router. The errors_to setting is now
4929 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4930 canonical form (fully expanded) before being placed in
4931 $sender_host_address.
4933 36. The table in the code that translates DNS record types into text (T_A to
4934 "A" for instance) was missing entries for NS and CNAME. It is just possible
4935 that this could have caused confusion if both these types were looked up
4936 for the same domain, because the text type is used as part of Exim's
4937 per-process caching. But the chance of anyone hitting this buglet seems
4940 37. The dnsdb lookup has been extended in a number of ways.
4942 (1) There is a new type, "zns", which walks up the domain tree until it
4943 finds some nameserver records. It should be used with care.
4945 (2) There is a new type, "mxh", which is like "mx" except that it returns
4946 just the host names, not the priorities.
4948 (3) It is now possible to give a list of domains (or IP addresses) to be
4949 looked up. The behaviour when one of the lookups defers can be
4950 controlled by a keyword.
4952 (4) It is now possible to specify the separator character for use when
4953 multiple records are returned.
4955 38. The dnslists ACL condition has been extended: it is now possible to supply
4956 a list of IP addresses and/or domains to be looked up in a particular DNS
4959 39. Added log_selector=+queue_time_overall.
4961 40. When running the queue in the test harness, wait just a tad after forking a
4962 delivery process, to get repeatability of debugging output.
4964 41. Include certificate and key file names in error message when GnuTLS fails
4965 to set them up, because the GnuTLS error message doesn't include the name
4966 of the failing file when there is a problem reading it.
4968 42. Allow both -bf and -bF in the same test run.
4970 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4972 44. The "Exiscan patch" is now merged into the mainline Exim source.
4974 45. Sometimes the final signoff response after QUIT could fail to get
4975 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4976 < 0 before doing a fflush(). This bug looks as though it goes back to the
4977 introduction of TLS in release 3.20, but "sometimes" must have been rare
4978 because the tests only now provoked it.
4980 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4981 (this can affect the format of dates).
4983 47. exim_tidydb, when checking for the continued existence of a message for
4984 which it has found a message-specific retry record, was not finding
4985 messages that were in split spool directories. Consequently, it was
4986 deleting retry records that should have stayed in existence.
4988 48. Steve fixed some bugs in eximstats.
4990 49. The SPA authentication driver was not abandoning authentication and moving
4991 on to the next authenticator when an expansion was forced to fail,
4992 contradicting the general specification for all authenticators. Instead it
4993 was generating a temporary error. It now behaves as specified.
4995 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4996 (the order specifies the preference for clients). The order is now AES256,
4997 AES128, 3DES, ARCFOUR128.
4999 51. Small patch to Sieve code - explicitly set From: when generating an
5002 52. Exim crashed if a remote delivery caused a very long error message to be
5003 recorded - for instance if somebody sent an entire SpamAssassin report back
5004 as a large number of 550 error lines. This bug was coincidentally fixed by
5005 increasing the size of one of Exim's internal buffers (big_buffer) that
5006 happened as part of the Exiscan merge. However, to be on the safe side, I
5007 have made the code more robust (and fixed the comments that describe what
5010 53. Now that there can be additional text after "Completed" in log lines (if
5011 the queue_time_overall log selector is set), a one-byte patch to exigrep
5012 was needed to allow it to recognize "Completed" as not the last thing in
5015 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5016 patch that reportedly fixes this has been added. I am not expert enough to
5017 create a test for it. This is what the patch creator wrote:
5019 "I found a little strange behaviour of ldap code when working with
5020 Windows 2003 AD Domain, where users was placed in more than one
5021 Organization Units. When I tried to give exim partial DN, the exit code
5022 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5023 But simultaneously result of request was absolutely normal ldap result,
5024 so I produce this patch..."
5026 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5027 REFERENCE, so I have modified the code to exclude the patch when that macro
5030 55. Some experimental protocols are using DNS PTR records for new purposes. The
5031 keys for these records are domain names, not reversed IP addresses. The
5032 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5033 leaves it alone. Component reversal etc. now happens only for IP addresses.
5036 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5038 57. Double the size of the debug message buffer (to 2048) so that more of very
5039 long debug lines gets shown.
5041 58. The exicyclog utility now does better if the number of log files to keep
5042 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5044 59. Two changes related to the smtp_active_hostname option:
5046 (1) $smtp_active_hostname is now available as a variable.
5047 (2) The default for smtp_banner uses $smtp_active_hostname instead
5048 of $primary_hostname.
5050 60. The host_aton() function is supposed to be passed a string that is known
5051 to be a valid IP address. However, in the case of IPv6 addresses, it was
5052 not checking this. This is a hostage to fortune. Exim now panics and dies
5053 if the condition is not met. A case was found where this could be provoked
5054 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5055 components; fortuitously, this particular loophole had already been fixed
5056 by change 4.50/55 above.
5058 If there are any other similar loopholes, the new check in host_aton()
5059 itself should stop them being exploited. The report I received stated that
5060 data on the command line could provoke the exploit when Exim was running as
5061 exim, but did not say which command line option was involved. All I could
5062 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5063 running as the user.
5066 61. There was a buffer overflow vulnerability in the SPA authentication code
5067 (which came originally from the Samba project). I have added a test to the
5068 spa_base64_to_bits() function which I hope fixes it.
5071 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5072 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5074 63. The daemon start-up calls getloadavg() while still root for those OS that
5075 need the first call to be done as root, but it missed one case: when
5076 deliver_queue_load_max is set with deliver_drop_privilege. This is
5077 necessary for the benefit of the queue runner, because there is no re-exec
5078 when deliver_drop_privilege is set.
5080 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5081 This has been fixed.
5083 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5084 was in use, was not putting the data itself into the right store pool;
5085 consequently, it could be overwritten for a subsequent message in the same
5086 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5089 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5091 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5092 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5093 the function were treating the return as a boolean value, which happened to
5094 work because 0=false and not-0=true, but is not correct code.
5096 68. The host_aton() function was not handling scoped IPv6 addresses (those
5097 with, for example, "%eth0" on the end) correctly.
5099 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5100 negated items (that is, ~something) in unsigned ints. Some compilers
5101 apparently mutter when there is no cast.
5103 70. If an address verification called from an ACL failed, and did not produce a
5104 user-specific message (i.e. there was only a "system" message), nothing was
5105 put in $acl_verify_message. In this situation, it now puts the system
5108 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5109 session; change 4.31/43 added the unwanted input to the log line - except
5110 that it did not do this in the start of session case. It now does.
5112 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5113 This is wrong and can cause the other end to generate a synchronization
5114 error if it is another Exim or anything else that does the synchronization
5115 check. A QUIT command is no longer sent after a timeout.
5117 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5118 during host lookups.
5120 74. The defer_ok option of callout verification was not working if it was used
5121 when verifying addresses in header lines, that is, for this case:
5123 verify = header_sender/callout=defer_ok
5125 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5126 those file descriptors could be used for SMTP connections. If anything
5127 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5128 could be sent to the SMTP client, causing chaos. The daemon now opens
5129 stdin, stdout, and stderr to /dev/null when it puts itself into the
5132 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5133 log by default. The user can override this with suitable Perl magic.
5135 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5136 the log message when discard triggers, was not working for the DATA ACL or
5137 for the non-SMTP ACL.
5139 78. Error message wording change in sieve.c.
5141 79. If smtp_accept_max_per_host was set, the number of connections could be
5142 restricted to fewer than expected, because the daemon was trying to set up
5143 a new connection before checking whether the processes handling previous
5144 connections had finished. The check for completed processes is now done
5145 earlier. On busy systems, this bug wouldn't be noticed because something
5146 else would have woken the daemon, and it would have reaped the completed
5149 80. If a message was submitted locally by a user whose login name contained one
5150 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5151 It caused a spool format error. I have fixed the spool reading code. A
5152 related problem was that the "from" clause in the Received: line became
5153 illegal because of the space(s). It is now covered by ${quote_local_part.
5155 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5158 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5160 83. Patch from Sieve maintainer to fix unterminated string problem in
5161 "vacation" handling.
5163 84. Some minor changes to the Linux configuration files to help with other
5164 OS variants using glibc.
5166 85. One more patch for Sieve to update vacation handling to latest spec.
5169 ----------------------------------------------------
5170 See the note above about the 4.44 and 4.50 releases.
5171 ----------------------------------------------------
5177 1. Change 4.43/35 introduced a bug that caused file counts to be
5178 incorrectly computed when quota_filecount was set in an appendfile
5181 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5182 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5185 3. Give more explanation in the error message when the command for a transport
5186 filter fails to execute.
5188 4. There are several places where Exim runs a non-Exim command in a
5189 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5190 was being done only for the command run by the queryprogram router. It is
5191 now done for all such subprocesses. The other cases are: ${run, transport
5192 filters, and the commands run by the lmtp and pipe transports.
5194 5. Some older OS have a limit of 256 on the maximum number of file
5195 descriptors. Exim was using setrlimit() to set 1000 as a large value
5196 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5197 systems. I've change it so that if it can't get 1000, it tries for 256.
5199 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5200 was an oversight, and furthermore, ever since the addition of extra
5201 controls (e.g. 4.43/32), the checks on when to allow different forms of
5202 "control" were broken. There should now be diagnostics for all cases when a
5203 control that does not make sense is encountered.
5205 7. $recipients is now available in the predata ACL (oversight).
5207 8. Tidy the search cache before the fork to do a delivery from a message
5208 received from the command line. Otherwise the child will trigger a lookup
5209 failure and thereby defer the delivery if it tries to use (for example) a
5210 cached ldap connection that the parent has called unbind on.
5212 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5213 of $address_data from the recipient verification was clobbered by the
5214 sender verification.
5216 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5217 was its contents. (It was OK if the option was not defined at all.)
5219 11. A "Completed" log line is now written for messages that are removed from
5220 the spool by the -Mrm option.
5222 12. $host_address is now set to the target address during the checking of
5223 ignore_target_hosts.
5225 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5226 being passed; this would have caused $sender_host_name to have been used if
5227 matching the list had actually called for a host name (not very likely,
5228 since this list is usually IP addresses). A host name is now passed as
5231 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5232 code that specifies a non-restarting handler (typically sigaction() in
5233 modern systems) in an attempt to fix a rare and obscure crash bug.
5235 15. Narrowed the window for a race in the daemon that could cause it to ignore
5236 SIGCHLD signals. This is not a major problem, because they are used only to
5237 wake it up if nothing else does.
5239 16. A malformed maildirsize file could cause Exim to calculate negative values
5240 for the mailbox size or file count. Odd effects could occur as a result.
5241 The maildirsize information is now recalculated if the size or filecount
5244 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5245 support for a long time. Removed HAVE_SYS_VFS_H.
5247 18. Updated exipick to current release from John Jetmore.
5249 19. Allow an empty sender to be matched against a lookup in an address list.
5250 Previously the only cases considered were a regular expression, or an
5253 20. Exim went into a mad DNS lookup loop when doing a callout where the
5254 host was specified on the transport, if the DNS lookup yielded more than
5257 21. The RFC2047 encoding function was originally intended for short strings
5258 such as real names; it was not keeping to the 75-character limit for
5259 encoded words that the RFC imposes. It now respects the limit, and
5260 generates multiple encoded words if necessary. To be on the safe side, I
5261 have increased the buffer size for the ${rfc2047: expansion operator from
5264 22. Failure to deliver a bounce message always caused it to be frozen, even if
5265 there was an errors_to setting on the router. The errors_to setting is now
5268 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5269 canonical form (fully expanded) before being placed in
5270 $sender_host_address.
5272 24. Updated eximstats to version 1.33
5274 25. Include certificate and key file names in error message when GnuTLS fails
5275 to set them up, because the GnuTLS error message doesn't include the name
5276 of the failing file when there is a problem reading it.
5278 26. Expand error message when OpenSSL has problems setting up cert/key files.
5281 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5282 (this can affect the format of dates).
5284 28. exim_tidydb, when checking for the continued existence of a message for
5285 which it has found a message-specific retry record, was not finding
5286 messages that were in split spool directories. Consequently, it was
5287 deleting retry records that should have stayed in existence.
5289 29. eximstats updated to version 1.35
5290 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5291 1.35 - bugfix such that pie charts by volume are generated correctly
5293 30. The SPA authentication driver was not abandoning authentication and moving
5294 on to the next authenticator when an expansion was forced to fail,
5295 contradicting the general specification for all authenticators. Instead it
5296 was generating a temporary error. It now behaves as specified.
5298 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5299 (the order specifies the preference for clients). The order is now AES256,
5300 AES128, 3DES, ARCFOUR128.
5302 31. Small patch to Sieve code - explicitly set From: when generating an
5305 32. Exim crashed if a remote delivery caused a very long error message to be
5306 recorded - for instance if somebody sent an entire SpamAssassin report back
5307 as a large number of 550 error lines. This bug was coincidentally fixed by
5308 increasing the size of one of Exim's internal buffers (big_buffer) that
5309 happened as part of the Exiscan merge. However, to be on the safe side, I
5310 have made the code more robust (and fixed the comments that describe what
5313 33. Some experimental protocols are using DNS PTR records for new purposes. The
5314 keys for these records are domain names, not reversed IP addresses. The
5315 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5316 leaves it alone. Component reversal etc. now happens only for IP addresses.
5319 34. The host_aton() function is supposed to be passed a string that is known
5320 to be a valid IP address. However, in the case of IPv6 addresses, it was
5321 not checking this. This is a hostage to fortune. Exim now panics and dies
5322 if the condition is not met. A case was found where this could be provoked
5323 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5324 components; fortuitously, this particular loophole had already been fixed
5325 by change 4.50/55 or 4.44/33 above.
5327 If there are any other similar loopholes, the new check in host_aton()
5328 itself should stop them being exploited. The report I received stated that
5329 data on the command line could provoke the exploit when Exim was running as
5330 exim, but did not say which command line option was involved. All I could
5331 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5332 running as the user.
5335 35. There was a buffer overflow vulnerability in the SPA authentication code
5336 (which came originally from the Samba project). I have added a test to the
5337 spa_base64_to_bits() function which I hope fixes it.
5340 36. The daemon start-up calls getloadavg() while still root for those OS that
5341 need the first call to be done as root, but it missed one case: when
5342 deliver_queue_load_max is set with deliver_drop_privilege. This is
5343 necessary for the benefit of the queue runner, because there is no re-exec
5344 when deliver_drop_privilege is set.
5346 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5347 was in use, was not putting the data itself into the right store pool;
5348 consequently, it could be overwritten for a subsequent message in the same
5349 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5352 38. Sometimes the final signoff response after QUIT could fail to get
5353 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5354 < 0 before doing a fflush(). This bug looks as though it goes back to the
5355 introduction of TLS in release 3.20, but "sometimes" must have been rare
5356 because the tests only now provoked it.
5362 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5363 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5364 Now it returns an integer. A number of calls were still expecting a T/F
5365 return. Fortuitously, in all cases, the tests worked in OK situations,
5366 which is the norm. However, things would have gone wrong on any write
5367 failures on the smtp file descriptor. This function is used when sending
5368 messages over SMTP and also when doing verify callouts.
5370 2. When Exim is called to do synchronous delivery of a locally submitted
5371 message (the -odf or -odi options), it no longer closes stderr before doing
5374 3. Implemented the mua_wrapper option.
5376 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5378 5. Implemented the functions header_remove(), header_testname(),
5379 header_add_at_position(), and receive_remove_recipient(), and exported them
5382 6. If an ACL "warn" statement specified the addition of headers, Exim already
5383 inserted X-ACL-Warn: at the start if there was no header name. However, it
5384 was not making this test for the second and subsequent header lines if
5385 there were newlines in the string. This meant that an invalid header could
5386 be inserted if Exim was badly configured.
5388 7. Allow an ACL "warn" statement to add header lines at the start or after all
5389 the Received: headers, as well as at the end.
5391 8. Added the rcpt_4xx retry error code.
5393 9. Added postmaster_mailfrom=xxx to callout verification option.
5395 10. Added mailfrom=xxxx to the callout verification option, for verify=
5398 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5399 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5401 12. Inserted some casts to stop certain compilers warning when using pointer
5402 differences as field lengths or precisions in printf-type calls (mostly
5403 affecting debugging statements).
5405 13. Added optional readline() support for -be (dynamically loaded).
5407 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5408 same clock tick as a message's arrival, so that its received time was the
5409 same as the "first fail" time on the retry record, and that message
5410 remained on the queue past the ultimate address timeout, every queue runner
5411 would try a delivery (because it was past the ultimate address timeout) but
5412 after another failure, the ultimate address timeout, which should have then
5413 bounced the address, did not kick in. This was a "< instead of <=" error;
5414 in most cases the first failure would have been in the next clock tick
5415 after the received time, and all would be well.
5417 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5418 being recognized when the domain list was tested by the match_domain
5419 condition in an expansion string.
5421 16. Added the ${str2b64: operator.
5423 17. Exim was always calling setrlimit() to set a large limit for the number of
5424 processes, without checking whether the existing limit was already
5425 adequate. (It did check for the limit on file descriptors.) Furthermore,
5426 errors from getrlimit() and setrlimit() were being ignored. Now they are
5427 logged to the main and panic logs, but Exim does carry on, to try to do its
5428 job under whatever limits there are.
5430 18. Imported PCRE 5.0.
5432 19. Trivial typo in log message " temporarily refused connection" (the leading
5435 20. If the log selector return_path_on_delivery was set and an address was
5436 redirected to /dev/null, the delivery process crashed because it assumed
5437 that a return path would always be set for a "successful" delivery. In this
5438 case, the whole delivery is bypassed as an optimization, and therefore no
5441 21. Internal re-arrangement: the function for sending a challenge and reading
5442 a response while authentication was assuming a zero-terminated challenge
5443 string. It's now changed to take a pointer and a length, to allow for
5444 binary data in such strings.
5446 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5448 23. Exim was not respecting finduser_retries when seeking the login of the
5449 uid under which it was called; it was always trying 10 times. (The default
5450 setting of finduser_retries is zero.) Also, it was sleeping after the final
5451 failure, which is pointless.
5453 24. Implemented tls_on_connect_ports.
5455 25. Implemented acl_smtp_predata.
5457 26. If the domain in control=submission is set empty, Exim assumes that the
5458 authenticated id is a complete email address when it generates From: or
5459 Sender: header lines.
5461 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5462 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5463 chown and chgrp in /bin and hostname in /usr/bin.
5465 28. Exim was keeping the "process log" file open after each use, just as it
5466 does for the main log. This opens the possibility of it remaining open for
5467 long periods when the USR1 signal hits a daemon. Occasional processlog
5468 errors were reported, that could have been caused by this. Anyway, it seems
5469 much more sensible not to leave this file open at all, so that is what now
5472 29. The long-running daemon process does not normally write to the log once it
5473 has entered its main loop, and it closes the log before doing so. This is
5474 so that log files can straightforwardly be renamed and moved. However,
5475 there are a couple of unusual error situations where the daemon does write
5476 log entries, and I had neglected to close the log afterwards.
5478 30. The text of an SMTP error response that was received during a remote
5479 delivery was being truncated at 512 bytes. This is too short for some of
5480 the long messages that one sometimes sees. I've increased the limit to
5483 31. It is now possible to make retry rules that apply only when a message has a
5484 specific sender, in particular, an empty sender.
5486 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5487 it possible to be selective about when SMTP synchronization is enforced.
5489 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5491 32. Implemented hosts_connection_nolog.
5493 33. Added an ACL for QUIT.
5495 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5498 35. Added mailbox_size and mailbox_filecount to appendfile.
5500 36. Added control = no_multiline_responses to ACLs.
5502 37. There was a bug in the logic of the code that waits for the clock to tick
5503 in the case where the clock went backwards by a substantial amount such
5504 that the microsecond fraction of "now" was more than the microsecond
5505 fraction of "then" (but the whole seconds number was less).
5507 38. Added support for the libradius Radius client library this is found on
5508 FreeBSD (previously only the radiusclient library was supported).
5514 1. When certain lookups returned multiple values in the form name=value, the
5515 quoting of the values was not always being done properly. Specifically:
5516 (a) If the value started with a double quote, but contained no whitespace,
5518 (b) If the value contained whitespace other than a space character (i.e.
5519 tabs or newlines or carriage returns) it was not quoted.
5520 This fix has been applied to the mysql and pgsql lookups by writing a
5521 separate quoting function and calling it from the lookup code. The fix
5522 should probably also be applied to nisplus, ibase and oracle lookups, but
5523 since I cannot test any of those, I have not disturbed their existing code.
5525 2. A hit in the callout cache for a specific address caused a log line with no
5526 reason for rejecting RCPT. Now it says "Previous (cached) callout
5527 verification failure".
5529 3. There was an off-by-one bug in the queryprogram router. An over-long
5530 return line was truncated at 256 instead of 255 characters, thereby
5531 overflowing its buffer with the terminating zero. As well as fixing this, I
5532 have increased the buffer size to 1024 (and made a note to document this).
5534 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5535 when Exim is waiting for an SMTP response from a remote server, Exim
5536 restarts its select() call on the socket, thereby resetting its timeout.
5537 This is not a problem when such interrupts are rare. Somebody set up a cron
5538 job to run exiwhat every 2 minutes, which is less than the normal select()
5539 timeout (5 or 10 minutes). This meant that the select() timeout never
5540 kicked in because it was always reset. I have fixed this by comparing the
5541 time when an interrupt arrives with the time at the start of the first call
5542 to select(). If more time than the timeout has elapsed, the interrupt is
5543 treated as a timeout.
5545 5. Some internal re-factoring in preparation for the addition of Sieve
5546 extensions (by MH). In particular, the "personal" test is moved to a
5547 separate function, and given an option for scanning Cc: and Bcc: (which is
5548 not set for Exim filters).
5550 6. When Exim created an email address using the login of the caller as the
5551 local part (e.g. when creating a From: or Sender: header line), it was not
5552 quoting the local part when it contained special characters such as @.
5554 7. Installed new OpenBSD configuration files.
5556 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5557 try to make them clearer.
5559 9. Callout options, other than the timeout value, were being ignored when
5560 verifying sender addresses in header lines. For example, when using
5562 verify = header_sender/callout=no_cache
5564 the cache was (incorrectly) being used.
5566 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5567 only the Cygwin environment.
5569 11. When return_path_on_delivery was set as a log selector, if different remote
5570 addresses in the same message used different return paths and parallel
5571 remote delivery occurred, the wrong values would sometimes be logged.
5572 (Whenever a remote delivery process finished, the return path value from
5573 the most recently started remote delivery process was logged.)
5575 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5576 lines when AUTH and/or TLS are in use. This is the "received protocol"
5577 field. Exim used to use "asmtp" for authenticated SMTP, without any
5578 indication (in the protocol name) for TLS use. Now it follows the RFC and
5579 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5580 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5581 names appear in log lines as well as in Received: header lines.
5583 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5584 extensions, and comparison tests, and to fix some bugs.
5586 14. Changes to the "personal" filter test:
5588 (1) The test was buggy in that it was just doing the equivalent of
5589 "contains" tests on header lines. For example, if a user's address was
5590 anne@some.where, the "personal" test would incorrectly be true for
5592 To: susanne@some.where
5594 This test is now done by extracting each address from the header in turn,
5595 and checking the entire address. Other tests that are part of "personal"
5596 are now done using regular expressions (for example, to check local parts
5597 of addresses in From: header lines).
5599 (2) The list of non-personal local parts in From: addresses has been
5600 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5601 taken from the Sieve specification recommendations.
5603 (3) If the message contains any header line starting with "List-" it is
5604 treated as non-personal.
5606 (4) The test for "circular" in the Subject: header line has been removed
5607 because it now seems ill-conceived.
5609 15. Minor typos in src/EDITME comments corrected.
5611 16. Installed latest exipick from John Jetmore.
5613 17. If headers_add on a router specified a text string that was too long for
5614 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5615 of string_sprintf() is now avoided.
5617 18. $message_body_size was not set (it was always zero) when running the DATA
5618 ACL and the local_scan() function.
5620 19. For the "mail" command in an Exim filter, no default was being set for
5621 the once_repeat time, causing a random time value to be used if "once" was
5622 specified. (If the value happened to be <= 0, no repeat happened.) The
5623 default is now 0s, meaning "never repeat". The "vacation" command was OK
5624 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5625 (I found it when inspecting the code).
5627 20. There is now an overall timeout for performing a callout verification. It
5628 defaults to 4 times the callout timeout, which applies to individual SMTP
5629 commands during the callout. The overall timeout applies when there is more
5630 than one host that can be tried. The timeout is checked before trying the
5631 next host. This prevents very long delays if there are a large number of
5632 hosts and all are timing out (e.g. when the network connections are timing
5633 out). The value of the overall timeout can be changed by specifying an
5634 additional sub-option for "callout", called "maxwait". For example:
5636 verify = sender/callout=5s,maxwait=20s
5638 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5639 to the end before writing, but this should make it even safer).
5641 22. Exim was forgetting that it had advertised PIPELINING for the second and
5642 subsequent messages on an SMTP connection. It was also not resetting its
5643 memory on STARTTLS and an internal HELO.
5645 23. When Exim logs an SMTP synchronization error within a session, it now
5646 records whether PIPELINING has been advertised or not.
5648 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5649 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5650 rather than long int.
5652 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5654 26. Added the never_mail option to autoreply.
5660 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5661 crash if the getsockname() call failed; this can happen if a connection is
5662 closed very soon after it is established. The problem was simply in the
5663 order in which certain operations were done, causing Exim to try to write
5664 to the SMTP stream before it had set up the file descriptor. The bug has
5665 been fixed by making things happen in the correct order.
5671 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5672 before the connection was closed, thus losing the rejection response.
5674 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5675 some early Solaris releases, but causes trouble in current releases where
5676 socklen_t is defined.
5678 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5681 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5684 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5685 that releasing the top of it at the end releases what was used for sub-
5686 expansions (unless the block got too big). However, discard this block if
5687 the first thing is a variable or header, so that we can use its block when
5688 it is dynamic (useful for very large $message_headers, for example).
5690 6. Lookups now cache *every* query, not just the most recent. A new, separate
5691 store pool is used for this. It can be recovered when all lookup caches are
5692 flushed. Lookups now release memory at the end of their result strings.
5693 This has involved some general refactoring of the lookup sources.
5695 7. Some code has been added to the store_xxx() functions to reduce the amount
5696 of flapping under certain conditions.
5698 8. log_incoming_interface used to affect only the <= reception log lines. Now
5699 it causes the local interface and port to be added to several more SMTP log
5700 lines, for example "SMTP connection from", and rejection lines.
5702 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5704 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5706 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5707 could be overwritten at the end of the current message (or the start of a
5708 new message if it was set in a HELO ACL). The value is now preserved for
5709 the duration of the SMTP connection.
5711 12. If a transport had a headers_rewrite setting, and a matching header line
5712 contained an unqualified address, that address was qualified, even if it
5713 did not match any rewriting rules. The underlying bug was that the values
5714 of the flags that permit the existence of unqualified sender and recipient
5715 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5716 non-local messages, and by -bnq for local messages) were not being
5717 preserved with the message after it was received.
5719 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5720 "next input=" as part of the text comprising the host identity instead of
5721 the correct text. The code was using the same buffer for two different
5722 strings. However, depending on which order the printing function evaluated
5723 its arguments, the bug did not always show up. Under Linux, for example, my
5724 test suite worked just fine.
5726 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5727 doesn't work with some older versions of Perl. It has been changed to "my",
5728 which in any case is probably the better facility to use.
5730 15. A really picky compiler found some instances of statements for creating
5731 error messages that either had too many or two few arguments for the format
5734 16. The size of the buffer for calls to the DNS resolver has been increased
5735 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5736 for addresses that have a lot of PTR records. This alleviates a problem; it
5737 does not fully solve it.
5739 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5740 buffer now truncates the list and logs the incident, which is the same
5741 action as happens when Exim is looking up a host name and its aliases.
5742 Previously in this situation something unpredictable would happen;
5743 sometimes it was "internal error: store_reset failed".
5745 18. If a server dropped the connection unexpectedly when an Exim client was
5746 using GnuTLS and trying to read a response, the client delivery process
5747 crashed while trying to generate an error log message.
5749 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5750 string, for example:
5752 warn message = H1: something\nH2: something
5754 the text was added as a single header line from Exim's point of view
5755 though it ended up OK in the delivered message. However, searching for the
5756 second and subsequent header lines using $h_h2: did not work. This has been
5757 fixed. Similarly, if a system filter added multiple headers in this way,
5758 the routers could not see them.
5760 20. Expanded the error message when iplsearch is called with an invalid key to
5761 suggest using net-iplsearch in a host list.
5763 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5764 ACLs are no longer actually imposed (and a message to that effect is
5767 22. If a "gecos" field in a passwd entry contained escaped characters, in
5768 particular, if it contained a \" sequence, Exim got it wrong when building
5769 a From: or a Sender: header from that name. A second bug also caused
5770 incorrect handling when an unquoted " was present following a character
5771 that needed quoting.
5773 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5774 was not being matched caselessly.
5776 24. Arranged for all hyphens in the exim.8 source to be escaped with
5779 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5780 properly. Recipient callout cache records were still being keyed to include
5781 the sender, even when use_sender was set false. This led to far more
5782 callouts that were necessary. The sender is no longer included in the key
5783 when use_sender is false.
5785 26. Added "control = submission" modifier to ACLs.
5787 27. Added the ${base62d: operator to decode base 62 numbers.
5789 28. dnsdb lookups can now access SRV records.
5791 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5792 the configuration file.
5794 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5795 (-v) mode. This makes the output for a verbose queue run more intelligible.
5797 31. Added a use_postmaster feature to recipient callouts.
5799 32. Added the $body_zerocount variable, containing the number of binary zero
5800 bytes in the message body.
5802 33. The time of last modification of the "new" subdirectory is now used as the
5803 "mailbox time last read" when there is a quota error for a maildir
5806 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5808 35. Added +ignore_unknown as a special item in host lists.
5810 36. Code for decoding IPv6 addresses in host lists is now included, even if
5811 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5812 address was recognized as an IP address, but was then not correctly decoded
5813 into binary, causing unexpected and incorrect effects when compared with
5820 1. Very minor rewording of debugging text in manualroute to say "list of
5821 hosts" instead of "hostlist".
5823 2. If verify=header_syntax was set, and a header line with an unqualified
5824 address (no domain) and a large number of spaces between the end of the
5825 name and the colon was received, the reception process suffered a buffer
5826 overflow, and (when I tested it) crashed. This was caused by some obsolete
5827 code that should have been removed. The fix is to remove it!
5829 3. When running in the test harness, delay a bit after writing a bounce
5830 message to get a bit more predictability in the log output.
5832 4. Added a call to search_tidyup() just before forking a reception process. In
5833 theory, someone could use a lookup in the expansion of smtp_accept_max_
5834 per_host which, without the tidyup, could leave open a database connection.
5836 5. Added the variables $recipient_data and $sender_data which get set from a
5837 lookup success in an ACL "recipients" or "senders" condition, or a router
5838 "senders" option, similar to $domain_data and $local_part_data.
5840 6. Moved the writing of debug_print from before to after the "senders" test
5843 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5844 problems for message scanning, either using a data ACL, or using
5845 local_scan() because the Received: header was not generated till after they
5846 were called (in order to set the time as the time of reception completion).
5847 I have revised the way this works. The header is now generated after the
5848 body is received, but before the ACL or local_scan() are called. After they
5849 are run, the timestamp in the header is updated.
5855 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5856 before starting a queue runner without re-exec. This happened only when
5857 deliver_drop_privilege was set or when the Exim user was set to root. The
5858 effect of the bug was that timeouts during subsequent deliveries caused
5859 crashes instead of being properly handled. The handler is now left at its
5860 default (and expected) setting.
5862 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5863 message, again when deliver_drop_privilege is set or Exim is run as root.
5864 The bug described in (1) was not present in this case, but the tidying up
5865 of the other signals was missing. I have made the two cases consistent.
5867 3. The ignore_target_hosts setting on a manualroute router was being ignored
5868 for hosts that were looked up using the /MX notation.
5870 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5873 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5874 operated on the sender address. After changing the $sender_address to <>
5875 for the sender address verify, Exim was re-instated it as the original
5876 (before rewriting) address, but remembering that it had rewritten it, so it
5877 wasn't rewriting it again. This bug also had the effect of breaking the
5878 sender address verification caching when the sender address was rewritten.
5880 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5881 This has been changed so that if the ip literal address matches
5882 ignore_target_hosts, the router declines.
5884 7. Added expansion conditions match_domain, match_address, and match_local_
5885 part (NOT match_host).
5887 8. The placeholder for the Received: header didn't have a length field set.
5889 9. Added code to Exim itself and to exim_lock to test for a specific race
5890 condition that could lead to file corruption when using MBX delivery. The
5891 issue is with the lockfile that is created in /tmp. If this file is removed
5892 after a process has opened it but before that process has acquired a lock,
5893 there is the potential for a second process to recreate the file and also
5894 acquire a lock. This could lead to two Exim processes writing to the file
5895 at the same time. The added code performs the same test as UW imapd; it
5896 checks after acquiring the lock that its file descriptor still refers to
5897 the same named file.
5899 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5900 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5903 11. Added the smtp_active_hostname option. If used, this will typically be made
5904 to depend on the incoming interface address. Because $interface_address is
5905 not set up until the daemon has forked a reception process, error responses
5906 that can happen earlier (such as "too many connections") no longer contain
5909 12. If an expansion in a condition on a "warn" statement fails because a lookup
5910 defers, the "warn" statement is abandoned, and the next ACL statement is
5911 processed. Previously this caused the whole ACL to be aborted.
5913 13. Added the iplsearch lookup type.
5915 14. Added ident_timeout as a log selector.
5917 15. Added tls_certificate_verified as a log selector.
5919 16. Added a global option tls_require_ciphers (compare the smtp transport
5920 option of the same name). This controls incoming TLS connections.
5922 17. I finally figured out how to make tls_require_ciphers do a similar thing
5923 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5924 before starting the TLS session.
5926 18. Tabs are now shown as \t in -bP output.
5928 19. If the log selector return_path_on_delivery was set, Exim crashed when
5929 bouncing a message because it had too many Received: header lines.
5931 20. If two routers both had headers_remove settings, and the first one included
5932 a superfluous trailing colon, the final name in the first list and the
5933 first name in the second list were incorrectly joined into one item (with a
5934 colon in the middle).
5940 1. Added -C and -D options to the exinext utility, mainly to make it easier
5941 to include in the automated testing, but these could be helpful when
5942 multiple configurations are in use.
5944 2. The exinext utility was not formatting the output nicely when there was
5945 an alternate port involved in the retry record key, nor when there was a
5946 message id as well (for retries that were specific to a specific message
5947 and a specific host). It was also confused by IPv6 addresses, because of
5948 the additional colons they contain. I have fixed the IPv4 problem, and
5949 patched it up to do a reasonable job for IPv6.
5951 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5952 delivery, the log line now contains "pipelined" if PIPELINING was used.
5954 4. An SMTP transport process used to panic and die if the bind() call to set
5955 an explicit outgoing interface failed. This has been changed; it is now
5956 treated in the same way as a connect() failure.
5958 5. A reference to $sender_host_name in the part of a conditional expansion
5959 that was being skipped was still causing a DNS lookup. This no longer
5962 6. The def: expansion condition was not recognizing references to header lines
5963 that used bh_ and bheader_.
5965 7. Added the _cache feature to named lists.
5967 8. The code for checking quota_filecount in the appendfile transport was
5968 allowing one more file than it should have been.
5970 9. For compatibility with Sendmail, the command line option
5978 and sets the incoming protocol and host name (for trusted callers). The
5979 host name and its colon can be omitted when only the protocol is to be set.
5980 Note the Exim already has two private options, -pd and -ps, that refer to
5981 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5982 "s", but I don't think that's a major issue.
5984 10. A number of refactoring changes to the code, none of which should affect
5987 (a) The number of logging options was getting close to filling up the
5988 32-bit word that was used as a bit map. I have split them into two classes:
5989 those that are passed in the argument to log_write(), and those that are
5990 only ever tested independently outside of that function. These are now in
5991 separate 32-bit words, so there is plenty of room for expansion again.
5992 There is no change in the user interface or the logging behaviour.
5994 (b) When building, for example, log lines, the code previously used a
5995 macro that called string_cat() twice, in order to add two strings. This is
5996 not really sufficiently general. Furthermore, there was one instance where
5997 it was actually wrong because one of the argument was used twice, and in
5998 one call a function was used. (As it happened, calling the function twice
5999 did not affect the overall behaviour.) The macro has been replaced by a
6000 function that can join an arbitrary number of extra strings onto a growing
6003 (c) The code for expansion conditions now uses a table and a binary chop
6004 instead of a serial search (which was left over from when there were very
6005 few conditions). Also, it now recognizes conditions like "pam" even when
6006 the relevant support is not compiled in: a suitably worded error message is
6007 given if an attempt is made to use such a condition.
6009 11. Added ${time_interval:xxxxx}.
6011 12. A bug was causing one of the ddress fields not to be passed back correctly
6012 from remote delivery subprocesses. The field in question was not being
6013 subsequently used, so this caused to problems in practice.
6015 13. Added new log selectors queue_time and deliver_time.
6017 14. Might have fixed a bug in maildirsizefile handling that threw up
6018 "unexpected character" debug warnings, and recalculated the data
6019 unnecessarily. In any case, I expanded the warning message to give more
6022 15. Added the message "Restricted characters in address" to the statements in
6023 the default ACL that block characters like @ and % in local parts.
6025 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6026 Three changes have been made:
6028 (a) There was a serious bug; a negative response to MAIL caused the whole
6029 recipient domain to be cached as invalid, thereby blocking all messages
6030 to all local parts at the same domain, from all senders. This bug has
6031 been fixed. The domain is no longer cached after a negative response to
6032 MAIL if the sender used is not empty.
6034 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6037 (c) A new callout option, "use_sender" has been added for people who want
6038 the modified behaviour.
6044 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6047 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6048 indeed breaks things for older releases.
6050 3. Added additional logging to the case where there is a problem reading data
6051 from a filter that is running in a subprocess using a pipe, in order to
6052 try to track down a specific problem.
6054 4. Testing facility fudge: when running in the test harness and attempting
6055 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6056 getting "No route to host". Convert this to a timeout.
6058 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6061 6. Some OS don't have socklen_t but use size_t instead. This affects the
6062 fifth argument of getsockopt() amongst other things. This is now
6063 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6064 can be set for individual OS. I have set it for SunOS5, OSF1, and
6065 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6066 some earlier ones do not.
6068 7. Change 4.30/15 was not doing the test caselessly.
6070 8. The standard form for an IPv6 address literal was being rejected by address
6071 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6072 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6073 this, as well as the form without the "IPv6" on the front (but only when
6074 address literals are enabled, of course).
6076 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6078 10. Exim crashed if a message with an empty sender address specified by -f
6079 encountered a router with an errors_to setting. This could be provoked only
6080 by a command such as
6084 where an empty string was supplied; "<>" did not hit this bug.
6086 11. Installed PCRE release 4.5.
6088 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6089 remained set. It is now erased.
6091 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6092 times from message ids (which are base 36 rather than the normal 62).
6094 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6095 were being counted as actual protocol errors, and logged if the log
6096 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6097 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6098 and DATA following a set of rejected RCPTs do not count as protocol errors.
6099 In other words, Exim assumes they were pipelined, though this may not
6100 actually be the case. Of course, in all cases the client gets an
6101 appropriate error code.
6103 15. If a lookup fails in an ACL condition, a message about the failure may
6104 be available; it is used if testing the ACL cannot continue, because most
6105 such messages specify what the cause of the deferral is. However, some
6106 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6107 that caused an old message to be retained and used if a later statement
6108 caused a defer, replacing the real cause of the deferral.
6110 16. If an IP address had so many PTR records that the DNS lookup buffer
6111 was not large enough to hold them, Exim could crash while trying to process
6112 the truncated data. It now detects and logs this case.
6114 17. Further to 4.21/58, another change has been made: if (and only if) the
6115 first line of a message (the first header line) ends with CRLF, a bare LF
6116 in a subsequent header line has a space inserted after it, so as not to
6117 terminate the header.
6119 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6120 unnecessarily, used atoi() instead of strtol(), and didn't check the
6121 termination when getting file sizes from file names by regex.
6123 19. Completely re-implemented the support for maildirsize files, in the light
6124 of a number of problems with the previous contributed implementation
6125 (4.30/29). In particular:
6127 . If the quota is zero, the maildirsize file is maintained, but no quota is
6130 . If the maildir directory does not exist, it is created before any attempt
6131 to write a maildirsize file.
6133 . The quota value in the file is just a cache; if the quota is changed in
6134 the transport, the new value overrides.
6136 . A regular expression is available for excluding directories from the
6139 20. The autoreply transport checks the characters in options that define the
6140 message's headers; it allows continued headers, but it was checking with
6141 isspace() after an embedded newline instead of explicitly looking for a
6144 21. If all the "regular" hosts to which an address was routed had passed their
6145 expiry times, and had not reached their retry times, the address was
6146 bounced, even if fallback hosts were defined. Now Exim should go on to try
6149 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6150 equivalent code in the SMTP transport. Some hosts send humungous responses
6151 to HELO/EHLO, more than 1024 it seems.
6153 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6154 gives compiler warnings in some environments. I've now done it "properly",
6157 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6158 (because of problems with the built-in one) was declared to return uschar *
6159 instead of char *, causing compiler failure.
6161 25. Fixed a file descriptor leak when processing alias/forward files.
6163 26. Fixed a minor format string issue in dbfn.c.
6165 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6167 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6168 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6171 29. When a sender address is verified, it is cached, to save repeating the test
6172 when there is more than one recipient in a message. However, when the
6173 verification involves a callout, it is possible for different callout
6174 options to be set for different recipients. It is too complicated to keep
6175 track of this in the cache, so now Exim always runs a verification when a
6176 callout is required, relying on the callout cache for the optimization.
6177 The overhead is duplication of the address routing, but this should not be
6180 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6181 to be verified with callout=postmaster, and the main callout worked but the
6182 postmaster check failed, the verification correctly failed. However, if a
6183 subsequent RCPT command asked for sender verification *without* the
6184 postmaster check, incorrect caching caused this verification also to fail,
6187 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6188 it was not caching the DNS options (qualify_single, search_parents) that
6189 were used when the lookup failed. A subsequent lookup with different
6190 options therefore always gave the same answer, though there were cases
6191 where it should not have. (Example: a "domains = !$mx_any" option on a
6192 dnslookup router: the "domains" option is always processed without any
6193 widening, but the router might have qualify_single set.) Now Exim uses the
6194 cached value only when the same options are set.
6196 32. Added John Jetmore's "exipick" utility to the distribution.
6198 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6199 than a timeout (e.g. a certificate is required, and is not provided), an
6200 Exim server now closes the connection immediately. Previously it waited for
6201 the client to close - but if the client is SSL, it seems that they each
6202 wait for each other, leading to a delay before one of them times out.
6204 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6205 maintained 0.8.x compatibility because I don't think many are using it, and
6206 it is clearly obsolete.
6208 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6211 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6212 client certificate was expired. A simple patch fixes this, though I don't
6213 understand the full logic of why the verify callback is called multiple
6216 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6217 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6218 which causes problems with some clients (such as the Certicom SSL Plus
6219 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6220 disables the coutermeasure allowing Eudora to connect."
6222 38. Exim was not checking that a write() to a log file succeeded. This could
6223 lead to Bad Things if a log got too big, in particular if it hit a file
6224 size limit. Exim now panics and dies if it cannot write to a log file, just
6225 as it does if it cannot open a log file.
6227 39. Modified OS/Makefile-Linux so that it now contains
6229 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6231 The two -D definitions ensure that Exim is compiled with large file
6232 support, which makes it possible to handle log files that are bigger than
6235 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6236 instance) a domain was checked against a named list that involved a lookup,
6237 causing $domain_data to be set, then another domain was checked against the
6238 same list, then the first domain was re-checked, the value of $domain_data
6239 after the final check could be wrong. In particular, if the second check
6240 failed, it could be set empty. This bug probably also applied to
6243 41. The strip_trailing_dot option was not being applied to the address given
6244 with the -f command-line option.
6246 42. The code for reading a message's header from the spool was incrementing
6247 $received_count, but never initializing it. This meant that the value was
6248 incorrect (doubled) while delivering a message in the same process in which
6249 it was received. In the most common configuration of Exim, this never
6250 happens - a fresh exec is done - but it can happen when
6251 deliver_drop_privilege is set.
6253 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6254 it now includes up to 150 characters of the unexpected data in the log
6257 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6258 and building data strings. The size of both of these buffers was 10 000
6259 bytes - far larger than anybody would *ever* want, thought I. Needless to
6260 say, somebody hit the limit. I have increased the maximum line length to
6261 20 000 and the maximum data length of concatenated lines to 100 000. I have
6262 also fixed two bugs, because there was no checking on these buffers. Tsk,
6263 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6264 buffer is too small.
6266 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6267 lsearch lookups. Now it does.
6269 46. When parsing a route_list item in a manualroute router, a fixed-length
6270 buffer was used for the list of hosts. I made this 1024 bytes long,
6271 thinking that nobody would ever have a list of hosts that long. Wrong.
6272 Somebody had a whole pile of complicated expansion conditions, and the
6273 string was silently truncated, leading to an expansion error. It turns out
6274 that it is easier to change to an unlimited length (owing to other changes
6275 that have happened since this code was originally written) than to build
6276 structure for giving a limitation error. The length of the item that
6277 expands into the list of hosts is now unlimited.
6279 47. The lsearch lookup could not handle data where the length of text line was
6280 more than 4095 characters. Such lines were truncated, leading to shortened
6281 data being returned. It should now handle lines of any length.
6283 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6284 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6287 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6288 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6289 no longer happens in comment lines. A list of replacements is now placed
6290 at the head of all of the source files, except those whose only change is
6291 to replace PERL_COMMAND in the very first #! line.
6293 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6294 messages on the queue, with a bottom-up merge sort, using code contributed
6295 by Michael Haardt. This should make operations like -bp somewhat faster on
6296 large queues. It won't affect queue runners, except when queue_run_in_order
6299 51. Installed eximstats 1.31 in the distribution.
6301 52. Added support for SRV lookups to the dnslookup router.
6303 53. If an ACL referred to $message_body or $message_body_end, the value was not
6304 reset for any messages that followed in the same SMTP session.
6306 54. The store-handling optimization for building very long strings was not
6307 differentiating between the different store pools. I don't think this
6308 actually made any difference in practice, but I've tidied it.
6310 55. While running the routers to verify a sender address, $sender_address
6311 was still set to the sender address. This is wrong, because when routing to
6312 send a bounce to the sender, it would be empty. Therefore, I have changed
6313 it so that, while verifying a sender address, $sender_address is set to <>.
6314 (There is no change to what happens when verifying a recipient address.)
6316 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6317 target A or AAAA records (if not already returned) without resetting the
6318 qualify_single or search_parents options of the DNS resolver. These are
6319 inappropriate in this case because the targets of MX and SRV records must
6320 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6321 target that, when qualified, matched something in the local domain. These
6322 two options are now turned off when doing these lookups.
6324 57. It seems that at least some releases of Reiserfs (which does not have the
6325 concept of a fixed number of inodes) returns zero and not -1 for the
6326 number of available inodes. This interacted badly with check_spool_inodes,
6327 which assumed that -1 was the "no such thing" setting. What I have done is
6328 to check that the total number of inodes is greater than zero before doing
6329 the test of how many are available.
6331 58. When a "warn" ACL statement has a log_message modifier, the message is
6332 remembered, and not repeated. This is to avoid a lot of repetition when a
6333 message has many recipients that cause the same warning to be written.
6334 However, Exim was preserving the list of already written lines for an
6335 entire SMTP session, which doesn't seem right. The memory is now reset if a
6336 new message is started.
6338 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6339 debugging output unless log_rewrite was also set.
6341 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6342 of (int)(handle) when we know that handle contains (void *)(-1).
6344 61. The Exim daemon panic-logs an error return when it closes the incoming
6345 connection. However "connection reset by peer" seems to be common, and
6346 isn't really an error worthy of noting specially, so that particular error
6349 62. When Exim is trying to find all the local interfaces, it used to panic and
6350 die if the ioctl to get the interface flags failed. However, it seems that
6351 on at least one OS (Solaris 9) it is possible to have an interface that is
6352 included in the list of interfaces, but for which you get a failure error
6353 for this call. This happens when the interface is not "plumbed" into a
6354 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6355 failure of the "get flags" call assumes that the interface is down.
6357 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6358 makes life easier for people who are doing arithmetic on fields extracted
6359 from dates, where you often get leading zeros that should not be
6360 interpreted as octal.
6362 64. Added qualify_domain to the redirect router, to override the global
6365 65. If a pathologically long header line contained very many addresses (the
6366 report of this problem mentioned 10 000) and each of them was rewritten,
6367 Exim could use up a very large amount of memory. (It kept on making new
6368 copies of the header line as it rewrote, and never released the old ones.)
6369 At the expense of a bit more processing, the header rewriting function has
6370 been changed so that it no longer eats memory in this way.
6372 66. The generation of the Received: header has been moved from the time that a
6373 message starts to be received, to the time that it finishes. The timestamp
6374 in the Received: header should now be very close to that of the <= log
6375 line. There are two side-effects of this change:
6377 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6378 logged header lines no longer include the local Received: line, because
6379 it has not yet been created. The same applies to a copy of the message
6380 that is returned to a non-SMTP sender when a message is rejected.
6382 (b) When a filter file is tested using -bf, no additional Received: header
6383 is added to the test message. After some thought, I decided that this
6386 This change does not affect the value of $received_for. It is still set
6387 after address rewriting, but before local_scan() is called.
6389 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6391 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6392 gave an unhelpful panic error message, and a defer error. I have managed to
6393 change this behaviour so that it now rejects any supplied certificate,
6394 which seems right, as the list of acceptable certificates is empty.
6396 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6397 gave an unhelpful defer error. I have not managed to make this reject any
6398 supplied certificates, but the error message it gives is "no certificate
6399 supplied", which is not helpful.
6401 70. exigrep's output now also includes lines that are not associated with any
6402 message, but which match the given pattern. Implemented by a patch from
6403 Martin Sluka, which also tidied up the Perl a bit.
6405 71. Recipient callout verification, like sender verification, was using <> in
6406 the MAIL FROM command. This isn't really the right thing, since the actual
6407 sender may affect whether the remote host accepts the recipient or not. I
6408 have changed it to use the actual sender in the callout; this means that
6409 the cache record is now keyed on a recipient/sender pair, not just the
6410 recipient address. There doesn't seem to be a real danger of callout loops,
6411 since a callout by the remote host to check the sender would use <>.
6412 [SEE ABOVE: changed after hitting problems.]
6414 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6415 temporary errors. However, in the case of such a code being given after
6416 the end of a data transmission (i.e. after ".") Exim was failing to write
6417 a retry record for the message. (Yes, there was some broken host that was
6418 actually sending 8xx at this point.)
6420 73. An unknown lookup type in a host list could cause Exim to panic-die when
6421 the list was checked. (An example that provoked this was putting <; in the
6422 middle of a list instead of at the start.) If this happened during a DATA
6423 ACL check, a -D file could be left lying around. This kind of configuration
6424 error no longer causes Exim to die; instead it causes a defer error. The
6425 incident is still logged to the main and panic logs.
6427 74. Buglet left over from Exim 3 conversion. The message "too many messages
6428 in one connection" was written to the rejectlog but not the mainlog, except
6429 when address rewriting (yes!) was being logged.
6431 75. Added write_rejectlog option.
6433 76. When a system filter was run not as root (that is, when system_filter_user
6434 was set), the values of the $n variables were not being returned to the
6435 main process; thus, they were not subsequently available in the $sn
6438 77. Added +return_path_on_delivery log selector.
6440 78. A connection timeout was being treated differently from recipients deferred
6441 when testing hosts_max_try with a message that was older than the host's
6442 retry timeout. (The host should not be counted, thus allowing all hosts to
6443 be tried at least once before bouncing.) This may have been the cause of an
6444 occasionally reported bug whereby a message would remain on the queue
6445 longer than the retry timeout, but would be bounced if a delivery was
6446 forced. I say "may" because I never totally pinned down the problem;
6447 setting up timeout/retry tests is difficult. See also the next item.
6449 79. The ultimate address timeout was not being applied to errors that involved
6450 a combination of host plus message (for example, a timeout on a MAIL
6451 command). When an address resolved to a number of possible hosts, and they
6452 were not all tried for each delivery (e.g. because of hosts_max_try), a
6453 message could remain on the queue longer than the retry timeout.
6455 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6458 81. Fixed an obscure SMTP outgoing bug which required at least the following
6459 conditions: (a) there was another message waiting for the same server;
6460 (b) the server returned 5xx to all RCPT commands in the first message so
6461 that the message was not completed; (c) the server dropped the connection
6462 or gave a negative response to the RSET that Exim sends to abort the
6463 transaction. The observed case was a dropped connection after DATA that had
6464 been sent in pipelining mode. That is, the server had advertised PIPELINING
6465 but was not implementing it correctly. The effect of the bug was incorrect
6466 behaviour, such as trying another host, and this could lead to a crash.
6472 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6473 and daemon.c were passed as pointers to ints; they should have been
6474 pointers to socklen_t variables (which are typically unsigned ints).
6476 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6479 3. Fixed a really odd bug that affected only the testing scheme; patching a
6480 certain fixed string in the binary changed the value of another string that
6481 happened to be identical to the end of the original first string.
6483 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6484 name", it returns that address as the IP address. On some operating
6485 systems (e.g. Solaris), it also passes back the IP address string as the
6486 "host name". However, on others (e.g. Linux), it passes back an empty
6487 string. Exim wasn't checking for this, and was changing the host name to an
6488 empty string, assuming it had been canonicalized.
6490 5. Although rare, it is permitted to have more than one PTR record for a given
6491 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6492 all the names associated with an address, because they do in Solaris.
6493 However, it seems that they do not in Linux for data that comes from the
6494 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6495 I found this out when I moved to a new Linux workstation and tried to run
6496 the Exim test suite.
6498 To get round this problem I have changed the code so that it now does its
6499 own call to the DNS to look up PTR records when searching for a host name.
6500 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6501 addresses that are only in /etc/hosts are still found.
6503 This behaviour is, however, controlled by an option called host_lookup_
6504 order, which defaults to "bydns:byaddr". If people want to use the other
6505 order, or indeed, just use one or the other means of lookup, they can
6506 specify it in this variable.
6508 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6509 some operating systems, this comes back from gethostbyaddr() as an empty
6510 string, and this is what Exim used to test for. However, it seems that in
6511 other systems, "." is yielded. Exim now tests for this case too.
6513 7. The values of check_spool_space and check_log_space are now held internally
6514 as a number of kilobytes instead of an absolute number of bytes. If a
6515 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6516 kilobyte. This means that much larger values can be stored.
6518 8. Exim monitor: an attempt to get the action menu when not actually pointing
6519 at a message produces an empty menu entitled "No message selected". This
6520 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6521 no entries in it ("Shell widget menu has zero width and/or height"). So I
6522 have added a single, blank menu entry in this case.
6524 9. Added ${quote_local_part.
6526 10. MIME decoding is now applied to the contents of Subject: header lines when
6529 11. Now that a reference to $sender_host_address automatically causes a reverse
6530 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6531 host lookup before query-style lookups in lists that might use this
6532 variable. This has therefore been abolished, and the "net-" prefix is no
6533 longer necessary for query-style lookups.
6535 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6536 have been a typo for LFLAGS, so it has been changed.
6538 13. The install script calls Exim with "-C /dev/null" in order to find the
6539 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6540 to be output. However, since Exim outputs its version number before the
6541 error, it didn't break the script. It just looked ugly. I fixed this by
6542 always allowing "-C /dev/null" if the caller is root.
6544 14. Ignore overlarge ACL variable number when reading spool file - insurance
6545 against a later release with more variables having written the file.
6547 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6548 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6549 this, as well as the form without the "IPv6" on the front.
6551 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6552 OS/Makefile-Darwin file.
6554 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6555 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6557 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6558 variable. However, some OS use ints for inodes. I've added cast to long int
6559 to get rid of the compiler warning.
6561 19. I had forgotten to lock out "/../" in configuration file names when
6562 ALT_CONFIG_PREFIX was set.
6564 20. Routers used for verification do not need to specify transports. However,
6565 if such a router generated a host list, and callout was configured, Exim
6566 crashed, because it could not find a port number from the (non-existent)
6567 transport. It now assumes port 25 in this circumstance.
6569 21. Added the -t option to exigrep.
6571 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6572 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6573 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6574 them is set, LOOKUP_LSEARCH is forced.
6576 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6578 24. Added sender and host information to the "rejected by local_scan()" log
6579 line; previously there was no indication of these.
6581 25. Added .include_if_exists.
6583 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6584 Linux. It turns out that not all file systems support this. Apparently some
6585 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6586 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6587 not supported on the file descriptor, is now ignored when Exim is trying to
6588 sync a directory. This applies only to Linux.
6590 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6592 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6593 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6596 29. Added support for maildirsize files from supplied patch (modified a bit).
6598 30. The use of :fail: followed by an empty string could lead Exim to respond to
6599 sender verification failures with (e.g.):
6601 550 Verification failed for <xxx>
6602 550 Sender verify failed
6604 where the first response line was missing the '-' that indicates it is not
6605 the final line of the response.
6607 31. The loop for finding the name of the user that called Exim had a hardwired
6608 limit of 10; it now uses the value of finduser_retries, which is used for
6609 all other user lookups.
6611 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6614 33. Exim was neglecting to zero errno before one call of strtol() when
6615 expanding a string and expecting an integer value. On some systems this
6616 resulted in spurious "integer overflow" errors. Also, it was casting the
6617 result into an int without checking.
6619 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6620 did not work. The code looks as if it has *never* worked, though it appears
6621 to have been documented since at least release 1.62. I have made it work.
6623 35. The "timeout_DNS" error in retry rules, also documented since at least
6624 1.62, also never worked. As it isn't clear exactly what this means, and
6625 clearly it isn't a major issue, I have abolished the feature by treating it
6626 as "timeout", and writing a warning to the main and panic logs.
6628 36. The display of retry rules for -brt wasn't always showing the error code
6631 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6632 timeout_connect_A, timeout_connect_MX.
6634 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6635 to the empty sender.
6637 39. The daemon was not analysing the content of -oX till after it had closed
6638 stderr and disconnected from the controlling terminal. This meant that any
6639 syntax errors were only noted on the panic log, and the return code from
6640 the command was 0. By re-arranging the code a little, I've made the
6641 decoding happen first, so such errors now appear on stderr, and the return
6642 code is 1. However, the actual setting up of the sockets still happens in
6643 the disconnected process, so errors there are still only recorded on the
6646 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6647 connections (as happens on some IP stacks) was logged at start up time as
6648 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6649 it from "IPv6 and IPv4", which means that two separate sockets are being
6652 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6653 says whether AF_INET or AF_INET6 was passed as an argument.
6655 42. Exiwhat output was messed up when time zones were included in log
6658 43. Exiwhat now gives more information about the daemon's listening ports,
6659 and whether -tls-on-connect was used.
6661 44. The "port" option of the smtp transport is now expanded.
6663 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6664 silently ignored. Now an error message is written to the main and panic
6667 46. There's a new ACL modifier called "logwrite" which writes to a log file
6668 as soon as it is encountered.
6670 47. Added $local_user_uid and $local_user_gid at routing time.
6672 48. Exim crashed when trying to verify a sender address that was being
6675 49. Exim was recognizing only a space character after ".include". It now also
6676 recognizes a tab character.
6678 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6679 extracting the relevant information from the specification. The man page no
6680 longer contains scrambled data for the -d option, and I've added a section
6681 at the front about calling Exim under different names.
6683 51. Added "extra_headers" argument to the "mail" command in filter files.
6685 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6688 53. Installed eximstats 1.29.
6690 54. Added transport_filter_timeout as a generic transport option.
6692 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6693 Cc: header lines. This was required by RFC 822, but it not required by RFC
6696 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6697 incoming messages that did not have them. Now it does so only if the
6698 message originates locally, that is, if there is no associated remote host
6699 address. When Resent- header lines are present, this applies to the Resent-
6700 lines rather than the non-Resent- lines.
6702 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6703 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6705 58. Messages for configuration errors now include the name of the main
6706 configuration file - useful now that there may be more than one file in a
6707 list (.included file names were always shown).
6709 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6710 for those rare installations that do not start the daemon as root or run it
6711 setuid root. I've cut out the call to initgroups() if the daemon is not
6714 60. The Exim user and group can now be bound into the binary as text strings
6715 that are looked up at the start of Exim's processing.
6717 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6719 62. Added $mailstore_basename variable.
6721 63. Installed patch to sieve.c from Michael Haardt.
6723 64. When Exim failed to open the panic log after failing to open the main log,
6724 the original message it was trying to log was written to stderr and debug
6725 output, but if they were not available (the usual case in production), it
6726 was lost. Now it is written to syslog before the two lines that record the
6727 failures to open the logs.
6729 65. Users' Exim filters run in subprocesses under the user's uid. It is
6730 possible for a "deliver" command or an alias in a "personal" command to
6731 provoke an address rewrite. If logging of address rewriting is configured,
6732 this fails because the process is not running as root or exim. There may be
6733 a better way of dealing with this, but for the moment (because 4.30 needs
6734 to be released), I have disabled address rewrite logging when running a
6735 filter in a non-root, non-exim process.
6741 1. The buildconfig auxiliary program wasn't quoting the value set for
6742 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6743 not defined. This bug was masked in 4.22 by the effect that was fixed in
6746 2. Some messages that were rejected after a message id was allocated were
6747 shown as "incomplete" by exigrep. It no longer does this for messages that
6748 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6750 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6751 have allow_domain_literals set, the ID did not get logged in the <= line.
6752 Domain literals are now always recognized in Message-ID: header lines.
6754 4. The first argument for a ${extract expansion item is the key name or field
6755 number. Leading and trailing spaces in this item were not being ignored,
6756 causing some misleading effects.
6758 5. When deliver_drop_privilege was set, single queue runner processes started
6759 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6760 same command in the process it spins off) were not dropping privilege.
6762 6. When the daemon running as "exim" started a queue runner, it always
6763 re-executed Exim in the spun-off process. This is a waste of effort when
6764 deliver_drop_privilege is set. The new process now just calls the
6765 queue-runner function directly.
6771 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6774 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6775 ignored. Though the use of -oP was forcing the writing of a pid file, it
6776 was always written to the default place.
6778 3. If the message "no IP address found for host xxxx" is generated during
6779 incoming verification, it is now followed by identification of the incoming
6780 connection (so you can more easily find what provoked it).
6782 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6784 5. Added some features to "harden" Exim a bit more against certain attacks:
6786 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6787 be put in Local/Makefile. This is like the never_users runtime option,
6788 but it cannot be overridden. The default setting is "root".
6790 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6791 prefix string with which any file named in a -C command line option
6794 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6795 is retained for -C and -D only if the caller of Exim is root. Without
6796 it, the exim user may also use -C and -D and retain privilege.
6798 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6799 command line option is disabled.
6801 6. Macro names set by the -D option must start with an upper case letter, just
6802 like macro names defined in the configuration file.
6804 7. Added "dereference=" facility to LDAP.
6806 8. Two instances of the typo "uknown" in the source files are fixed.
6808 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6809 the Configure-Makefile script screwed up while processing it.
6811 10. Incorporated PCRE 4.4.
6813 11. The SMTP synchronization check was not operating right at the start of an
6814 SMTP session. For example, it could not catch a HELO sent before the client
6815 waited for the greeting. There is now a check for outstanding input at the
6816 point when the greeting is written. Because of the duplex, asynchronous
6817 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6818 way, but not yet received, when the check is performed.
6820 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6821 on TCP/IP sockets, because this apparently causes some broken clients to
6824 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6825 unchanged) from the Cygwin maintainer.
6827 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6828 was supported instead of testing for mbx. Effectively a typo.
6830 15. The spa authenticator server code was not checking that the input it
6831 received was valid base64.
6833 16. The debug output line for the "set" modifier in ACLs was not showing the
6834 name of the variable that was being set.
6836 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6838 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6839 lookup on its own. Something else was needed to trigger the lookup. For
6840 example, a match in host_lookup or the need for a host name in a host list.
6841 Now, if $sender_host_name is referenced and the host name has not yet been
6842 looked up, a lookup is performed. If the lookup fails, the variable remains
6843 empty, and $host_lookup_failed is set to "1".
6845 19. Added "eqi" as a case-independent comparison operator.
6847 20. The saslauthd authentication condition could segfault if neither service
6848 nor realm was specified.
6850 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6851 error message that was logged was misleading, and incoming SMTP
6852 connections were dropped. The message is now more accurate, and temporary
6853 errors are given to SMTP connections.
6855 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6856 (except RSET) with a 421 error, until QUIT is received. However, it was
6857 failing to send a response to QUIT.
6859 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6860 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6861 ACL is now run after the helo_try_verify_hosts code.
6863 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6864 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6865 case-independent, so other case variants are also recognized). Apparently
6866 some systems use these upper case variants.
6868 25. If more than two messages were waiting for the same host, and a transport
6869 filter was specified for the transport, Exim sent two messages over the
6870 same TCP/IP connection, and then failed with "socket operation on non-
6871 socket" when it tried to send the third.
6873 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6875 27. The extern definition of crypt16() in expand.c was not being excluded when
6876 the OS had its own crypt16() function.
6878 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6879 as a preferred synonym for return_size_limit, both as an option and as an
6882 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6884 30. Changed the default configuration ACL to relax the local part checking rule
6885 for addresses that are not in any local domains. For these addresses,
6886 slashes and pipe symbols are allowed within local parts, but the sequence
6887 /../ is explicitly forbidden.
6889 31. SPA server authentication was not clearing the challenge buffer before
6892 32. log_message in a "warn" ACL statement was writing to the reject log as
6893 well as to the main log, which contradicts the documentation and doesn't
6894 seem right (because no rejection is happening). So I have stopped it.
6896 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6897 However, I am unable to do any testing of this.
6899 34. Fixed an infelicity in the appendfile transport. When checking directories
6900 for a mailbox, to see if any needed to be created, it was accidentally
6901 using path names with one or more superfluous leading slashes; tracing
6902 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6904 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6905 contents are added to the log line that is written for every discarded
6906 recipient. (Previously a log_message setting was ignored.)
6908 36. The ${quote: operator now quotes the string if it is empty.
6910 37. The install script runs exim in order to find its version number. If for
6911 some reason other than non-existence or emptiness, which it checks, it
6912 could not run './exim', it was installing it with an empty version number,
6913 i.e. as "exim-". This error state is now caught, and the installation is
6916 38. An argument was missing from the function that creates an error message
6917 when Exim fails to connect to the socket for saslauthd authentication.
6918 This could cause Exim to crash, or give a corrupted message.
6920 39. Added isip, isip4, and isip6 to ${if conditions.
6922 40. The ACL variables $acl_xx are now saved with the message, and can be
6923 accessed later in routers, transports, and filters.
6925 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6926 strings in the file are not string-expanded.
6928 42. If a MAIL command specified a SIZE value that was too large to fit into an
6929 int variable, the check against message_size_limit failed. Such values are
6930 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6931 day this will have to be increased, but I don't think I want to be around
6932 when emails are that large.
6939 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6940 iconv() is not standard in FreeBSD.
6942 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6943 IPv6 enabled. The observed symptom was a segmentation fault on return from
6944 the function os_common_find_running_interfaces() in src/os.c.
6946 3. In the check_special_case() function in daemon.c I had used "errno" as an
6947 argument name, which causes warnings on some systems. This was basically a
6948 typo, since it was named "eno" in the comments!
6950 4. The code that waits for the clock to tick (at a resolution of some fraction
6951 of a second) so as to ensure message-id uniqueness was always waiting for
6952 at least one whole tick, when it could have waited for less. [This is
6953 almost certainly not relevant at current processor speeds, where it is
6954 unlikely to ever wait at all. But we try to future-proof.]
6956 5. The function that sleeps for a time interval that includes fractions of a
6957 second contained a race. It did not block SIGALRM between setting the
6958 timer, and suspending (a couple of lines later). If the interval was short
6959 and the sigsuspend() was delayed until after it had expired, the suspension
6960 never ended. On busy systems this could lead to processes getting stuck for
6963 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6964 process, before it forks any delivery processes. The open lookup caching
6965 mechanism meant that the open file or database connection was passed into
6966 the delivery process. The problem was that delivery processes always tidy
6967 up cached lookup data. This could cause a problem for the next delivery
6968 process started by the queue runner, because the external queue runner
6969 process does not know about the closure. So the next delivery process
6970 still has data in the lookup cache. In the case of a file lookup, there was
6971 no problem because closing a file descriptor in a subprocess doesn't affect
6972 the parent. However, if the lookup was caching a connection to a database,
6973 the connection was closed, and the second delivery process was likely to
6974 see errors such as "PGSQL: query failed: server closed the connection
6975 unexpectedly". The problem has been fixed by closing all cached lookups
6976 in a queue runner before running a delivery process.
6978 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6979 seem to have the "const" qualifier which it has on other OS. I've
6982 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6983 *of the same type* (client or server) with the same public name that an
6984 error should be diagnosed.
6986 9. When Exim looked up a host name for an IP address, but failed to find the
6987 original IP address when looking up the host name (a safety check), it
6988 output the message "<ip address> does not match any IP for NULL", which was
6989 confusing, to say the least. The bug was that the host name should have
6990 appeared instead of "NULL".
6992 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6993 user that is built into the binary, and the -C or -D options is used, root
6994 privilege is dropped before the configuration file is read. In addition,
6995 logging is switched to stderr instead of the normal log files. If the
6996 configuration then re-defines the Exim user, the unprivileged environment
6997 is probably not what is expected, so Exim logs a panic warning message (but
7000 However, if deliver_drop_privilege is set, the unprivileged state may well
7001 be exactly what is intended, so the warning has been cut out in that case,
7002 and Exim is allowed to try to write to its normal log files.
7008 1. smtp_return_error_details was not giving details for temporary sender
7009 or receiver verification errors.
7011 2. Diagnose a configuration error if two authenticators have the same public
7014 3. Exim used not to create the message log file for a message until the first
7015 delivery attempt. This could be confusing when incoming messages were held
7016 for policy or load reasons. The message log file is now created at the time
7017 the message is received, and an initial "Received" line is written to it.
7019 4. The automatically generated man page for command line options had a minor
7020 bug that caused no ill effects; however, a more serious problem was that
7021 the procedure for building the man page automatically didn't always
7022 operate. Consequently, release 4.20 contains an out-of-date version. This
7023 shouldn't happen again.
7025 5. When building Exim with embedded Perl support, the script that builds the
7026 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7027 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7029 6. The freeze_tell option was not being used for messages that were frozen on
7030 arrival, either by an ACL or by local_scan().
7032 7. Added the smtp_incomplete_transaction log selector.
7034 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7035 was accepting AUTH without a new EHLO.
7037 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7038 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7041 10. Logging of TCP/IP connections (when configured) now happens in the main
7042 daemon process instead of the child process, so that the TCP/IP connection
7043 count is more accurate (but it can never be perfect).
7045 11. The use of "drop" in a nested ACL was not being handled correctly in the
7046 outer ACL. Now, if condition failure induced by the nested "drop" causes
7047 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7048 or "require"), the connection is dropped.
7050 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7051 that yield "discard" can now be used with an "accept" or a "discard" verb,
7052 but an error is generated for any others (because I can't see a useful way
7053 to define what should happen).
7055 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7056 are now processed in the same way as lines in the Exim configuration file.
7057 In particular, continuation lines are supported.
7059 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7061 15. Added -ti meaning -t -i.
7063 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7064 domains, and warn by logging if others are found.
7066 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7067 returned from GIFCONF when Exim is trying to find the list of interfaces on
7068 a host. The code in os.c has been modified to copy each ifreq to an aligned
7069 structure in all cases.
7071 Also, in some cases, the returned ifreq's were being copied to a 'struct
7072 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7073 means the last couple of bytes of an IPv6 address could be chopped if the
7074 ifreq contained only a normal sockaddr (14 bytes storage).
7076 18. Named domain lists were not supported in the hosts_treat_as_local option.
7077 An entry such as +xxxx was not recognized, and was treated as a literal
7080 19. Ensure that header lines added by a DATA ACL are included in the reject log
7081 if the ACL subsequently rejects the message.
7083 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7084 MD5 (which is deprecated).
7086 21. When testing a filter file using -bf, Exim was writing a message when it
7087 took the sender from a "From " line in the message, but it was not doing so
7088 when it took $return_path from a Return-Path: header line. It now does.
7090 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7091 with a valid header line field name (a series of printing characters
7092 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7094 23. Changed "disc" in the source to "disk" to conform to the documentation and
7095 the book and for uniformity.
7097 24. Ignore Sendmail's -Ooption=value command line item.
7099 25. When execve() failed while trying to run a command in a pipe transport,
7100 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7101 could be confused with a return value of 69 from the command itself. This
7102 has been changed to 127, the value the shell returns if it is asked to run
7103 a non-existent command. The wording for the related log line suggests a
7104 non-existent command as the problem.
7106 26. If received_header_text expands to an empty string, do not add a Received:
7107 header line to the message. (Well, it adds a token one on the spool, but
7108 marks it "old" so that it doesn't get used or transmitted.)
7110 27. Installed eximstats 1.28 (addition of -nt option).
7112 28. There was no check for failure on the call to getsockname() in the daemon
7113 code. This can fail if there is a shortage of resources on the system, with
7114 ENOMEM, for example. A temporary error is now given on failure.
7116 29. Contrary to the C standard, it seems that in some environments, the
7117 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7118 program. Exim now does this explicitly; it affects the formatting of
7119 timestamps using strftime().
7121 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7122 complaints. I've now initialized all the variables, to avoid this.
7124 32. Header lines added by a system filter were not being "seen" during
7125 transport-time rewrites.
7127 33. The info_callback() function passed to OpenSSL is set up with type void
7128 (*)(SSL *, int, int), as described somewhere. However, when calling the
7129 function (actually a macro) that sets it up, the type void(*)() is
7130 expected. I've put in a cast to prevent warnings from picky compilers.
7132 34. If a DNS black list lookup found a CNAME record, but there were no A
7133 records associated with the domain it pointed at, Exim crashed.
7135 35. If a DNS black list lookup returned more than one A record, Exim ignored
7136 all but the first. It now scans all returned addresses if a particular IP
7137 value is being sought. In this situation, the contents of the
7138 $dnslist_value variable are a list of all the addresses, separated by a
7141 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7142 to accept a host name and all its aliases if the forward lookup for any of
7143 them yielded the IP address of the incoming connection. Now it accepts only
7144 those names whose forward lookup yields the correct IP address. Any other
7145 names are discarded. This closes a loophole whereby a rogue DNS
7146 administrator could create reverse DNS records to break through a
7147 wildcarded host restriction in an ACL.
7149 37. If a user filter or a system filter that ran in a subprocess used any of
7150 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7151 the wrong values were passed to the pipe command ($thisaddress had the
7152 value of $0, $0 had the value of $1, etc). This bug was introduced by
7153 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7155 38. Improved the line breaking for long SMTP error messages from ACLs.
7156 Previously, if there was no break point between 40 and 75 characters, Exim
7157 left the rest of the message alone. Two changes have been made: (a) I've
7158 reduced the minimum length to 35 characters; (b) if it can't find a break
7159 point between 35 and 75 characters, it looks ahead and uses the first one
7160 that it finds. This may give the occasional overlong line, but at least the
7161 remaining text gets split now.
7163 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7164 file descriptors might be low, and that setting 1000 would always raise it.
7165 It turns out that in some environments, the limit is already over 1000 and
7166 that lowering it causes trouble. So now Exim takes care not to decrease it.
7168 40. When delivering a message, the value of $return_path is set to $sender_
7169 address at the start of routing (routers may change the value). By an
7170 oversight, this default was not being set up when an address was tested by
7171 -bt or -bv, which affected the outcome if any router or filter referred to
7174 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7175 log only when "message" or "log_message" are set. However, if one of the
7176 conditions was an address verification, or a call to a nested ACL, the
7177 messages generated by the underlying test were being passed through. This
7178 no longer happens. The underlying message is available in $acl_verify_
7179 message for both "message" and "log_message" expansions, so it can be
7180 passed through if needed.
7182 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7183 new expansion $bh_ to give the encoded byte string without charset
7184 translation. Translation happens only if iconv() is available; HAVE_ICONV
7185 indicates this at build time. HEADERS_CHARSET gives the charset to
7186 translate to; headers_charset can change it in the configuration, and
7187 "headers charset" can change it in an individual filter file.
7189 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7190 that creates RFC 2047 encoded "words" labels them as that charset instead
7191 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7192 expansion operator; (ii) when Exim creates a From: line for a local
7193 message; (iii) when a header line is rewritten to include a "phrase" part.
7195 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7196 buggy, causing it to skip the first lines of messages whose message ID
7197 ended in 'D'. This would not have bitten before Exim release 4.14, because
7198 message IDs were unlikely to end in 'D' before then. The effect was to have
7199 incorrect size information for certain domains.
7201 45. #include "config.h" was missing at the start of the crypt16.c module. This
7202 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7205 46. If there was a timeout during a "random" callout check, Exim treated it as
7206 a failure of the random address, and carried on sending RSET and the real
7207 address. If the delay was just some slowness somewhere, the response to the
7208 original RCPT would be taken as a response to RSET and so on, causing
7209 mayhem of various kinds.
7211 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7212 when I implemented it. It didn't allow for the fact that some option values
7213 may legitimately be negative (e.g. size_addition), and it didn't even do
7214 the right test for positive values.
7216 48. Domain names in DNS records are case-independent. Exim always looks them up
7217 in lower case. Some resolvers return domain names in exactly the case they
7218 appear in the zone file, that is, they may contain uppercase letters. Not
7219 all resolvers do this - some return always lower case. Exim was treating a
7220 change of case by a resolver as a change of domain, similar to a widening
7221 of a domain abbreviation. This triggered its re-routing code and so it was
7222 trying to route what was effectively the same domain again. This normally
7223 caused routing to fail (because the router wouldn't handle the domain
7224 twice). Now Exim checks for this case specially, and just changes the
7225 casing of the domain that it ultimately uses when it transmits the message
7228 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7231 50. If a filter generated a file delivery with a non-absolute name (possible if
7232 no home directory exists for the router), the forbid_file option was not
7235 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7236 the existing equality matching.
7238 52. Exim was using ints instead of ino_t variables in some places where it was
7239 dealing with inode numbers.
7241 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7242 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7243 called TMPDIR, and if it finds it is different, it changes its value.
7245 54. The smtp_printf() function is now made available to local_scan() so
7246 additional output lines can be written before returning. There is also an
7247 smtp_fflush() function to enable the detection of a dropped connection.
7248 The variables smtp_input and smtp_batched_input are exported to
7251 55. Changed the default runtime configuration: the message "Unknown user"
7252 has been removed from the ACL, and instead placed on the localuser router,
7253 using the cannot_route_message feature. This means that any verification
7254 failures that generate their own messages won't get overridden. Similarly,
7255 the "Unrouteable address" message that was in the ACL for unverifiable
7256 relay addresses has also been removed.
7258 56. Added hosts_avoid_esmtp to the smtp transport.
7260 57. The exicyclog script was not checking for the esoteric option
7261 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7262 will work only if exicyclog is run under the appropriate euid.
7264 58. Following a discussion on the list, the rules by which Exim recognises line
7265 endings on incoming messages have been changed. The -dropcr and drop_cr
7266 options are now no-ops, retained only for backwards compatibility. The
7267 following line terminators are recognized: LF CRLF CR. However, special
7268 processing applies to CR:
7270 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7271 nor a local message in the state where . is a terminator.
7273 (ii) If a bare CR is encountered in a header line, an extra space is added
7274 after the line terminator so as not to end the header. The reasoning
7275 behind this is that bare CRs in header lines are most likely either
7276 to be mistakes, or people trying to play silly games.
7278 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7279 was being incorrectly given as 18 bytes larger than it should have been.
7280 This is a VOB (very old bug).
7282 60. This may never have affected anything current, but just in case it has:
7283 When the local host is found other than at the start of a list of hosts,
7284 the local host, those with the same MX, and any that follow, are discarded.
7285 When the list in question was part of a longer list of hosts, the following
7286 hosts (not currently being processed) were also being discarded. This no
7287 longer happens. I'm not sure if this situation could ever has previously
7290 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7293 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7294 header. This is something that is recommended in a new Internet Draft, and
7295 is something that is documented as being done by Sendmail. There are two
7296 possible values. For messages generated by the autoreply transport, Exim
7299 Auto-Submitted: auto-replied
7301 whereas for all other generated messages (e.g. bounces) it adds
7303 Auto-Submitted: auto-generated
7305 63. The "personal" condition in filters now includes a test for the
7306 Auto-Submitted: header. If it contains the string "auto-" the message it
7307 not considered personal.
7309 64. Added rcpt_include_affixes as a generic transport option.
7311 65. Added queue_only_override (default true).
7313 66. Added the syslog_duplication option.
7315 67. If what should have been the first header line of a message consisted of
7316 a space followed by a colon, Exim was mis-interpreting it as a header line.
7317 It isn't of course - it is syntactically invalid and should therefore be
7318 treated as the start of the message body. The misbehaviour could have
7319 caused a number of strange effects, including loss of data in subsequent
7320 header lines, and spool format errors.
7322 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7323 client host had authenticated. This control can now be exercised by an ACL
7324 for more flexibility.
7326 69. By default, callouts do not happen when testing with -bh. There is now a
7327 variant, -bhc, which does actually run the callout code, including
7328 consulting and updating the callout cache.
7330 70. Added support for saslauthd authentication, courtesy of Alexander
7333 71. If statvfs() failed on the spool or log directories while checking their
7334 size for availability, Exim confusingly gave the error "space shortage".
7335 Furthermore, in debugging mode it crashed with a floating point exception.
7336 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7337 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7338 serious problem, Exim now writes to the main and panic logs when this
7339 happens, with details of the failure. It then refuses to accept the
7340 incoming message, giving the message "spool directory problem" or "log
7341 directory problem" with a 421 code for SMTP messages.
7343 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7344 0, 1, and 2 exist, because some OS complain for execs without them (see
7345 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7346 descriptors. However, the code omitted to check that the open succeeded,
7347 causing mysterious errors if for some reason the permissions on /dev/null
7348 got screwed. Now Exim writes a message to the main and panic logs, and
7349 bombs out if it can't open /dev/null.
7351 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7352 interact so that it is all more flexible. It is supposed to remain
7353 backwards compatible. Also added extra_local_interfaces.
7355 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7356 to bomb out with an assertion failure - to the client this appears as a
7357 connection drop. This problem occurs in the part of the code that was taken
7358 from the Samba project. Fortunately, the assertion is in a very simple
7359 function, so I have fixed this by reproducing the function inline in the
7360 one place where it is called, and arranging for authentication to fail
7361 instead of killing the process with assert().
7363 75. The SPA client code was not working when the server requested OEM rather
7364 than Unicode encoding.
7366 76. Added code to make require_files with a specific uid setting more usable in
7367 the case where statting the file as root fails - usually a non-root-mounted
7368 NFS file system. When this happens and the failure is EACCES, Exim now
7369 forks a subprocess and does the per-uid checking as the relevant uid.
7371 77. Added process_log_path.
7373 78. If log_file_path was not explicitly set, a setting of check_log_space or
7374 check_log_inodes was ignored.
7376 79. If a space check for the spool or log partitions fails, the incident is now
7377 logged. Of course, in the latter case the data may get lost...
7379 80. Added the %p formatting code to string_format() so that it can be used to
7380 print addresses in debug_print(). Adjusted all the address printing in the
7381 debugging in store.c to use %p rather than %d.
7383 81. There was a concern that a line of code in smtp_in.c could overflow a
7384 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7385 initially expressed, the concern was not well-founded, because trailing
7386 spaces are removed early. However, if the trailing spaces were followed by
7387 a NULL, they did not get removed, so the overflow was possible. Two fixes
7390 (a) I re-wrote the offending code in a cleaner fashion.
7391 (b) If an incoming SMTP command contains a NULL character, it is rejected
7394 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7395 runs initgroups(), so that if the Exim user is in any additional groups,
7396 they will be used during message reception.
7402 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob