1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
12 buffer overrun for (non-chunking) other transports.
14 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
15 TLS1.3, means that a server rejecting a client certificate is not visible
16 to the client until the first read of encrypted data (typically the
17 response to EHLO). Add detection for that case and treat it as a failed
18 TLS connection attempt, so that the normal retry-in-clear can work (if
25 JH/01 Remove code calling the customisable local_scan function, unless a new
26 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
28 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
29 non-signal-safe functions being used.
31 JH/03 Bug 2269: When presented with a received message having a stupidly large
32 number of DKIM-Signature headers, disable DKIM verification to avoid
33 a resource-consumption attack. The limit is set at twenty.
35 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
36 report of oldest_pass in ${authres } in consequence, and separate out
37 some descriptions of reasons for verification fail.
39 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
40 files in the spool were present and unlocked. A queue-runner could spot
41 them, resulting in a duplicate delivery. Fix that by doing the unlock
42 after the unlink. Investigation by Tim Stewart. Take the opportunity to
43 add more error-checking on spoolfile handling while that code is being
46 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
47 No known attacks, no CVE, this is defensive hardening.
49 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
50 a queue-runner could start a delivery while other operations were ongoing.
51 Cutthrough delivery was a common victim, resulting in duplicate delivery.
52 Found and investigated by Tim Stewart. Fix by using the open message data
53 file handle rather than opening another, and not locally closing it (which
54 releases a lock) for that case, while creating the temporary .eml format
55 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
57 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
58 $sender_verify_failure/$recipient_verify_failure to "random".
60 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
63 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
64 Previously this would segfault.
66 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
69 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
70 like zero, since the resolver should be doing this for us, But we need one
71 as a CNAME but no MX presence gets the CNAME returned; we need to check
72 that doesn't point to an MX to declare it "no MX returned" rather than
73 "error, loop". A new main option is added so the older capability of
74 following some limited number of chain links is maintained.
76 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
78 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
79 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
80 now-deprecated earlier definitions used only specified the range up to TLS
81 1.2 (in the older-version library docs).
83 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
85 JH/15 Rework TLS client-side context management. Stop using a global, and
86 explicitly pass a context around. This enables future use of TLS for
87 connections to service-daemons (eg. malware scanning) while a client smtp
88 connection is using TLS; with cutthrough connections this is quite likely.
90 JH/16 Fix ARC verification to do AS checks in reverse order.
92 JH/17 Support a "tls" option on the ${readsocket } expansion item.
94 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
95 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
96 Previously the "utf8" would be re-prepended for every additional message.
98 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
99 Previously thery were accepted, resulting in issues when attempting to
100 forward messages to a non-supporting MTA.
102 PP/02 Let -n work with printing macros too, not just options.
104 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
105 one parent address was copied, and bogus data was used at delivery-logging
106 time. Either a crash (after delivery) or bogus log data could result.
107 Discovery and analysis by Tim Stewart.
109 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
110 Previously if the string ended mid-character, we did not insert the
111 promised '?' replacement.
113 PP/04 Documentation: current string operators work on bytes, not codepoints.
115 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
116 should pack well giving a smaller memory footprint so better caching and
117 therefore performance. Group the declarations where this can't be done so
118 that the byte-sized flag variables are not interspersed among pointer
119 variables, giving a better chance of good packing by the compiler.
121 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
122 non-null, to avoid issues with sites running BATV. Previously reports were
123 sent with an empty envelope sender so looked like bounces.
125 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
126 The ignore_error flag wasn't being returned from the filter subprocess so
127 was not set for later routers. Investigation and fix by Matthias Kurz.
129 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
130 and a msg:complete for the whole, when a message is manually removed using
131 -Mrm. Developement by Matthias Kurz, hacked on by JH.
133 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
134 a "Gnu special" function, asprintf() in the DB utility binary builds; I
135 hope that is portable enough.
137 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
138 requiring a known-CA anchor certificate; make it now rely entirely on the
139 TLSA as an anchor. Checking the name on the leaf cert against the name
140 on the A-record for the host is still done for TA (but not for EE mode).
142 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
143 included in delivery lines for non-proxied connections, when compiled with
144 SUPPORT_SOCKS and running with proxy logging enabled.
146 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
147 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
148 move the existing event to fire before the normal logging of message
149 failure so that custom logging is bracketed by normal logging.
151 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
152 msg:fail:internal event. Developement by Matthias Kurz.
154 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
155 far too small for todays use of crypto signatures stored there. Go all
156 the way to the max DNS message size of 64kB, even though this might be
157 overmuch for IOT constrained device use.
159 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
160 copy a string over itself. The library routine is documented as not
161 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
163 JH/32 For main options check_spool_space and check_inode_space, where the
164 platform supports 64b integers, support more than the previous 2^31 kB
165 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
166 the previous G, M, k.
168 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
169 $authenticated_fail_id variable on authentication failure. Previously
172 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
173 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
174 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
175 GNUTLS_SEC_PARAM_MEDIUM.
177 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
178 side. Previously we would continue as if no SNI had been received.
180 JH/36 Harden the handling of string-lists. When a list consisted of a sole
181 "<" character, which should be a list-separator specification, we walked
182 off past the nul-terimation.
184 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
185 causes) even when the retry time is not yet met. Previously they were
186 not, meaning that when (say) an account was over-quota and temp-rejecting,
187 and multiple senders' messages were queued, only one sender would get
188 notified on each configured delay_warning cycle.
190 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
192 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
193 cipher-suites, an error can be left on the stack even for a succeeding
194 accept; this results in impossible error messages when a later operation
197 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
198 return error codes indicating retry. Under TLS1.3 this becomes required.
200 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
201 it only wrote the new authenticators, resulting in a lack of tracking of
202 peer changes of ESMTP extensions until the next cache flush.
204 JH/41 Fix the loop reading a message header line to check for integer overflow,
205 and more-often against header_maxsize. Previously a crafted message could
206 induce a crash of the recive process; now the message is cleanly rejected.
208 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
209 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
215 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
216 When redis_servers is set to a list of > 1 element, and the Redis servers
217 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
218 case of MOVED into a DEFER case instead, thus moving the query onto the
219 next server in the list. For a cluster of N elements, all N servers must
220 be defined in redis_servers.
222 GF/02 Catch and remove uninitialized value warning in exiqsumm
223 Check for existence of @ARGV before looking at $ARGV[0]
225 JH/01 Replace the store_release() internal interface with store_newblock(),
226 which internalises the check required to safely use the old one, plus
227 the allocate and data copy operations duplicated in both (!) of the
228 extant use locations.
230 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
231 modifier. This matches the restriction on the commandline.
233 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
234 Previously only the last row was returned.
236 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
237 we assumed that tags in the header were well-formed, and parsed the
238 element content after inspecting only the first char of the tag.
239 Assumptions at that stage could crash the receive process on malformed
242 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
243 While running the DKIM ACL we operate on the Permanent memory pool so that
244 variables created with "set" persist to the DATA ACL. Also (at any time)
245 DNS lookups that fail create cache records using the Permanent pool. But
246 expansions release any allocations made on the current pool - so a dnsdb
247 lookup expansion done in the DKIM ACL releases the memory used for the
248 DNS negative-cache, and bad things result. Solution is to switch to the
249 Main pool for expansions.
250 While we're in that code, add checks on the DNS cache during store_reset,
251 active in the testsuite.
252 Problem spotted, and debugging aided, by Wolfgang Breyha.
254 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
255 When none of the hosts presented to a transport match an already-open
256 connection, close it and proceed with the list. Previously we would
257 queue the message. Spotted by Lena with Yahoo, probably involving
260 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
261 Previously a spurious "250 OK id=" response was appended to the proper
264 JH/08 The "support for" informational output now, which built with Content
265 Scanning support, has a line for the malware scanner interfaces compiled
266 in. Interface can be individually included or not at build time.
268 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
269 by the template makefile "src/EDITME". The "STREAM" support for an older
270 ClamAV interface method is removed.
272 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
273 rows affected is given instead).
275 JH/11 The runtime Berkeley DB library version is now additionally output by
276 "exim -d -bV". Previously only the compile-time version was shown.
278 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
279 SMTP connection. Previously, when one had more receipients than the
280 first, an abortive onward connection was made. Move to full support for
281 multiple onward connections in sequence, handling cutthrough connection
282 for all multi-message initiating connections.
284 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
285 routers. Previously, a multi-recipient message would fail to match the
286 onward-connection opened for the first recipient, and cause its closure.
288 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
289 a timeout on read on a GnuTLS initiating connection, resulting in the
290 initiating connection being dropped. This mattered most when the callout
291 was marked defer_ok. Fix to keep the two timeout-detection methods
294 JH/15 Relax results from ACL control request to enable cutthrough, in
295 unsupported situations, from error to silently (except under debug)
296 ignoring. This covers use with PRDR, frozen messages, queue-only and
299 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
301 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
302 metadata, resulting in a crash in free().
304 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
305 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
306 Broken also in d185889f4, with init system revamp.
308 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
309 Previously we abruptly closed the connection after reading a malware-
310 found indication; now we go on to read the "scan ok" response line,
313 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
314 ACL. Previously, a crash would result.
316 JH/19 Speed up macro lookups during configuration file read, by skipping non-
317 macro text after a replacement (previously it was only once per line) and
318 by skipping builtin macros when searching for an uppercase lead character.
320 JH/20 DANE support moved from Experimental to mainline. The Makefile control
321 for the build is renamed.
323 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
324 was allocated for every new TLS startup, meaning one per message. Fix
325 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
327 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
328 reported the original. Fix to report (as far as possible) the ACL
329 result replacing the original.
331 JH/23 Fix memory leak during multi-message connections using STARTTLS under
332 OpenSSL. Certificate information is loaded for every new TLS startup,
333 and the resources needed to be freed.
335 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
337 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
340 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
341 DATA response info to the (existing) per-recipient response info for
342 the "C=" log element. It can have useful tracking info from the
343 destination system. Patch from Simon Arlott.
345 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
346 length value. Previously this would segfault.
348 HS/02 Support Avast multiline protoocol, this allows passing flags to
349 newer versions of the scanner.
351 JH/28 Ensure that variables possibly set during message acceptance are marked
352 dead before release of memory in the daemon loop. This stops complaints
353 about them when the debug_store option is enabled. Discovered specifically
354 for sender_rate_period, but applies to a whole set of variables.
355 Do the same for the queue-runner and queue-list loops, for variables set
356 from spool message files. Do the same for the SMTP per-message loop, for
357 certain variables indirectly set in ACL operations.
359 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
360 as a multi-recipient message from a mailinglist manager). The coding had
361 an arbitrary cutoff number of characters while checking for more input;
362 enforced by writing a NUL into the buffer. This corrupted long / fast
363 input. The problem was exposed more widely when more pipelineing of SMTP
364 responses was introduced, and one Exim system was feeding another.
365 The symptom is log complaints of SMTP syntax error (NUL chars) on the
366 receiving system, and refused recipients seen by the sending system
367 (propating to people being dropped from mailing lists).
368 Discovered and pinpointed by David Carter.
370 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
371 replaced by the ${authresults } expansion.
373 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
375 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
376 allows proper process termination in container environments.
378 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
379 Previously the "final dot" had a newline after it; ensure it is CR,LF.
381 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
382 and "err_perm", deprecated since 4.83 when the RFC-defined words
383 "temperror" and "permerror" were introduced.
385 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
386 transport-filters or DKIM-signing. The restriction was lost in the
387 consolidation of verify-callout and delivery SMTP handling.
388 Extend the restriction to also cover ARC-signing.
390 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
391 in defer=pass mode supply a 450 to the initiator. Previously the message
394 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
395 tls_require_ciphers is used as before.
397 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
398 "pass_unscanned". Only tmpfails from the scanner are written to
399 the paniclog, as they may require admin intervention (permission
400 denied, license issues). Other scanner errors (like decompression
401 bombs) do not cause a paniclog entry.
403 JH/36 Fix reinitialisation of DKIM logging variable between messages.
404 Previously it was possible to log spurious information in receive log
407 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
408 triggered odd behaviour from Outlook Express clients.
410 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
413 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
414 since the IETF WG has not yet settled on that versus the original
415 "bare" representation.
417 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
418 Previously the millisecond value corrupted the output.
419 Fix also for syslog_pid=no and log_selector +pid, for which the pid
420 corrupted the output.
426 JH/01 Rework error string handling in TLS interface so that the caller in
427 more cases is responsible for logging. This permits library-sourced
428 string to be attached to addresses during delivery, and collapses
429 pairs of long lines into single ones.
431 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
432 during configuration. Wildcards are allowed and expanded.
434 JH/02 Rework error string handling in DKIM to pass more info back to callers.
435 This permits better logging.
437 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
438 do not close it down and have the child transport start it up again on
439 the passed-on TCP connection. Instead, proxy the child (and any
440 subsequent ones) for TLS via a unix-domain socket channel. Logging is
441 affected: the continued delivery log lines do not have any DNSSEC, TLS
442 Certificate or OCSP information. TLS cipher information is still logged.
444 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
445 identical IP addresses on different listening ports. Will also affect
448 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
449 add noisy ifdef guards to special-case this sillyness.
450 Patch from Bernd Kuhls.
452 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
453 than 255 are no longer allowed.
455 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
456 Disable the session-cache too, which might reduce our load. Since we
457 currrectly use a new context for every connection, both as server and
458 client, there is no benefit for these.
459 GnuTLS appears to not support tickets server-side by default (we don't
460 call gnutls_session_ticket_enable_server()) but client side is enabled
461 by default on recent versions (3.1.3 +) unless the PFS priority string
464 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
465 <https://reproducible-builds.org/specs/source-date-epoch/>.
467 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
468 the check for any unsuccessful recipients did not notice the limit, and
469 erroneously found still-pending ones.
471 JH/08 Pipeline CHUNKING command and data together, on kernels that support
472 MSG_MORE. Only in-clear (not on TLS connections).
474 JH/09 Avoid using a temporary file during transport using dkim. Unless a
475 transport-filter is involved we can buffer the headers in memory for
476 creating the signature, and read the spool data file once for the
477 signature and again for transmission.
479 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
480 4.77 as the kernel support then wasn't solid, having issues in 64bit
481 mode. Now, it's been long enough. Add support for FreeBSD also.
483 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
484 case where the routing stage had gathered several addresses to send to
485 a host before calling the transport for the first, we previously failed
486 to close down TLS in the old transport process before passing the TCP
487 connection to the new process. The new one sent a STARTTLS command
488 which naturally failed, giving a failed delivery and bloating the retry
489 database. Investigation and fix prototype from Wolfgang Breyha.
491 JH/12 Fix check on SMTP command input synchronisation. Previously there were
492 false-negatives in the check that the sender had not preempted a response
493 or prompt from Exim (running as a server), due to that code's lack of
494 awareness of the SMTP input buffering.
496 PP/04 Add commandline_checks_require_admin option.
497 Exim drops privileges sanely, various checks such as -be aren't a
498 security problem, as long as you trust local users with access to their
499 own account. When invoked by services which pass untrusted data to
500 Exim, this might be an issue. Set this option in main configuration
501 AND make fixes to the calling application, such as using `--` to stop
504 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
505 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
506 responses to those, into a single TLS record each way (this usually means
507 a single packet). As a side issue, smtp_enforce_sync now works on TLS
510 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
511 affects you only if you're dancing at the edge of the param size limits.
512 If you are, and this message makes sense to you, then: raise the
513 configured limit or use OpenSSL 1.1. Nothing we can do for older
516 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
517 cmdline element to get the documented default one. Previously it was
520 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
521 are unsuable for non-message-specific reasons - eg. connection timeout,
522 banner-time rejection.
524 JH/16 Fix logging of delivery remote port, when specified by router, under
527 PP/06 Repair manualroute's ability to take options in any order, even if one
528 is the name of a transport.
531 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
533 JH/17 Change the list-building routines interface to use the expanding-string
534 triplet model, for better allocation and copying behaviour.
536 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
537 Previously it was constructed the first time a possibly-matching string
538 was met in the configuration file input during startup; now it is done
541 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
542 compatible one, to avoid the (poorly documented) possibility of a config
543 file in the working directory redirecting the DB files, possibly correpting
544 some existing file. CVE-2017-10140 assigned for BDB.
546 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
547 cache-hot. Previously, although the result was properly cached, the
548 initial verify call returned a defer.
550 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
551 the main verify for receipient in uncached-mode.
553 JH/22 Retire historical build files to an "unsupported" subdir. These are
554 defined as "ones for which we have no current evidence of testing".
556 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
557 if present. Previously it was ignored.
559 JH/24 Start using specified-initialisers in C structure init coding. This is
560 a C99 feature (it's 2017, so now considered safe).
562 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
563 if was a fixed-sized field and bitmask ops via macros; it is now more
566 PP/07 GitHub PR 56: Apply MariaDB build fix.
567 Patch provided by Jaroslav Škarvada.
569 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
570 during Coverity cleanups [4.87 JH/47]
571 Diagnosis and fix provided by Michael Fischer v. Mollard.
573 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
574 the right size to place the terminating semicolon on its own folded
575 line, the header hash was calculated to an incorrect value thanks to
576 the (relaxed) space the fold became.
578 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
579 and confused the parent.
581 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
582 which could crash as a result. This could lead to undeliverable messages.
584 JH/28 Logging: "next input sent too soon" now shows where input was truncated
587 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
588 matters on fast-turnover and PID-randomising systems, which were getting
589 out-of-order delivery.
591 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
592 a possibly-overlapping copy. The symptom was that "Remote host closed
593 connection in response to HELO" was logged instead of the actual 4xx
596 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
597 Previously only that bufferd was discarded, resulting in SYMTP command
600 JH/32 DKIM: when a message has multiple signatures matching an identity given
601 in dkim_verify_signers, run the dkim acl once for each. Previously only
602 one run was done. Bug 2189.
604 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
605 "panic the current process" to "deliberately defer". The panic log is
606 still written with the problem list name; the mail and reject logs now
607 get a temp-reject line for the message that was being handled, saying
608 something like "domains check lookup or other defer". The SMTP 451
609 message is still "Temporary local problem".
611 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
612 A crafted sequence of BDAT commands could result in in-use memory beeing
613 freed. CVE-2017-16943.
615 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
616 from SMTP input. Previously it was always done; now only done for DATA
617 and not BDAT commands. CVE-2017-16944.
619 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
620 to the message (such as an overlong header line). Previously this was
621 not done and we did not exit BDAT mode. Followon from the previous item
622 though a different problem.
628 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
629 than -2003 did; needs libidn2 in addition to libidn.
631 JH/02 The path option on a pipe transport is now expanded before use.
633 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
634 Patch provided by "Björn", documentation fix added too.
636 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
637 missing a wire-to-host endian conversion.
639 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
640 close after a BDAT command line could be taken as a following command,
641 giving a synch failure. Fix by only checking for synch immediately
642 before acknowledging the chunk.
644 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
645 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
646 macro. Patches provided by Josh Soref.
648 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
649 Previously we did not; the RFC seems ambiguous and VRFY is not listed
650 by IANA as a service extension. However, John Klensin suggests that we
653 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
654 the dkim code may be unix-mode line endings rather than smtp wire-format
655 CRLF, so prepend a CR to any bare LF.
657 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
658 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
660 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
661 the first were themselves being wrongly included in the feed into dkim
662 processing; with most chunk sizes in use this resulted in an incorrect
663 body hash calculated value.
665 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
666 DKIM signature block, for verification. Although advised against by
667 standards it is specifically not ruled illegal.
669 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
671 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
672 missing a body hash (the bh= tag).
674 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
675 It seems that HAProxy sends the Proxy Protocol information in clear and
676 only then does a TLS startup, so do the same.
678 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
679 TCP connections (such as for Spamd) unless the daemon successfully set
680 Fast Open mode on its listening sockets. This fixes breakage seen on
681 too-old kernels or those not configured for Fast Open, at the cost of
682 requiring both directions being enabled for TFO, and TFO never being used
683 by non-daemon-related Exim processes.
685 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
686 endings, at least on the first header line. Try to canonify any that get
687 past that check, despite the cost.
689 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
690 now limited to an arbitrary five deep, while parsing addresses with the
691 strip_excess_angle_brackets option enabled.
693 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
694 instead leave the unprompted TLS handshake in socket buffer for the
695 TLS library to consume.
697 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
699 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
701 JH/16 Drop variables when they go out of scope. Memory management drops a whole
702 region in one operation, for speed, and this leaves assigned pointers
703 dangling. Add checks run only under the testsuite which checks all
704 variables at a store-reset and panics on a dangling pointer; add code
705 explicitly nulling out all the variables discovered. Fixes one known
706 bug: a transport crash, where a dangling pointer for $sending_ip_address
707 originally assigned in a verify callout, is re-used.
709 PP/06 Drop '.' from @INC in various Perl scripts.
711 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
713 PP/08 Reduce a number of compilation warnings under clang; building with
714 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
715 should be warning-free.
717 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
719 HS/01 Fix portability problems introduced by PP/08 for platforms where
720 realloc(NULL) is not equivalent to malloc() [SunOS et al].
722 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
723 chunk. This allows us to accept broken chunked messages. We need a more
724 general solution here.
726 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
727 already-broken messages in the queue.
729 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
731 JH/19 Fix reference counting bug in routing-generated-address tracking.
737 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
738 supports it and a size is available (ie. the sending peer gave us one).
740 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
741 years of being deprecated). The replacements are the ACLs
742 acl_smtp_mime and acl_not_smtp_mime.
744 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
745 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
746 or even an in-clear connection were permitted. Now, if the host lookup
747 was dnssec and dane was requested then the host is only used if the
748 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
749 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
750 if one fails this test.
751 This means that a poorly-configured remote DNS will make it incommunicado;
752 but it protects against a DNS-interception attack on it.
754 JH/04 Bug 1810: make continued-use of an open smtp transport connection
755 non-noisy when a race steals the message being considered.
757 JH/05 If main configuration option tls_certificate is unset, generate a
758 self-signed certificate for inbound TLS connections.
760 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
761 in rewrites and routers.
763 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
764 and logged a warning sing 4.83; now they are a configuration file error.
766 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
767 (lacking @domain). Apply the same qualification processing as RCPT.
769 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
771 JH/10 Support ${sha256:} applied to a string (as well as the previous
774 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
775 a cutthrough deliver is pending, as we always want to make a connection.
776 This also avoids re-routing the message when later placing the cutthrough
777 connection after a verify cache hit.
778 Do not update it with the verify result either.
780 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
781 when routing results in more than one destination address.
783 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
784 signing (which inhibits the cutthrough capability). Previously only
785 the presence of an option was tested; now an expansion evaluating as
786 empty is permissible (obviously it should depend only on data available
787 when the cutthrough connection is made).
789 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
790 the relevant preceding SMTP command did not note the pipelining mode.
792 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
793 Previously they were not counted.
795 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
796 as one having no matching records. Previously we deferred the message
797 that needed the lookup.
799 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
800 distinguished as "(=".
802 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
803 for missing MX records. Previously it only worked for missing A records.
805 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
807 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
808 after the data-go-ahead and data-ack. Patch from Jason Betts.
810 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
811 even for a "none" policy. Patch from Tony Meyer.
813 JH/22 Fix continued use of a connection for further deliveries. If a port was
814 specified by a router, it must also match for the delivery to be
817 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
818 When one of the recipients of a message was unsuitable for the connection
819 (has no matching addresses), we lost track of needing to mark it
820 deferred. As a result mail would be lost.
822 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
824 JH/25 Decoding ACL controls is now done using a binary search; the source code
825 takes up less space and should be simpler to maintain. Merge the ACL
826 condition decode tables also, with similar effect.
828 JH/26 Fix problem with one_time used on a redirect router which returned the
829 parent address unchanged. A retry would see the parent address marked as
830 delivered, so not attempt the (identical) child. As a result mail would
833 JH/27 Fix a possible security hole, wherein a process operating with the Exim
834 UID can gain a root shell. Credit to http://www.halfdog.net/ for
835 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
838 JH/28 Enable {spool,log} filesystem space and inode checks as default.
839 Main config options check_{log,spool}_{inodes,space} are now
840 100 inodes, 10MB unless set otherwise in the configuration.
842 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
843 Previously it only applied to the main-section connection policy
846 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
848 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
849 by me. Added RFC7919 DH primes as an alternative.
851 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
852 are not in the system include path.
854 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
855 GnuTLS, when a session startup failed (eg because the client disconnected)
856 Exim did stdio operations after fclose. This was exposed by a recent
857 change which nulled out the file handle after the fclose.
859 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
860 signed directly by the cert-signing cert, rather than an intermediate
861 OCSP-signing cert. This is the model used by LetsEncrypt.
863 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
865 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
866 an incoming connection.
868 HS/02 Bug 1802: Do not half-close the connection after sending a request
871 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
872 fallback to "prime256v1".
874 JH/34 SECURITY: Use proper copy of DATA command in error message.
875 Could leak key material. Remotely exploitable. CVE-2016-9963.
881 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
882 and 3.4.4 - once the server is enabled to respond to an OCSP request
883 it does even when not requested, resulting in a stapling non-aware
884 client dropping the TLS connection.
886 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
887 support variable-length bit vectors. No functional change.
889 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
890 The I= interface field on outgoing lines is now after the H= remote
891 host field, same as incoming lines. There is a separate
892 outgoing_interface log selector which allows you to disable the
895 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
896 If not running log_selector +smtp_connection the mainlog would be held
897 open indefinitely after a "too many connections" event, including to a
898 deleted file after a log rotate. Leave the per net connection logging
899 leaving it open for efficiency as that will be quickly detected by the
900 check on the next write.
902 HS/01 Bug 1671: Fix post transport crash.
903 Processing the wait-<transport> messages could crash the delivery
904 process if the message IDs didn't exist for some reason. When
905 using 'split_spool_directory=yes' the construction of the spool
906 file name failed already, exposing the same netto behaviour.
908 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
909 mime_regex ACL conditions.
911 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
912 to DSN fail messages (bounces): remote IP, remote greeting, remote response
913 to HELO, local diagnostic string.
915 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
916 log line to debug. Even when configured with a tls authenticator many
917 client connections are expected to not authenticate in this way, so
918 an authenticate fail is not an error.
920 HS/02 Add the Exim version string to the process info. This way exiwhat
921 gives some more detail about the running daemon.
923 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
924 matter for fast-change records such as DNSBLs.
926 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
927 retry record, even if constant. There may be multiple transports with
928 different interface settings and the retry behaviour needs to be kept
931 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
933 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
935 JH/10 Bug 840: fix log_defer_output option of pipe transport
937 JH/11 Bug 830: use same host for all RCPTS of a message, even under
938 hosts_randomize. This matters a lot when combined with mua_wrapper.
940 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
941 ${quote_pgsql:<string>} operator.
943 JH/13 Bug 1708: avoid misaligned access in cached lookup.
945 JH/14 Change header file name for freeradius-client. Relevant if compiling
946 with Radius support; from the Gentoo tree and checked under Fedora.
948 JH/15 Bug 1712: Introduce $prdr_requested flag variable
950 JH/16 Bug 1714: Permit an empty string as expansion result for transport
951 option transport_filter, meaning no filtering.
953 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
955 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
956 defaults to "*" (all hosts). The variable is now available when not built
957 with TLS, default unset, mainly to enable keeping the testsuite sane.
958 If a server certificate is not supplied (via tls_certificate) an error is
959 logged, and clients will find TLS connections fail on startup. Presumably
960 they will retry in-clear.
961 Packagers of Exim are strongly encouraged to create a server certificate
962 at installation time.
964 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
965 with the $config_file variable.
967 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
968 in transport context, after the attempt, and per-recipient. The latter type
969 is per host attempted. The event data is the error message, and the errno
970 information encodes the lookup type (A vs. MX) used for the (first) host,
971 and the trailing two digits of the smtp 4xx response.
973 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
974 to write to mainlog (or rejectlog, paniclog) in the window between file
975 creation and permissions/ownership being changed. Particularly affects
976 installations where exicyclog is run as root, rather than exim user;
977 result is that the running daemon panics and dies.
979 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
981 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
982 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
983 "pri" and "weight". Note that the previous implicit priority given by the
984 list order is no longer honoured.
986 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
989 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
990 by defining SUPPORT_SOCKS.
992 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
993 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
994 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
995 variables are renamed to proxy_{local,external}_{address,port}.
997 JH/27 Move Internationalisation support from Experimental to mainline, enabled
998 for a build by defining SUPPORT_I18N
1000 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1001 of the query string, and make ${quote_redis:} do that quoting.
1003 JH/29 Move Events support from Experimental to mainline, enabled by default
1004 and removable for a build by defining DISABLE_EVENT.
1006 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1008 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1009 cached by the daemon.
1011 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1012 by defining LOOKUP_REDIS. The libhiredis library is required.
1014 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1015 keys are given for lookup.
1017 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1018 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1019 only supported when built with TLS support. The PolarSSL SHA routines
1020 are still used when the TLS library is too old for convenient support.
1022 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1023 openssl_options), for security. OpenSSL forces this from version 1.1.0
1024 server-side so match that on older versions.
1026 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1027 allocation for $value could be released as the expansion processing
1028 concluded, but leaving the global pointer active for it.
1030 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1031 and to use the domains and local_parts ACL conditions.
1033 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1034 incorrectly not doubled on cutthrough transmission, hence seen as a
1035 body-termination at the receiving system - resulting in truncated mails.
1036 Commonly the sender saw a TCP-level error, and retransmitted the message
1037 via the normal store-and-forward channel. This could result in duplicates
1038 received - but deduplicating mailstores were liable to retain only the
1039 initial truncated version.
1041 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1043 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1045 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1046 we're in there, support oversigning also; bug 1309.
1048 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1050 HS/04 Add support for keep_environment and add_environment options.
1052 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1053 either intentional arithmetic overflow during PRNG, or testing config-
1056 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1057 delivery resulted in actual delivery. Cancel cutthrough before DATA
1060 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1061 rejects a recipient: pass the reject to the originator.
1063 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1064 Many were false-positives and ignorable, but it's worth fixing the
1067 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1068 for the new environment-manipulation done at startup. Move the routines
1069 from being local to tls.c to being global via the os.c file.
1071 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1072 an extract embedded as result-arg for a map, the first arg for extract
1073 is unavailable so we cannot tell if this is a numbered or keyed
1074 extraction. Accept either.
1080 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1083 JH/02 The smtp transport option "multi_domain" is now expanded.
1085 JH/03 The smtp transport now requests PRDR by default, if the server offers
1088 JH/04 Certificate name checking on server certificates, when exim is a client,
1089 is now done by default. The transport option tls_verify_cert_hostnames
1090 can be used to disable this per-host. The build option
1091 EXPERIMENTAL_CERTNAMES is withdrawn.
1093 JH/05 The value of the tls_verify_certificates smtp transport and main options
1094 default to the word "system" to access the system default CA bundle.
1095 For GnuTLS, only version 3.0.20 or later.
1097 JH/06 Verification of the server certificate for a TLS connection is now tried
1098 (but not required) by default. The verification status is now logged by
1099 default, for both outbound TLS and client-certificate supplying inbound
1102 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1105 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1106 Status Notification (bounce) messages are now MIME format per RFC 3464.
1107 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1108 under the control of the dsn_advertise_hosts option, and routers may
1109 have a dsn_lasthop option.
1111 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1112 default, modifiable by a malware= option. The list separator for
1113 the options can now be changed in the usual way. Bug 68.
1115 JH/10 The smtp_receive_timeout main option is now expanded before use.
1117 JH/11 The incoming_interface log option now also enables logging of the
1118 local interface on delivery outgoing connections.
1120 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1121 if the interface and destination host and port all match.
1123 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1126 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1127 Patch from Andrew Lewis.
1129 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1130 now supports optional time-restrictions, weighting, and priority
1131 modifiers per server. Patch originally by <rommer@active.by>.
1133 JH/16 The spamd_address main option now supports a mixed list of local
1134 and remote servers. Remote servers can be IPv6 addresses, and
1135 specify a port-range.
1137 JH/17 Bug 68: The spamd_address main option now supports an optional
1138 timeout value per server.
1140 JH/18 Bug 1581: Router and transport options headers_add/remove can
1141 now have the list separator specified.
1143 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1146 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1149 JH/21 Support for the A6 type of dns record is withdrawn.
1151 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1152 rather than the verbs used.
1154 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1155 from 255 to 1024 chars.
1157 JH/24 Verification callouts now attempt to use TLS by default.
1159 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1160 are generic router options now. The defaults didn't change.
1162 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1163 Original patch from Alexander Shikoff, worked over by JH.
1165 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1166 files if ZCAT_COMMAND is not executable.
1168 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1170 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1172 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1173 Normally benign, it bites when the pair was led to by a CNAME;
1174 modern usage is to not canonicalize the domain to a CNAME target
1175 (and we were inconsistent anyway for A-only vs AAAA+A).
1177 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1179 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1180 when evaluating $sender_host_dnssec.
1182 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1183 $sender_helo_dnssec variable.
1185 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1187 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1189 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1191 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1192 documented as working, but never had. Support all but $spam_report.
1194 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1195 added for tls authenticator.
1197 HS/03 Add perl_taintmode main config option
1203 TL/01 When running the test suite, the README says that variables such as
1204 no_msglog_check are global and can be placed anywhere in a specific
1205 test's script, however it was observed that placement needed to be near
1206 the beginning for it to behave that way. Changed the runtest perl
1207 script to read through the entire script once to detect and set these
1208 variables, reset to the beginning of the script, and then run through
1209 the script parsing/test process like normal.
1211 TL/02 The BSD's have an arc4random API. One of the functions to induce
1212 adding randomness was arc4random_stir(), but it has been removed in
1213 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1214 function when detected.
1216 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1217 cause callback expansion.
1219 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1220 syntax errors in an expansion can be treated as a string instead of
1221 logging or causing an error, due to the internal use of bool_lax
1222 instead of bool when processing it.
1224 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1225 server certificates when making smtp deliveries.
1227 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1229 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1231 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1233 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1234 Merged patch from Sebastian Wiedenroth.
1236 JH/05 Fix results-pipe from transport process. Several recipients, combined
1237 with certificate use, exposed issues where response data items split
1238 over buffer boundaries were not parsed properly. This eventually
1239 resulted in duplicates being sent. This issue only became common enough
1240 to notice due to the introduction of connection certificate information,
1241 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1243 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1244 size buffer was used, resulting in syntax errors when an expansion
1247 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1248 version 3.3.6 or later.
1250 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1251 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1252 both become "event_action", the variables become $event_name, $event_data
1253 and $event_defer_errno. There is a new variable $verify_mode, usable in
1254 routers, transports and related events. The tls:cert event is now also
1255 raised for inbound connections, if the main configuration event_action
1258 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1259 early OCSP support, but no stapling (appears to be less than 1.0.0).
1261 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1262 server certificate names available under the smtp transport option
1263 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1266 JH/10 Time-related extraction expansions from certificates now use the main
1267 option "timezone" setting for output formatting, and are consistent
1268 between OpenSSL and GnuTLS compilations. Bug 1541.
1270 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1271 encoded parameter in the incoming message. Bug 1558.
1273 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1274 include certificate info, eximon was claiming there were spoolfile
1277 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1279 JH/14 Log delivery-related information more consistently, using the sequence
1280 "H=<name> [<ip>]" wherever possible.
1282 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1283 are problematic for Debian distribution, omit them from the release
1286 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1288 JH/16 Fix string representation of time values on 64bit time_t architectures.
1291 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1292 output list separator was used.
1297 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1298 checkers that were complaining about end of non-void function with no
1301 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1302 This was a regression introduced in 4.83 by another bugfix.
1304 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1306 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1307 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1313 TF/01 Correctly close the server side of TLS when forking for delivery.
1315 When a message was received over SMTP with TLS, Exim failed to clear up
1316 the incoming connection properly after forking off the child process to
1317 deliver the message. In some situations the subsequent outgoing
1318 delivery connection happened to have the same fd number as the incoming
1319 connection previously had. Exim would try to use TLS and fail, logging
1320 a "Bad file descriptor" error.
1322 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1323 utilities have not been installed.
1325 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1326 temporary space as the ACL may create new global variables.
1328 TL/01 LDAP support uses per connection or global context settings, depending
1329 upon the detected version of the libraries at build time.
1331 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1332 to extract and use the src ip:port in logging and expansions as if it
1333 were a direct connection from the outside internet. PPv2 support was
1334 updated based on HAProxy spec change in May 2014.
1336 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1338 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1339 Properly escape header and check for NULL return.
1341 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1344 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1346 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1347 characters in header names, implemented as a verify condition.
1348 Contributed by Michael Fischer v. Mollard.
1350 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1351 results permerror and temperror. Previous values are deprecated but
1352 still accepted. In a future release, err_perm and err_temp will be
1353 completely removed, which will be a backward incompatibility if the
1354 ACL tests for either of these two old results. Patch contributed by
1355 user bes-internal on the mailing list.
1357 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1359 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1360 selectors, in both main and reject logs.
1362 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1365 JH/07 Add malware type "sock" for talking to simple daemon.
1367 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1369 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1370 routers/transports under cutthrough routing.
1372 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1373 numbers. Touch up "bool" conditional to keep the same definition.
1375 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1377 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1379 JH/12 Expand items in router/transport headers_add or headers_remove lists
1380 individually rather than the list as a whole. Bug 1452.
1382 Required for reasonable handling of multiple headers_ options when
1383 they may be empty; requires that headers_remove items with embedded
1384 colons must have them doubled (or the list-separator changed).
1386 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1387 view the policy declared in the DMARC record. Currently, $dmarc_status
1388 is a combined value of both the record presence and the result of the
1391 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1393 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1394 dnslookup router and the smtp transport (applying to the forward
1397 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1398 of ldap servers used for a specific lookup. Patch provided by Heiko
1401 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1402 New variable $lookup_dnssec_authenticated for observability.
1404 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1405 Patch submitted by Lars Timman.
1407 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1409 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1410 Requires trusted mode and valid format message id, aborts otherwise.
1411 Patch contributed by Heiko Schlichting.
1413 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1414 certextract with support for various fields. Bug 1358.
1416 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1417 is requested by default, modifiable by smtp transport option
1420 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1421 operate on certificate variables to give certificate fingerprints
1422 Also new ${sha256:cert_variable}.
1424 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1426 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1427 Christian Aistleitner.
1429 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1431 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1432 file. Patch from Wolfgang Breyha.
1434 JH/25 Expand the coverage of the delivery $host and $host_address to
1435 client authenticators run in verify callout. Bug 1476.
1437 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1438 align with daemon_smtp_ports. Bug 72.
1440 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1441 support and error reporting did not work properly.
1443 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1444 and is readable. Patch from Andrew Colin Kissa.
1446 TL/14 Enhance documentation of ${run expansion and how it parses the
1447 commandline after expansion, particularly in the case when an
1448 unquoted variable expansion results in an empty value.
1450 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1452 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1453 by renaming away from it. Addresses GH issue 15, reported by
1456 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1458 TL/15 SECURITY: prevent double expansion in math comparison functions
1459 (can expand unsanitized data). Not remotely exploitable.
1466 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1468 PP/02 Make -n do something, by making it not do something.
1469 When combined with -bP, the name of an option is not output.
1471 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1474 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1475 $sender_host_name and config options to manage this, and basic check
1478 PP/05 DSCP support for outbound connections and control modifier for inbound.
1480 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1481 (Only plugin which currently uses this is kerberos4, which nobody should
1482 be using, but we should make it available and other future plugins might
1483 conceivably use it, even though it would break NAT; stuff *should* be
1484 using channel bindings instead).
1486 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1487 name; added for Sendmail compatibility; requires admin caller.
1488 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1489 just ignore it); requires trusted caller.
1490 Also parse but ignore: -Ac -Am -X<logfile>
1493 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1495 TL/02 Add +smtp_confirmation as a default logging option.
1497 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1498 Patch by Magnus Holmgren from 2007-02-20.
1500 TL/04 Bugzilla 1281 - Spec typo.
1501 Bugzilla 1283 - Spec typo.
1502 Bugzilla 1290 - Spec grammar fixes.
1504 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1506 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1508 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1509 mailing list by by Dmitry Isaikin.
1511 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1513 JH/02 Support "G" suffix to numbers in ${if comparisons.
1515 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1517 NM/01 Bugzilla 1197 - Spec typo
1518 Bugzilla 1196 - Spec examples corrections
1520 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1522 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1523 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1526 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1527 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1529 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1530 "acl {{name}{arg}...}", and optional args on acl condition
1533 JH/05 Permit multiple router/transport headers_add/remove lines.
1535 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1537 JH/07 Avoid using a waiting database for a single-message-only transport.
1538 Performance patch from Paul Fisher. Bugzilla 1262.
1540 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1543 JH/09 Add $headers_added variable, with content from use of ACL modifier
1544 add_header (but not yet added to the message). Bugzilla 199.
1546 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1547 Pulled from Bugzilla 817 by Wolfgang Breyha.
1549 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1551 (nb: this is the same fix as in Exim 4.80.1)
1553 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1556 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1559 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1561 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1562 advertise SMTP AUTH mechanism to us, instead of a generic
1563 protocol violation error. Also, make Exim more robust to bad
1564 data from the Dovecot auth socket.
1566 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1568 When a queue runner is handling a message, Exim first routes the
1569 recipient addresses, during which it prunes them based on the retry
1570 hints database. After that it attempts to deliver the message to
1571 any remaining recipients. It then updates the hints database using
1574 So if a recipient address works intermittently, it can get repeatedly
1575 deferred at routing time. The retry hints record remains fresh so the
1576 address never reaches the final cutoff time.
1578 This is a fairly common occurrence when a user is bumping up against
1579 their storage quota. Exim had some logic in its local delivery code
1580 to deal with this. However it did not apply to per-recipient defers
1581 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1583 This change adds a proper retry rule check during routing so that the
1584 final cutoff time is checked against the message's age. We only do
1585 this check if there is an address retry record and there is not a
1586 domain retry record; this implies that previous attempts to handle
1587 the address had the retry_use_local_parts option turned on. We use
1588 this as an approximation for the destination being like a local
1589 delivery, as in LMTP.
1591 I suspect this new check makes the old local delivery cutoff check
1592 redundant, but I have not verified this so I left the code in place.
1594 TF/02 Correct gecos expansion when From: is a prefix of the username.
1596 Test 0254 submits a message to Exim with the header
1600 When I ran the test suite under the user fanf2, Exim expanded
1601 the header to contain my full name, whereas it should have added
1602 a Resent-Sender: header. It erroneously treats any prefix of the
1603 username as equal to the username.
1605 This change corrects that bug.
1607 GF/01 DCC debug and logging tidyup
1608 Error conditions log to paniclog rather than rejectlog.
1609 Debug lines prefixed by "DCC: " to remove any ambiguity.
1611 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1613 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1614 Bug spotted by Jeremy Harris; was flawed since initial commit.
1615 Would have resulted in OCSP responses post-SNI triggering an Exim
1616 NULL dereference and crash.
1618 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1620 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1621 Bug detection, analysis and fix by Samuel Thibault.
1622 Bugzilla 1331, Debian bug #698092.
1624 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1626 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1627 Server implementation by Todd Lyons, client by JH.
1628 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1629 config variable "prdr_enable" controls whether the server
1630 advertises the facility. If the client requests PRDR a new
1631 acl_data_smtp_prdr ACL is called once for each recipient, after
1632 the body content is received and before the acl_smtp_data ACL.
1633 The client is controlled by both of: a hosts_try_prdr option
1634 on the smtp transport, and the server advertisement.
1635 Default client logging of deliveries and rejections involving
1636 PRDR are flagged with the string "PRDR".
1638 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1639 fclose(). Diagnosis by Todd Lyons.
1641 PP/17 Update configure.default to handle IPv6 localhost better.
1642 Patch by Alain Williams (plus minor tweaks).
1645 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1646 This is now consistent with GnuTLS, and is now documented: the
1647 previous undocumented portable approach to treating the option as
1648 unset was to force an expansion failure. That still works, and
1649 an empty string is now equivalent.
1651 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1652 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1653 not performing validation itself.
1655 PP/20 Added force_command boolean option to pipe transport.
1656 Patch from Nick Koston, of cPanel Inc.
1658 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1661 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1663 PP/21 Fix eximon continuous updating with timestamped log-files.
1664 Broken in a format-string cleanup in 4.80, missed when I repaired the
1665 other false fix of the same issue.
1666 Report and fix from Heiko Schlichting.
1669 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1670 Report from Prashanth Katuri.
1672 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1673 It's SecureTransport, so affects any MacOS clients which use the
1674 system-integrated TLS libraries, including email clients.
1676 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1677 using a MIME ACL for non-SMTP local injection.
1678 Report and assistance in diagnosis by Warren Baker.
1680 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1682 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1684 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1685 last id that failed so it may be referenced in subsequent ACL's.
1687 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1690 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1691 ldap library initialization, allowing self-signed CA's to be
1692 used. Also properly sets require_cert option later in code by
1693 using NULL (global ldap config) instead of ldap handle (per
1694 session). Bug diagnosis and testing by alxgomz.
1696 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1697 the src/util/ subdirectory.
1699 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1700 renamed to Transport Post Delivery Action by Jeremy Harris, as
1703 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1704 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1705 redis_servers = needs to be configured which will be used by the redis
1706 lookup. Patch from Warren Baker, of The Packet Hub.
1708 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1710 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1711 hostname or reverse DNS when processing a host list. Used suggestions
1712 from multiple comments on this bug.
1714 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1716 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1717 Missed a few lines, added it to make the runtest require no keyboard
1720 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1721 contains upper case chars. Make router use caseful_local_part.
1723 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1724 support when GnuTLS has been built with p11-kit.
1730 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1732 This, or similar/improved, will also be change PP/11 of 4.82.
1738 PP/01 Handle short writes when writing local log-files.
1739 In practice, only affects FreeBSD (8 onwards).
1740 Bugzilla 1053, with thanks to Dmitry Isaikin.
1742 NM/01 Bugzilla 949 - Documentation tweak
1744 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1747 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1749 PP/02 Implemented gsasl authenticator.
1751 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1753 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1754 `pkg-config foo` for cflags/libs.
1756 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1757 with rest of GSASL and with heimdal_gssapi.
1759 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1760 `pkg-config foo` for cflags/libs for the TLS implementation.
1762 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1763 properties get this fed in as external SSF. A number of robustness
1764 and debugging improvements to the cyrus_sasl authenticator.
1766 PP/08 cyrus_sasl server now expands the server_realm option.
1768 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1769 Patch by Jeremy Harris.
1771 PP/10 Added dbmjz lookup type.
1773 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1775 PP/12 MAIL args handles TAB as well as SP, for better interop with
1776 non-compliant senders.
1777 Analysis and variant patch by Todd Lyons.
1779 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1780 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1781 Patch from Dirk Mueller <dmueller@suse.com>
1783 PP/13 tls_peerdn now print-escaped for spool files.
1784 Observed some $tls_peerdn in wild which contained \n, which resulted
1785 in spool file corruption.
1787 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1788 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1789 or write after TLS renegotiation, which otherwise led to messages
1792 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1793 as a tracking header (ie: a signed header comes before the signature).
1794 Patch from Wolfgang Breyha.
1796 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1797 comma-sep list; embedded commas doubled.
1799 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1801 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1803 Report and patch from Dmitry Banschikov.
1805 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1806 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1807 failures. We appear to now support TLS1.1+ with Exim.
1809 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1810 lets Exim select keys and certificates based upon TLS SNI from client.
1811 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1812 before an outbound SMTP session. New log_selector, +tls_sni.
1814 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1815 NULL dereference. Report and patch from Alun Jones.
1817 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1818 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1819 Not seeing resolver debug output on NetBSD, but suspect this is a
1820 resolver implementation change.
1822 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1823 Left warnings. Added "eximon gdb" invocation mode.
1825 PP/21 Defaulting "accept_8bitmime" to true, not false.
1827 PP/22 Added -bw for inetd wait mode support.
1829 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1830 locate the relevant includes and libraries. Made this the default.
1832 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1833 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1835 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1836 This may cause build issues on older platforms.
1838 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1839 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1840 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1841 Added SNI support via GnuTLS too.
1842 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1844 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1846 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1847 Applied second patch from Janne, implementing suggestion to default
1848 multiple-strings-in-record handling to match SPF spec.
1850 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1852 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1853 read-only, out of scope).
1854 Patch from Wolfgang Breyha, report from Stuart Northfield.
1856 PP/29 Fix three issues highlighted by clang analyser static analysis.
1857 Only crash-plausible issue would require the Cambridge-specific
1858 iplookup router and a misconfiguration.
1859 Report from Marcin Mirosław.
1861 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1863 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1864 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1865 As part of this, removing so much warning spew let me fix some minor
1866 real issues in debug logging.
1868 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1869 assignment on my part. Fixed.
1871 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1872 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1873 Janne Snabb (who went above and beyond: thank you).
1875 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1876 string otherwise requires a connection and a bunch more work and it's
1877 relatively easy to get wrong. Should also expose TLS library linkage
1880 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1881 64-bit ${eval} (JH/03).
1883 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1884 GNU libc to support some of the 64-bit stuff, should not lead to
1885 conflicts. Defined before os.h is pulled in, so if a given platform
1886 needs to override this, it can.
1888 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1889 protection layer was required, which is not implemented.
1890 Bugzilla 1254, patch from Wolfgang Breyha.
1892 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1893 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1894 tls_dhparam take prime identifiers. Also unbreak combination of
1895 OpenSSL+DH_params+TLSSNI.
1897 PP/39 Disable SSLv2 by default in OpenSSL support.
1903 PP/01 Solaris build fix for Oracle's LDAP libraries.
1904 Bugzilla 1109, patch from Stephen Usher.
1906 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1908 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1911 TF/02 Fix a couple more cases where we did not log the error message
1912 when unlink() failed. See also change 4.74-TF/03.
1914 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1915 lock up or crash if it happened to be inside a call to libc when it
1916 got a SIGUSR1 from exiwhat.
1918 The SIGUSR1 handler appends the current process status to the process
1919 log which is later printed by exiwhat. It used to use the general
1920 purpose logging code to do this, but several functions it calls are
1921 not safe for signals.
1923 The new output code in the SIGUSR1 handler is specific to the process
1924 log, and simple enough that it's easy to inspect for signal safety.
1925 Removing some special cases also simplifies the general logging code.
1926 Removing the spurious timestamps from the process log simplifies
1929 TF/04 Improved ratelimit ACL condition.
1931 The /noupdate option has been deprecated in favour of /readonly which
1932 has clearer semantics. The /leaky, /strict, and /readonly update modes
1933 are mutually exclusive. The update mode is no longer included in the
1934 database key; it just determines when the database is updated. (This
1935 means that when you upgrade Exim will forget old rate measurements.)
1937 Exim now checks that the per_* options are used with an update mode that
1938 makes sense for the current ACL. For example, when Exim is processing a
1939 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1940 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1941 must specify per_mail/readonly. If you omit the update mode it defaults to
1942 /leaky where that makes sense (as before) or /readonly where required.
1944 The /noupdate option is now undocumented but still supported for
1945 backwards compatibility. It is equivalent to /readonly except that in
1946 ACLs where /readonly is required you may specify /leaky/noupdate or
1947 /strict/noupdate which are treated the same as /readonly.
1949 A useful new feature is the /count= option. This is a generalization
1950 of the per_byte option, so that you can measure the throughput of other
1951 aggregate values. For example, the per_byte option is now equivalent
1952 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1954 The per_rcpt option has been generalized using the /count= mechanism
1955 (though it's more complicated than the per_byte equivalence). When it is
1956 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1957 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1958 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1959 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1960 Note that using per_rcpt with a non-readonly update mode in more than
1961 one ACL will cause the recipients to be double-counted. (The per_mail
1962 and per_byte options don't have this problem.)
1964 The handling of very low rates has changed slightly. If the computed rate
1965 is less than the event's count (usually one) then this event is the first
1966 after a long gap. In this case the rate is set to the same as this event's
1967 count, so that the first message of a spam run is counted properly.
1969 The major new feature is a mechanism for counting the rate of unique
1970 events. The new per_addr option counts the number of different
1971 recipients that someone has sent messages to in the last time period. It
1972 behaves like per_rcpt if all the recipient addresses are different, but
1973 duplicate recipient addresses do not increase the measured rate. Like
1974 the /count= option this is a general mechanism, so the per_addr option
1975 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1976 example, measure the rate that a client uses different sender addresses
1977 with the options per_mail/unique=$sender_address. There are further
1978 details in the main documentation.
1980 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1982 TF/06 Removed a few PCRE remnants.
1984 TF/07 Automatically extract Exim's version number from tags in the git
1985 repository when doing development or release builds.
1987 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1988 Bugzilla 879. Patch from Paul Fisher.
1990 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1991 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1994 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1996 PP/05 Variable $av_failed, true if the AV scanner deferred.
1997 Bugzilla 1078. Patch from John Horne.
1999 PP/06 Stop make process more reliably on build failure.
2000 Bugzilla 1087. Patch from Heiko Schlittermann.
2002 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2003 Bugzilla 1089. Patch from Heiko Schlittermann.
2005 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2006 Bugzilla 1131. Patch from Holger Weiß.
2008 PP/09 Handle IPv6 addresses with SPF.
2009 Bugzilla 860. Patch from Wolfgang Breyha.
2011 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2013 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2016 PP/11 match_* no longer expand right-hand-side by default.
2017 New compile-time build option, EXPAND_LISTMATCH_RHS.
2018 New expansion conditions, "inlist", "inlisti".
2020 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2022 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2024 PP/14 fix log_write() format string regression from TF/03.
2025 Bugzilla 1152. Patch from Dmitry Isaikin.
2031 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2033 PP/02 Harmonised TLS library version reporting; only show if debugging.
2034 Layout now matches that introduced for other libraries in 4.74 PP/03.
2036 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2038 PP/04 New "dns_use_edns0" global option.
2040 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2043 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2044 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2046 TK/01 Updated PolarSSL code to 0.14.2.
2047 Bugzilla 1097. Patch from Andreas Metzler.
2049 PP/07 Catch divide-by-zero in ${eval:...}.
2050 Fixes bugzilla 1102.
2052 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2055 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2056 format-string attack -- SECURITY: remote arbitrary code execution.
2058 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2059 time unintentionally subject to list matching rules, letting the header
2060 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2061 arbitrary string expansion). This allowed for information disclosure.
2063 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2064 INT_MIN/-1 -- value coerced to INT_MAX.
2070 NM/01 Workaround for PCRE version dependency in version reporting
2073 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2074 This fixes portability to compilers other than gcc, notably
2075 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2077 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2078 makefiles for portability to HP-UX and POSIX correctness.
2080 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2081 Also via indented variable definition in the Makefile.
2082 (Debugging by Oliver Heesakkers).
2084 PP/02 Restore caching of spamd results with expanded spamd_address.
2085 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2087 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2088 Improves build reliability. Fix from: Frank Elsner
2090 NM/02 Fix wide character breakage in the rfc2047 coding
2091 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2093 NM/03 Allow underscore in dnslist lookups
2094 Fixes bug 1026. Patch from Graeme Fowler
2096 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2097 Code patches from Adam Ciarcinski of NetBSD.
2099 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2102 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2103 is logged, to avoid truncation. Patch from John Horne.
2105 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2106 Patch from Jakob Hirsch.
2108 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2109 SQL string expansion failure details.
2110 Patch from Andrey Oktyabrski.
2112 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2113 Patch from Simon Arlott.
2115 PP/09 New lookups functionality failed to compile on old gcc which rejects
2116 extern declarations in function scope.
2117 Patch from Oliver Fleischmann
2119 PP/10 Use sig_atomic_t for flags set from signal handlers.
2120 Check getgroups() return and improve debugging.
2121 Fixed developed for diagnosis in bug 927 (which turned out to be
2124 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2125 Patch from Mark Zealey.
2127 PP/12 Bugzilla 1056: Improved spamd server selection.
2128 Patch from Mark Zealey.
2130 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2131 Based on patch from Heiko Schlittermann.
2133 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2134 Patch from Uwe Doering, sign-off by Michael Haardt.
2136 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2137 variable declaration deep within a block. Bug and patch from
2140 PP/15 lookups-Makefile IRIX compatibility coercion.
2142 PP/16 Make DISABLE_DKIM build knob functional.
2144 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2145 Patch by Simon Arlott
2147 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2148 variable argument macros. Our copy now differs from upstream.
2154 TF/01 Failure to get a lock on a hints database can have serious
2155 consequences so log it to the panic log.
2157 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2158 controlled using the smtp_confirmation log selector.
2160 TF/03 Include the error message when we fail to unlink a spool file.
2162 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2163 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2164 for maintaining out-of-tree patches for some time.
2166 PP/01 Bugzilla 139: Documentation and portability issues.
2167 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2168 Handle per-OS dynamic-module compilation flags.
2170 PP/02 Let /dev/null have normal permissions.
2171 The 4.73 fixes were a little too stringent and complained about the
2172 permissions on /dev/null. Exempt it from some checks.
2173 Reported by Andreas M. Kirchwitz.
2175 PP/03 Report version information for many libraries, including
2176 Exim version information for dynamically loaded libraries. Created
2177 version.h, now support a version extension string for distributors
2178 who patch heavily. Dynamic module ABI change.
2180 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2181 privilege escalation vulnerability whereby the Exim run-time user
2182 can cause root to append content of the attacker's choosing to
2185 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2188 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2189 If dropping privileges for untrusted macros, we disabled normal logging
2190 on the basis that it would fail; for the Exim run-time user, this is not
2191 the case, and it resulted in successful deliveries going unlogged.
2192 Fixed. Reported by Andreas Metzler.
2198 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2199 only prepend for the Resent-* case. Fixes regression introduced in
2200 Exim 4.70 by NM/22 for Bugzilla 607.
2202 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2203 increasing numbers of administrators be bitten by this.
2205 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2207 PP/03 Bugzilla 994: added openssl_options main configuration option.
2209 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2211 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2213 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2215 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2216 without a peer certificate, leading to a segfault because of an
2217 assumption that peers always have certificates. Be a little more
2218 paranoid. Problem reported by Martin Tscholak.
2220 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2221 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2222 NB: ClamAV planning to remove STREAM in "middle of 2010".
2223 CL also introduces -bmalware, various -d+acl logging additions and
2224 more caution in buffer sizes.
2226 PP/09 Implemented reverse_ip expansion operator.
2228 PP/10 Bugzilla 937: provide a "debug" ACL control.
2230 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2232 PP/12 Bugzilla 973: Implement --version.
2234 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2236 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2238 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2240 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2241 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2242 ignore trailing whitespace.
2244 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2246 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2249 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2250 Notification from Dr Andrew Aitchison.
2252 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2253 ExtendedDetectionInfo response format.
2254 Notification from John Horne.
2256 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2259 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2260 XSL and documented dependency on system catalogs, with examples of how
2263 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2266 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2267 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2268 configuration file which is writeable by the Exim user or group.
2270 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2271 of configuration files to cover files specified with the -C option if
2272 they are going to be used with root privileges, not just the default
2275 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2276 option (effectively making it always true).
2278 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2279 files to be used while preserving root privileges.
2281 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2282 that rogue child processes cannot use them.
2284 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2285 run-time user, instead of root.
2287 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2288 Exim run-time user without dropping privileges.
2290 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2291 result string, instead of calling string_vformat() twice with the same
2294 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2295 for other users. Others should always drop root privileges if they use
2296 -C on the command line, even for a whitelisted configure file.
2298 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2300 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2306 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2307 $data_path, and $header_path variables; fixed documentation bugs and
2310 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2311 exipick to access non-standard spools, including the "frozen" queue
2314 NM/01 Bugzilla 965: Support mysql stored procedures.
2315 Patch from Alain Williams
2317 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2319 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2320 Patch from Andreas Metzler
2322 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2323 Patch from Kirill Miazine
2325 NM/05 Bugzilla 671: Added umask to procmail example.
2327 JJ/03 installed exipick 20100323.0, fixing doc bug
2329 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2330 directory. Notification and patch from Dan Rosenberg.
2332 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2334 TK/02 Improve log output when DKIM signing operation fails.
2336 MH/01 Treat the transport option dkim_domain as a colon separated
2337 list, not as a single string, and sign the message with each element,
2338 omitting multiple occurences of the same signer.
2340 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2341 Bugzilla 985, 986. Patch by Simon Arlott
2343 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2344 Patch by Simon Arlott
2346 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2347 MBX locking. Notification from Dan Rosenberg.
2353 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2355 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2357 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2359 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2361 NM/04 Bugzilla 915: Flag broken perl installation during build.
2367 TK/01 Added patch by Johannes Berg that expands the main option
2368 "spamd_address" if it starts with a dollar sign.
2370 TK/02 Write list of recipients to X-Envelope-Sender header when building
2371 the mbox-format spool file for content scanning (suggested by Jakob
2374 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2375 (http://www.dcc-servers.net/) support via dccifd. Activated by
2376 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2378 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2379 by Mark Daniel Reidel <mr@df.eu>.
2381 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2382 When building exim an external PCRE library is now needed -
2383 PCRE is a system library on the majority of modern systems.
2384 See entry on PCRE_LIBS in EDITME file.
2386 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2387 conversation. Added nologin parameter to request.
2388 Patch contributed by Kirill Miazine.
2390 TF/01 Do not log submission mode rewrites if they do not change the address.
2392 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2394 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2395 log files in place. Contributed by Roberto Lima.
2397 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2399 TF/03 Bugzilla 615: When checking the local_parts router precondition
2400 after a local_part_suffix or local_part_prefix option, Exim now
2401 does not use the address's named list lookup cache, since this
2402 contains cached lookups for the whole local part.
2404 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2405 Robert Millan. Documentation is in experimental-spec.txt.
2407 TF/04 Bugzilla 668: Fix parallel build (make -j).
2409 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2411 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2412 Patch provided by Jan Srzednicki.
2414 TF/05 Leading white space used to be stripped from $spam_report which
2415 wrecked the formatting. Now it is preserved.
2417 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2418 that they are available at delivery time.
2420 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2422 TF/08 TLS error reporting now respects the incoming_interface and
2423 incoming_port log selectors.
2425 TF/09 Produce a more useful error message if an SMTP transport's hosts
2426 setting expands to an empty string.
2428 NM/06 Bugzilla 744: EXPN did not work under TLS.
2429 Patch provided by Phil Pennock.
2431 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2432 Patch provided by Richard Godbee.
2434 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2435 acl_smtp_notquit, added index entry.
2437 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2438 Patch provided by Eugene Bujak.
2440 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2441 accept(). Patch provided by Maxim Dounin.
2443 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2444 Patch provided by Phil Pennock.
2446 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2448 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2449 Patch provided by Brad "anomie" Jorsch.
2451 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2452 Patch provided by Dean Brooks.
2454 TK/05 Add native DKIM support (does not depend on external libraries).
2456 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2457 Patch provided by Graeme Fowler.
2459 NM/16 Bugzilla 851: Documentation example syntax fix.
2461 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2463 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2466 NM/19 Bugzilla 745: TLS version reporting.
2467 Patch provided by Phil Pennock.
2469 NM/20 Bugzilla 167: bool: condition support.
2470 Patch provided by Phil Pennock.
2472 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2473 clients. Patch provided by Phil Pennock.
2475 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2476 Patch provided by Brad "anomie" Jorsch.
2478 NM/23 Bugzilla 687: Fix misparses in eximstats.
2479 Patch provided by Heiko Schlittermann.
2481 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2482 Patch provided by Heiko Schlittermann.
2484 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2485 plus update to original patch.
2487 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2489 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2490 Patch provided by David Brownlee.
2492 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2494 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2496 NM/30 Bugzilla 888: TLS documentation bugfixes.
2498 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2500 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2501 Unlike the original bugzilla I have changed all shell scripts in src tree.
2503 NM/33 Bugzilla 898: Transport filter timeout fix.
2504 Patch by Todd Rinaldo.
2506 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2507 Patch by Serge Demonchaux.
2509 NM/35 Bugzilla 39: Base64 decode bug fixes.
2510 Patch by Jakob Hirsch.
2512 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2514 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2516 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2518 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2524 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2525 ALT-N's libdkim that I have put here:
2526 http://duncanthrax.net/exim-experimental/
2528 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2529 were called 'true' and it seems that C99 defines that as a reserved
2530 keyword to be used with 'bool' variable types. That means you could
2531 not include C99-style headers which use bools without triggering
2532 build errors in sieve.c.
2534 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2535 as mailq or other aliases. Changed the --help handling significantly
2536 to do whats expected. exim_usage() emits usage/help information.
2538 SC/01 Added the -bylocaldomain option to eximstats.
2540 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2542 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2544 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2550 PH/01 Another patch from the Sieve maintainer.
2552 PH/02 When an IPv6 address is converted to a string for single-key lookup
2553 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2554 dots are used instead of colons so that keys in lsearch files need not
2555 contain colons. This was done some time before quoting was made available
2556 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2557 (notated using the quote facility) so as to distinguish them from IPv4
2558 keys. This meant that lookups for IP addresses in host lists did not work
2559 for iplsearch lookups.
2561 This has been fixed by arranging for IPv6 addresses to be expressed with
2562 colons if the lookup type is iplsearch. This is not incompatible, because
2563 previously such lookups could never work.
2565 The situation is now rather anomalous, since one *can* have colons in
2566 ordinary lsearch keys. However, making the change in all cases is
2567 incompatible and would probably break a number of configurations.
2569 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2572 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2573 conversion specification without a maximum field width, thereby enabling
2574 a rogue spamd server to cause a buffer overflow. While nobody in their
2575 right mind would setup Exim to query an untrusted spamd server, an
2576 attacker that gains access to a server running spamd could potentially
2577 exploit this vulnerability to run arbitrary code as the Exim user.
2579 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2580 $primary_hostname instead of what libspf2 thinks the hosts name is.
2582 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2583 a directory entry by the name of the lookup key. Previously, if a
2584 symlink pointed to a non-existing file or a file in a directory that
2585 Exim lacked permissions to read, a lookup for a key matching that
2586 symlink would fail. Now it is enough that a matching directory entry
2587 exists, symlink or not. (Bugzilla 503.)
2589 PH/03 The body_linecount and body_zerocount variables are now exported in the
2592 PH/04 Added the $dnslist_matched variable.
2594 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2595 This means they are set thereafter only if the connection becomes
2598 PH/06 Added the client_condition to authenticators so that some can be skipped
2599 by clients under certain conditions.
2601 PH/07 The error message for a badly-placed control=no_multiline_responses left
2602 "_responses" off the end of the name.
2604 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2606 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2607 (without spaces) instead of just copying the configuration text.
2609 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2611 PH/11 Added $max_received_linelength.
2613 PH/12 Added +ignore_defer and +include_defer to host lists.
2615 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2616 way in which PCRE > 7.0 is built.
2618 PH/14 Implemented queue_only_load_latch.
2620 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2621 MAIL command. The effect was to mangle the value on 64-bit systems.
2623 PH/16 Another patch from the Sieve maintainer.
2625 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2627 PH/18 If a system quota error occurred while trying to create the file for
2628 a maildir delivery, the message "Mailbox is full" was not appended to the
2629 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2630 only to a quota excession during the actual writing of the file.
2632 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2633 characters?) which causes problems in log lines. The DN values are now
2634 passed through string_printing() before being added to log lines.
2636 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2637 and InterBase are left for another time.)
2639 PH/21 Added message_body_newlines option.
2641 PH/22 Guard against possible overflow in moan_check_errorcopy().
2643 PH/23 POSIX allows open() to be a macro; guard against that.
2645 PH/24 If the recipient of an error message contained an @ in the local part
2646 (suitably quoted, of course), incorrect values were put in $domain and
2647 $local_part during the evaluation of errors_copy.
2653 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2654 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2657 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2658 issue a MAIL command.
2660 PH/02 In an ACL statement such as
2662 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2664 if a client was not listed at all, or was listed with a value other than
2665 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2666 the condition was not true (as it should be), so access was not denied.
2667 The bug was that the ! inversion was incorrectly passed on to the second
2668 item. This has been fixed.
2670 PH/03 Added additional dnslists conditions == and =& which are different from
2671 = and & when the dns lookup returns more than one IP address.
2673 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2674 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2676 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2677 FSYNC, which compiles an option called disable_fsync that allows for
2678 bypassing fsync(). The documentation is heavily laced with warnings.
2680 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2682 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2683 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2684 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2685 including adding "make clean"; (3) Added -fPIC when compiling the test
2686 dynamically loaded module, to get rid of a warning.
2688 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2689 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2690 = 0s. The bug is otherwise harmless.
2692 PH/07 There was a bug in the dovecot authenticator such that the value of
2693 $auth1 could be overwritten, and so not correctly preserved, after a
2694 successful authentication. This usually meant that the value preserved by
2695 the server_setid option was incorrect.
2697 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2699 PH/09 Installed PCRE release 7.0.
2701 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2702 run for batched SMTP input. It is now run at the start of every message
2703 in the batch. While fixing this I discovered that the process information
2704 (output by running exiwhat) was not always getting set for -bs and -bS
2705 input. This is fixed, and it now also says "batched" for BSMTP.
2707 PH/11 Added control=no_pipelining.
2709 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2710 patch, slightly modified), and move the expansion of helo_data till after
2711 the connection is made in the smtp transport (so it can use these
2714 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2716 PH/14 Added log_selector = +pid.
2718 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2720 PH/16 Add ${if forany and ${if forall.
2722 PH/17 Added dsn_from option to vary the From: line in DSNs.
2724 PH/18 Flush SMTP output before performing a callout, unless control =
2725 no_callout_flush is set.
2727 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2728 was true (the default) a successful delivery failed to delete the retry
2729 item, thus causing premature timeout of the address. The bug is now
2732 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2734 PH/21 Long custom messages for fakedefer and fakereject are now split up
2735 into multiline responses in the same way that messages for "deny" and
2736 other ACL rejections are.
2738 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2739 with slight modification.
2741 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2742 draft, changing the syntax and factoring some duplicate code".
2744 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2745 for deliveries of the second and subsequent messages over the same SMTP
2748 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2749 ${reduce, with only minor "tidies".
2751 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2753 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2754 expansion side effects.
2756 PH/27 When a message times out after an over-quota error from an Exim-imposed
2757 quota, the bounce message says "mailbox is full". This message was not
2758 being given when it was a system quota that was exceeded. It now should
2761 MH/03 Made $recipients available in local_scan(). local_scan() already has
2762 better access to the recipient list through recipients_list[], but
2763 $recipients can be useful in postmaster-provided expansion strings.
2765 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2766 in the case of a MAIL command with additional options following the
2767 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2768 were accidentally chopped off.
2770 PH/29 SMTP synchronization checks are implemented when a command is read -
2771 there is a check that no more input is waiting when there shouldn't be
2772 any. However, for some commands, a delay in an ACL can mean that it is
2773 some time before the response is written. In this time, more input might
2774 arrive, invalidly. So now there are extra checks after an ACL has run for
2775 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2776 pipelining has not been advertised.
2778 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2780 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2781 returned in the SMTP response when smtp_return_error_details was false.
2782 This has been fixed.
2784 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2785 instead of the C I/O that was originally supplied, because problems were
2786 reported on Solaris.
2788 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2789 Exim which did not show up earlier: it was assuming that a call to
2790 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2791 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2792 was a macro that became an assignment, so it seemed to work. This has
2793 changed to a proper function call with a void return, hence the compile
2794 error. Exim's code has been fixed.
2796 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2799 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2801 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2803 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2804 because it turns out to be too big or there is a timeout) there is no
2805 "Completed" line in the log. When some messages of this type were
2806 selected by exigrep, they were listed as "not completed". Others were
2807 picked up by some special patterns. I have improved the selection
2808 criteria to be more general.
2810 PH/38 The host_find_failed option in the manualroute router can now be set
2811 to "ignore", to completely ignore a host whose IP address cannot be
2812 found. If all hosts are ignored, the behaviour is controlled by the new
2813 host_all_ignored option.
2815 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2816 homing or because of multiple MX records with /mx) generated more than
2817 one IP address, and the following item turned out to be the local host,
2818 all the secondary addresses of the first item were incorrectly removed
2819 from the list, along with the local host and any following hosts (which
2820 is what is supposed to happen).
2822 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2823 whoever called Exim into the -H file. In the case of the daemon it was
2824 behaving confusingly. When first started, it used values for whoever
2825 started the daemon, but after a SIGHUP it used the Exim user (because it
2826 calls itself on a restart). I have changed the code so that it now always
2829 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2830 message are rejected with the same error (e.g. no authentication or bad
2831 sender address), and a DATA command is nevertheless sent (as can happen
2832 with PIPELINING or a stupid MUA), the error message that was given to the
2833 RCPT commands is included in the rejection of the DATA command. This is
2834 intended to be helpful for MUAs that show only the final error to their
2837 PH/42 Another patch from the Sieve maintainer.
2839 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2840 Eximstats - Fixed some broken HTML links and added missing column headers
2842 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2843 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2845 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2847 PH/43 Yet another patch from the Sieve maintainer.
2849 PH/44 I found a way to check for a TCP/IP connection going away before sending
2850 the response to the final '.' that terminates a message, but only in the
2851 case where the client has not sent further data following the '.'
2852 (unfortunately, this is allowed). However, in many cases there won't be
2853 any further data because there won't be any more messages to send. A call
2854 to select() can be used: if it shows that the input is "ready", there is
2855 either input waiting, or the socket has been closed. An attempt to read
2856 the next input character can distinguish the two cases. Previously, Exim
2857 would have sent an OK response which the client would never have see.
2858 This could lead to message repetition. This fix should cure that, at
2859 least in a lot of common cases.
2861 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2862 advertised in response to EHLO.
2868 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2869 fixed by 4.65/MH/01 (is this a record?) are fixed:
2871 (i) An empty string was always treated as zero by the numeric comparison
2872 operators. This behaviour has been restored.
2874 (ii) It is documented that the numeric comparison operators always treat
2875 their arguments as decimal numbers. This was broken in that numbers
2876 starting with 0 were being interpreted as octal.
2878 While fixing these problems I realized that there was another issue that
2879 hadn't been noticed. Values of message_size_limit (both the global option
2880 and the transport option) were treated as octal if they started with 0.
2881 The documentation was vague. These values are now always treated as
2882 decimal, and I will make that clear in the documentation.
2888 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2889 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2892 MH/01 Don't check that the operands of numeric comparison operators are
2893 integers when their expansion is in "skipping" mode (fixes bug
2894 introduced by 4.64-PH/07).
2896 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2897 child addresses, Exim now panics and dies. Previously, because the count
2898 is held in a short int, deliveries were likely to be lost. As such a
2899 large number of recipients for a single message is ridiculous
2900 (performance will be very, very poor), I have chosen to impose a limit
2901 rather than extend the field.
2907 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2908 leftover -K file (the existence of which was triggered by #402).
2909 While we were at it, introduced process PID as part of the -K
2910 filename. This should rule out race conditions when creating
2913 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2914 processing considerably. Previous code took too long for large mails,
2915 triggering a timeout which in turn triggers #401.
2917 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2918 in the DK code in transports.c. sendfile() is not really portable,
2919 hence the _LINUX specificness.
2921 TF/01 In the add_headers option to the mail command in an Exim filter,
2922 there was a bug that Exim would claim a syntax error in any
2923 header after the first one which had an odd number of characters
2926 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2927 callout verification, Exim cached a "reject" for the entire domain. This
2928 is correct for most verifications, but it is not correct for a recipient
2929 verification with use_sender or use_postmaster set, because in that case
2930 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2931 case of MAIL FROM:<> rejection from other early rejections (e.g.
2932 rejection of HELO). When verifying a recipient using a non-null MAIL
2933 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2934 Whatever the result of the callout, the value of the domain cache is
2935 left unchanged (for any other kind of callout, getting as far as trying
2936 RCPT means that the domain itself is ok).
2938 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2941 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2942 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2943 session, the daemon ignores ECONNECT errors and logs others; it now
2944 ignores EPIPE as well.
2946 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2947 (quoted-printable decoding).
2949 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2950 later the small subsequent patch to fix an introduced bug.
2952 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2954 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2956 PH/08 An error is now given if message_size_limit is specified negative.
2958 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2959 to be given (somewhat) arbitrary names.
2961 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2964 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2965 miscellaneous code fixes
2967 PH/10 Added the log_reject_target ACL modifier to specify where to log
2970 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2971 hostname. This is wrong, because it relates to the incoming message (and
2972 probably the interface on which it is arriving) and not to the outgoing
2973 callout (which could be using a different interface). This has been
2974 changed to use the value of the helo_data option from the smtp transport
2975 instead - this is what is used when a message is actually being sent. If
2976 there is no remote transport (possible with a router that sets up host
2977 addresses), $smtp_active_hostname is used.
2979 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2980 tweaks were necessary in order to get it to work (see also 21 below):
2981 (a) The code assumed that strncpy() returns a negative number on buffer
2982 overflow, which isn't the case. Replaced with Exim's string_format()
2984 (b) There were several signed/unsigned issues. I just did the minimum
2985 hacking in of casts. There is scope for a larger refactoring.
2986 (c) The code used strcasecmp() which is not a standard C function.
2987 Replaced with Exim's strcmpic() function.
2988 (d) The code set only $1; it now sets $auth1 as well.
2989 (e) A simple test gave the error "authentication client didn't specify
2990 service in request". It would seem that Dovecot has changed its
2991 interface. Fortunately there's a specification; I followed it and
2992 changed what the client sends and it appears to be working now.
2994 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2997 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2998 address A is aliased to B and C, where B exists and C does not. Without
2999 -v the output is "A verified" because verification stops after a
3000 successful redirection if more than one address is generated. However,
3001 with -v the child addresses are also verified. Exim was outputting "A
3002 failed to verify" and then showing the successful verification for C,
3003 with its parentage. It now outputs "B failed to verify", showing B's
3004 parentage before showing the successful verification of C.
3006 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3007 look up a TXT record in a specific list after matching in a combined
3010 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3011 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3012 they consult the DNS. I had assumed they would set it the way they
3013 wanted; and indeed my experiments on Linux seem to show that in some
3014 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3015 To be on the safe side, however, I have now made the interface to
3016 host_find_byname() similar to host_find_bydns(), with an argument
3017 containing the DNS resolver options. The host_find_byname() function now
3018 sets these options at its start, just as host_find_bydns() does. The smtp
3019 transport options dns_qualify_single and dns_search_parents are passed to
3020 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3021 of host_find_byname() use the default settings of RES_DEFNAMES
3022 (qualify_single) but not RES_DNSRCH (search_parents).
3024 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3025 spool_read_header() do less string testing, by means of a preliminary
3026 switch on the second character of optional "-foo" lines. (This is
3027 overdue, caused by the large number of possibilities that now exist.
3028 Originally there were few.) While I was there, I also converted the
3029 str(n)cmp tests so they don't re-test the leading "-" and the first
3030 character, in the hope this might squeeze out yet more improvement.
3032 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3033 flag allowing group syntax was set by the header_syntax check but not
3034 turned off, possible causing trouble later; (2) The flag was not being
3035 set at all for the header_verify test, causing "group"-style headers to
3036 be rejected. I have now set it in this case, and also caused header_
3037 verify to ignore an empty address taken from a group. While doing this, I
3038 came across some other cases where the code for allowing group syntax
3039 while scanning a header line wasn't quite right (mostly, not resetting
3040 the flag correctly in the right place). These bugs could have caused
3041 trouble for malformed header lines. I hope it is now all correct.
3043 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3044 with the "reply" argument non-NULL. The code, however (which originally
3045 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3046 but it didn't always do it. This confused somebody who was copying the
3047 code for some other use. I have removed all the tests.
3049 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3050 feature that was used to support insecure browsers during the U.S. crypto
3051 embargo. It requires special client support, and Exim is probably the
3052 only MTA that supported it -- and would never use it because real RSA is
3053 always available. This code has been removed, because it had the bad
3054 effect of slowing Exim down by computing (never used) parameters for the
3055 RSA_EXPORT functionality.
3057 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3058 authenticator to fail if there's a tab character in the incoming data
3059 (there should never be unless someone is messing about, as it's supposed
3060 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3061 if the connection is using TLS or if the remote IP is the same as the
3062 local IP, and the "valid-client-cert option" if a client certificate has
3065 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3066 authenticators. This can be used for authorization after authentication
3067 succeeds. (In the case of plaintext, it servers for both authentication
3070 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3071 if any retry times were supplied.
3073 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3074 connection, where there is no client IP address to check. In this
3075 situation, the verify now always succeeds.
3077 PH/25 Applied John Jetmore's -Mset patch.
3079 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3081 PH/27 In a string expansion for a processed (not raw) header when multiple
3082 headers of the same name were present, leading whitespace was being
3083 removed from all of them, but trailing whitespace was being removed only
3084 from the last one. Now trailing whitespace is removed from each header
3085 before concatenation. Completely empty headers in a concatenation (as
3086 before) are ignored.
3088 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3089 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3091 PH/29 [Removed. This was a change that I later backed out, and forgot to
3092 correct the ChangeLog entry (that I had efficiently created) before
3093 committing the later change.]
3095 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3096 address errors (4xx response to RCPT) over the same connection as other
3097 messages routed to the same hosts. Such deliveries are always "forced",
3098 so retry times are not inspected. This resulted in far too many retries
3099 for the affected addresses. The effect occurred only when there were more
3100 hosts than the hosts_max_try setting in the smtp transport when it had
3101 the 4xx errors. Those hosts that it had tried were not added to the list
3102 of hosts for which the message was waiting, so if all were tried, there
3103 was no problem. Two fixes have been applied:
3105 (i) If there are any address or message errors in an SMTP delivery, none
3106 of the hosts (tried or untried) are now added to the list of hosts
3107 for which the message is waiting, so the message should not be a
3108 candidate for sending over the same connection that was used for a
3109 successful delivery of some other message. This seems entirely
3110 reasonable: after all the message is NOT "waiting for some host".
3111 This is so "obvious" that I'm not sure why it wasn't done
3112 previously. Hope I haven't missed anything, but it can't do any
3113 harm, as the worst effect is to miss an optimization.
3115 (ii) If, despite (i), such a delivery is accidentally attempted, the
3116 routing retry time is respected, so at least it doesn't keep
3117 hammering the server.
3119 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3120 in ${readsocket because some servers need this prod.
3122 PH/32 Added some extra debug output when updating a wait-xxx database.
3124 PH/33 The hint "could be header name not terminated by colon", which has been
3125 given for certain expansion errors for a long time, was not being given
3126 for the ${if def:h_colon_omitted{... case.
3128 PH/34 The spec says: "With one important exception, whenever a domain list is
3129 being scanned, $domain contains the subject domain." There was at least
3130 one case where this was not true.
3132 PH/35 The error "getsockname() failed: connection reset by peer" was being
3133 written to the panic log as well as the main log, but it isn't really
3134 panic-worthy as it just means the connection died rather early on. I have
3135 removed the panic log writing for the ECONNRESET error when getsockname()
3138 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3139 runs only) independently of the message's sender address. This meant
3140 that, if the 4xx error was in fact related to the sender, a different
3141 message to the same recipient with a different sender could confuse
3142 things. In particular, this can happen when sending to a greylisting
3143 server, but other circumstances could also provoke similar problems.
3144 I have changed the default so that the retry time for these errors is now
3145 based a combination of the sender and recipient addresses. This change
3146 can be overridden by setting address_retry_include_sender=false in the
3149 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3150 remote server are returned as part of bounce messages. This was not
3151 happening for LMTP over a pipe (the lmtp transport), but now it is the
3152 same for both kinds of LMTP.
3154 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3155 in header lines that were in fact CNAMEs. This is no longer the case.
3157 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3158 and queue runs started by the daemon processed all messages. This has
3159 been fixed so that -R and -S can now usefully be given with -q<time>.
3161 PH/40 Import PCRE release 6.7 (fixes some bugs).
3163 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3165 PH/42 Give an error if -q is specified more than once.
3167 PH/43 Renamed the variables $interface_address and $interface_port as
3168 $received_ip_address and $received_port, to make it clear that these
3169 values apply to message reception, and not to the outgoing interface when
3170 a message is delivered. (The old names remain recognized, of course.)
3172 PH/44 There was no timeout on the connect() call when using a Unix domain
3173 socket in the ${readsocket expansion. There now is.
3175 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3176 be meaningful with "accept".
3178 SC/01 Eximstats V1.43
3179 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3181 SC/02 Eximstats V1.44
3182 Use a glob alias rather than an array ref in the generated
3183 parser. This improves both readability and performance.
3185 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3186 Collect SpamAssassin and rejection statistics.
3187 Don't display local sender or destination tables unless
3188 there is data to show.
3189 Added average volumes into the top table text output.
3191 SC/04 Eximstats V1.46
3192 Collect data on the number of addresses (recipients)
3193 as well as the number of messages.
3195 SC/05 Eximstats V1.47
3196 Added 'Message too big' to the list of mail rejection
3197 reasons (thanks to Marco Gaiarin).
3199 SC/06 Eximstats V1.48
3200 Mainlog lines which have GMT offsets and are too short to
3201 have a flag are now skipped.
3203 SC/07 Eximstats V1.49 (Alain Williams)
3204 Added the -emptyok flag.
3206 SC/08 Eximstats V1.50
3207 Fixes for obtaining the IP address from reject messages.
3209 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3210 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3211 whitespace changes from 4.64-PH/27
3213 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3216 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3217 are found, allow negative numbers in numeric criteria)
3219 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3221 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3222 to match changes made in 4.64-PH/43
3224 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3226 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3227 "domain" is actually an IP address, and force a failure. This locks out
3228 those revolvers/nameservers that support "A-for-A" lookups, in
3229 contravention of the specifications.
3231 PH/48 When a host name was looked up from an IP address, and the subsequent
3232 forward lookup of the name timed out, the host name was left in
3233 $sender_host_name, contrary to the specification.
3235 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3236 restricted to single-key lookups, Exim was not diagnosing an error if
3237 * or *@ was used with a query-style lookup.
3239 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3241 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3242 long ago, but noone interested enough thought of it. Let's just say that
3243 the "1.1" means that there are some new functions that weren't there at
3244 some point in the past.
3246 PH/51 Error processing for expansion failure of helo_data from an smtp
3247 transport during callout processing was broken.
3249 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3250 tested/used via the -bh/-bhc/-bs options.
3252 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3253 bug, fixed in subsequent PCRE releases).
3255 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3256 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3258 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3264 SC/01 Use a glob alias rather than an array ref in eximstats generated
3265 parser. This improves both readability and performance.
3267 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3268 Don't display local sender or destination tables in eximstats unless
3269 there is data to show.
3270 Added average volumes into the eximstats top table text output.
3272 SC/03 Collect data on the number of addresses (recipients) as well
3273 as the number of messages in eximstats.
3275 TF/01 Correct an error in the documentation for the redirect router. Exim
3276 does (usually) call initgroups() when daemonizing.
3278 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3279 with consistent privilege compared to when running as a daemon.
3281 TF/03 Note in the spec that $authenticated_id is not set for local
3282 submissions from trusted users.
3284 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3285 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3287 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3288 by adding some example configuration directives to the default
3289 configuration file. A little bit of work is required to uncomment the
3290 directives and define how usernames and passwords are checked, but
3291 there is now a framework to start from.
3293 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3294 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3295 without this. I don't know how relevant this is to other LDAP libraries.
3297 PH/02 Add the verb name to the "unknown ACL verb" error.
3299 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3301 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3303 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3304 directory not expanded when it should be if an expanded home directory
3305 was set for the address (which is overridden by the transport).
3307 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3310 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3311 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3312 because it is too late at that time, and has no effect.
3314 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3315 security issue with \' (bugzilla #107). I could not use the
3316 PQescapeStringConn() function, because it needs a PGconn value as one of
3319 PH/08 When testing addresses using -bt, indicate those final addresses that
3320 are duplicates that would not cause an additional delivery. At least one
3321 person was confused, thinking that -bt output corresponded to deliveries.
3322 (Suppressing duplicates isn't a good idea as you lose the information
3323 about possibly different redirections that led to the duplicates.)
3325 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3326 systems where poll() doesn't work, in particular OS X.
3328 PH/10 Added more information to debugging output for retry time not reached.
3330 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3331 operations in malware.c.
3333 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3336 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3337 syslog_duplication set false, log lines that would normally be written
3338 both the the main log and to the reject log were not written to syslog at
3341 PH/14 In the default configuration, change the use of "message" in ACL warn
3342 statements to "add_header".
3344 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3345 not followed by a command (e.g. "seen endif").
3347 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3348 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3351 PH/17 Added extra conditions to the default value of delay_warning_condition
3355 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3356 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3357 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3360 The Auto-Submitted: and various List- headers are standardised, whereas I
3361 don't think Precedence: ever was.
3363 PH/18 Refactored debugging code in route_finduser() to show more information,
3364 in particular, the error code if getpwnam() issues one.
3366 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3367 This is apparently needed in addition to the PH/07 change above to avoid
3368 any possible encoding problems.
3370 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3371 but not after initializing Perl.
3373 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3374 output them only if debugging. By default they are written stderr,
3375 apparently, which is not desirable.
3377 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3380 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3383 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3385 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3386 authenticated or an ident call has been made. Suppress the default
3387 values for $authenticated_id and $authenticated_sender (but permit -oMai
3388 and -oMas) when testing with -bh.
3390 PH/24 Re-jigged the order of the tests in the default configuration so that the
3391 tests for valid domains and recipients precede the DNS black list and CSA
3392 tests, on the grounds that those ones are more expensive.
3394 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3395 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3396 This bug exists in every version of Exim that I still have, right back to
3399 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3400 However, an attempt to turn on case-sensitivity in a regex key by
3401 including (?-i) didn't work because the subject string was already
3402 lowercased, and the effects were non-intuitive. It turns out that a
3403 one-line patch can be used to allow (?-i) to work as expected.
3409 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3410 other effects) broke the use of negated acl sub-conditions.
3412 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3415 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3416 "Deny" causes Exim to reject the incoming connection with a 554 error.
3417 Unfortunately, if there is a major crisis, such as a disk failure,
3418 tcp-wrappers gives "deny", whereas what one would like would be some
3419 kind of temporary error. A kludge has been added to help with this.
3420 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3421 554 error is used if errno is still zero or contains ENOENT (which occurs
3422 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3425 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3427 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3428 errors. Otherwise a message that provokes a temporary error (when other
3429 messages do not) can cause a whole host to time out.
3431 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3432 addresses were routed directly to files or pipes from a redirect router.
3433 File deliveries just didn't batch; pipe deliveries might have suffered
3436 PH/06 A failure to get a lock for a hints database would erroneously always say
3437 "Failed to get write lock", even when it was really a read lock.
3439 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3440 of 0600. This has been changed to use the value of the lockfile_mode
3441 option (which defaults to 0600).
3443 PH/08 Applied small patch from the Sieve maintainer.
3445 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3446 folder from quota calculations, a direct delivery into this folder messed
3447 up the contents of the maildirsize file. This was because the regex was
3448 used only to exclude .Trash (or whatever) when the size of the mailbox
3449 was calculated. There was no check that a delivery was happening into an
3450 excluded directory. This bug has been fixed by ignoring all quota
3451 processing for deliveries into excluded directories.
3453 PH/10 Added the maildirfolder_create_regex option to appendfile.
3459 PH/01 The code for finding all the local interface addresses on a FreeBSD
3460 system running IPv6 was broken. This may well have applied to all BSD
3461 systems, as well as to others that have similar system calls. The broken
3462 code found IPv4 interfaces correctly, but gave incorrect values for the
3463 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3464 that it would not match correctly against @[] and not recognize the IPv6
3467 PH/02 The ipliteral router was not recognizing addresses of the form user@
3468 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3470 PH/03 Added disable_ipv6.
3472 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3473 decoded form, because it is most often used to construct To: headers
3474 lines in autoreplies, and the decoded form may well be syntactically
3475 invalid. However, $reply_address has leading white space removed, and all
3476 newlines turned into spaces so that the autoreply transport does not
3479 PH/05 If group was specified without a user on a router, and no group or user
3480 was specified on a transport, the group from the router was ignored.
3482 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3483 for visible compile-time settings that can be used to change these
3484 numbers, for those that want even more. Backwards compatibility with old
3485 spool files has been maintained. However, going back to a previous Exim
3486 release will lost any variables that are in spool files.
3488 PH/07 Two small changes when running in the test harness: increase delay when
3489 passing a TCP/IP connection to a new process, in case the original
3490 process has to generate a bounce, and remove special handling of
3491 127.0.0.2 (sic), which is no longer necessary.
3493 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3494 be the same on different OS.
3496 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3499 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3500 whether --show-vars was specified or not
3502 JJ/02 exipick: Added support for new ACL variable spool format introduced
3505 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3506 syntactically invalid From: or Reply-to: line, and a filter used this to
3507 generate an autoreply, and therefore failed to obtain an address for the
3508 autoreply, Exim could try to deliver to a non-existent relative file
3509 name, causing unrelated and misleading errors. What now happens is that
3510 it logs this as a hard delivery error, but does not attempt to create a
3513 PH/11 The exinext utility has a -C option for testing purposes, but although
3514 the given file was scanned by exinext itself; it wasn't being passed on
3515 when Exim was called.
3517 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3518 an end-of-file indication when reading a command response.
3520 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3521 compiled. In many other places in Exim, IPv6 addresses are always
3522 recognized, so I have changed this. It also means that IPv4 domain
3523 literals of the form [IPV4:n.n.n.n] are now always recognized.
3525 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3526 used if the router is not running as root, for example, when verifying at
3527 ACL time, or when using -bh. The debugging output from this situation was
3528 non-existent - all you got was a failure to exec. I have made two
3531 (a) Failures to set uid/gid, the current directory, or a process leader
3532 in a subprocess such as that created by queryprogram now generate
3533 suitable debugging output when -d is set.
3535 (b) The queryprogram router detects when it is not running as root,
3536 outputs suitable debugging information if -d is set, and then runs
3537 the subprocess without attempting to change uid/gid.
3539 PH/15 Minor change to Makefile for building test_host (undocumented testing
3542 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3543 additional section of a DNS packet that returns MX or SRV records.
3544 Instead, it always explicitly searches for A/AAAA records. This avoids
3545 major problems that occur when a DNS server includes only records of one
3546 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3547 fixed another bug: if SRV records were looked up and the corresponding
3548 address records were *not* found in the additional section, the port
3549 values from the SRV records were lost.
3551 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3552 using the correct key (the original address) when searching the retry
3553 rules in order to find which one to use for generating the retry hint.
3555 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3556 adding the default one. Similarly, if it contains a Reply-To: header, the
3557 errors_reply_to option, if set, is not used.
3559 PH/19 When calculating a retry time, Exim used to measure the "time since
3560 failure" by looking at the "first failed" field in the retry record. Now
3561 it does not use this if it is later than than the arrival time of the
3562 message. Instead it uses the arrival time. This makes for better
3563 behaviour in cases where some deliveries succeed, thus re-setting the
3564 "first failed" field. An example is a quota failure for a huge message
3565 when small messages continue to be delivered. Without this change, the
3566 "time since failure" will always be short, possible causing more frequent
3567 delivery attempts for the huge message than are intended.
3568 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3570 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3571 $1, $2, $3) because the numerical variables can be reset during some
3572 expansion items (e.g. "match"), thereby losing the authentication data.
3574 PH/21 Make -bV show the size of off_t variables so that the test suite can
3575 decide whether to run tests for quotas > 2G.
3577 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3578 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3579 filecount value is greater than 2G or if a quota value is greater than 2G
3580 on a system where the size of off_t is not greater than 4, a panic error
3583 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3584 never match. The debug and -bh output now contains an explicit error
3585 message indicating a malformed IPv4 address or mask.
3587 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3588 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3589 PH/23 above applies.
3591 PH/25 Do not write to syslog when running in the test harness. The only
3592 occasion when this arises is a failure to open the main or panic logs
3593 (for which there is an explicit test).
3595 PH/26 Added the /no_tell option to "control=freeze".
3597 PH/27 If a host name lookup failed very early in a connection, for example, if
3598 the IP address matched host_lookup and the reverse lookup yielded a name
3599 that did not have a forward lookup, an error message of the form "no IP
3600 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3601 could be logged. Now it outputs the IP address instead of "NULL".
3603 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3604 allows the sender and the authenticated sender to be set when
3605 submitting a message from within Exim. Since child_open_exim() is
3606 documented for local_scan(), the new function should be too.
3608 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3609 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3610 results in an empty string is now treated as unset.
3612 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3614 PH/31 Added sender_verify_fail logging option.
3616 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3617 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3618 tidied the source and removed it altogether.
3620 PH/33 When a queue run was abandoned because the load average was too high, a
3621 log line was always written; now it is written only if the queue_run log
3622 selector is set. In addition, the log line for abandonment now contains
3623 information about the queue run such as the pid. This is always present
3624 in "start" and "stop" lines but was omitted from the "abandon" line.
3626 PH/34 Omit spaces between a header name and the colon in the error message that
3627 is given when verify = headers_syntax fails (if there are lots of them,
3628 the message gets confusing).
3630 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3631 names, as there are now some PTR records that contain slashes. This check
3632 is only to protect against broken name servers that fall over on strange
3633 characters, so the fact that it applies to all lookups doesn't matter.
3635 PH/36 Now that the new test suite is complete, we can remove some of the
3636 special code in Exim that was needed for the old test suite. For example,
3637 sorting DNS records because real resolvers return them in an arbitrary
3638 order. The new test suite's fake resolver always returns records in the
3641 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3642 bounces) except when queue_only is set, to avoid logging races between
3643 the different processes.
3645 PH/38 Panic-die if .include specifies a non-absolute path.
3647 PH/39 A tweak to the "H" retry rule from its user.
3649 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3650 a label. They prevented compilation on older perls.
3652 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3653 a warning to be raised on newish perls.
3655 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3656 on queue. Changes to match documented behaviour of showing count of
3657 messages matching specified criteria.
3659 PH/40 Changed the default ident timeout from 30s to 5s.
3661 PH/41 Added support for the use of login_cap features, on those BSD systems
3662 that have them, for controlling the resources used by pipe deliveries.
3664 PH/42 The content-scanning code uses fopen() to create files in which to put
3665 message data. Previously it was not paying any attention to the mode of
3666 the files. Exim runs with umask(0) because the rest of the code creates
3667 files with open(), and sets the required mode explicitly. Thus, these
3668 files were ending up world-writeable. This was not a big issue, because,
3669 being within the spool directory, they were not world-accessible. I have
3670 created a function called modefopen, which takes an additional mode
3671 argument. It sets umask(777), creates the file, chmods it to the required
3672 mode, then resets the umask. All the relevant calls to fopen() in the
3673 content scanning code have been changed to use this function.
3675 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3676 to 24 hours. This avoids potential overflow problems when processing G
3677 and H retry rules. I suspect nobody ever tinkers with this value.
3679 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3681 PH/45 When the plaintext authenticator is running as a client, the server's
3682 challenges are checked to ensure they are valid base64 strings. By
3683 default, the authentication attempt is cancelled if an invalid string is
3684 received. Setting client_ignore_invalid_base64 true ignores these errors.
3685 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3686 they are received. Thus, the responses can be made to depend on the
3687 challenges. If an invalid string is ignored, an empty string is placed in
3690 PH/46 Messages that are created by the autoreply transport now contains a
3691 References: header, in accordance with RFCs 2822 and 3834.
3693 PH/47 Added authenticated_sender_force to the smtp transport.
3695 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3697 PH/49 Installed latest patch from the Sieve maintainer.
3699 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3700 was also set, the appendfile transport was unnecessarily scanning a
3701 directory of message files (e.g. for maildir delivery) to find the count
3702 of files (along with the size), even though it did not need this
3703 information. It now does the scan only if it needs to find either the
3704 size of the count of files.
3706 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3708 PH/52 Two bugs concerned with error handling when the smtp transport is
3711 (i) Exim was not creating retry information for temporary errors given
3712 for individual recipients after the DATA command when the smtp transport
3713 was used in LMTP mode. This meant that they could be retried too
3714 frequently, and not timed out correctly.
3716 (ii) Exim was setting the flag that allows error details to be returned
3717 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3718 recipients that were returned after the DATA command.
3720 PH/53 This is related to PH/52, but is more general: for any failing address,
3721 when detailed error information was permitted to be returned to the
3722 sender, but the error was temporary, then after the final timeout, only
3723 "retry timeout exceeded" was returned. Now it returns the full error as
3724 well as "retry timeout exceeded".
3726 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3727 do this, and (what is worse) MTAs that accept it.
3729 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3730 will now be deprecated.
3732 PH/56 New os.c-cygwin from the Cygwin maintainer.
3734 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3735 formats (previously only available in exim formats via -bpr, -bpru,
3736 and -bpra. Now also available in native and exiqgrep formats)
3738 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3739 with very large, slow to parse queues
3741 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3743 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3745 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3746 responses to authentication challenges, though it was showing the
3747 challenges; (ii) I've removed the CR characters from the debug output for
3750 PH/58 Allow for the insertion of a newline as well as a space when a string
3751 is turned into more than one encoded-word during RFC 2047 encoding. The
3752 Sieve code now uses this.
3754 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3755 data_4xx, lost_connection, tls_required.
3757 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3758 message was being sent as an SMTP response.
3760 PH/61 Add -l and -k options to exicyclog.
3762 PH/62 When verifying, if an address was redirected to one new address, so that
3763 verification continued, and the new address failed or deferred after
3764 having set something in $address_data, the value of $address_data was not
3765 passed back to the ACL. This was different to the case when no
3766 redirection occurred. The value is now passed back in both cases.
3768 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3769 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3770 use login_cap.h, so on its own it isn't the distinguishing feature. The
3771 new name refers directly to the setclassresources() function.
3773 PH/65 Added configuration files for NetBSD3.
3775 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3777 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3778 is preferred over IPv4.
3780 PH/68 The bounce_return_message and bounce_return_body options were not being
3781 honoured for bounces generated during the reception of non-SMTP messages.
3782 In particular, this applied to messages rejected by the ACL. This bug has
3783 been fixed. However, if bounce_return_message is true and bounce_return_
3784 body is false, the headers that are returned for a non-SMTP message
3785 include only those that have been read before the error was detected.
3786 (In the case of an ACL rejection, they have all been read.)
3788 PH/69 The HTML version of the specification is now built in a directory called
3789 spec_html instead of spec.html, because the latter looks like a path with
3790 a MIME-type, and this confuses some software.
3792 PH/70 Catch two compiler warnings in sieve.c.
3794 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3795 function verify_get_ident() calls ip_connect() to connect a socket, but
3796 if the "connect()" function timed out, ip_connect() used to close the
3797 socket. However, verify_get_ident() also closes the socket later, and in
3798 between Exim writes to the log, which may get opened at this point. When
3799 the socket was closed in ip_connect(), the log could get the same file
3800 descriptor number as the socket. This naturally causes chaos. The fix is
3801 not to close the socket in ip_connect(); the socket should be closed by
3802 the function that creates it. There was only one place in the code where
3803 this was missing, in the iplookup router, which I don't think anybody now
3804 uses, but I've fixed it anyway.
3806 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3807 well as to direct DNS lookups. Otherwise the handling of names in host
3808 lists is inconsistent and therefore confusing.
3814 PH/01 Two changes to the default runtime configuration:
3816 (1) Move the checks for relay_from_hosts and authenticated clients from
3817 after to before the (commented out) DNS black list checks.
3819 (2) Add control=submission to the relay_from_hosts and authenticated
3820 clients checks, on the grounds that messages accepted by these
3821 statements are most likely to be submissions.
3823 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3825 (1) Generate an error if the third argument for the ${prvs expansion is
3828 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3831 (3) Reset the variables that are obtained from the first argument of
3832 ${prvscheck and used in the second argument before leaving the code,
3833 because their memory is reclaimed, so using them afterwards may do
3836 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3837 one (it's much easier than Tom thought :-).
3839 (5) Because of (4), we can now allow for the use of $prvscheck_result
3840 inside the third argument.
3842 PH/03 For some reason, the default setting of PATH when running a command from
3843 a pipe transport was just "/usr/bin". I have changed it to
3846 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3847 anything to be listed in the output from -bV.
3849 PH/05 When a filter generated an autoreply, the entire To: header line was
3850 quoted in the delivery log line, like this:
3852 => >A.N.Other <ano@some.domain> <original@ddress> ...
3854 This has been changed so that it extracts the operative address. There
3855 may be more than one such address. If so, they are comma-separated, like
3858 => >ano@some.domain,ona@other.domain <original@ddress> ...
3860 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3861 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3862 not being looked up in the rDNS to get a host name, Exim was showing the
3863 IP address twice in Received: lines, even though the IP addresses were
3864 identical. For example:
3866 Received: from [1.2.3.4] (helo=[1.2.3.4])
3868 However, if the real host name was known, it was omitting the HELO data
3869 if it matched the actual IP address. This has been tidied up so that it
3870 doesn't show the same IP address twice.
3872 PH/07 When both +timestamp and +memory debugging was on, the value given by
3873 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3874 called by the debug printing, thereby overwriting the timestamp buffer.
3875 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3877 PH/08 When the original message was included in an autoreply transport, it
3878 always said "this is a copy of the message, including all the headers",
3879 even if body_only or headers_only was set. It now gives an appropriate
3882 PH/09 Applied a patch from the Sieve maintainer which:
3884 o fixes some comments
3885 o adds the (disabled) notify extension core
3886 o adds some debug output for the result of if/elsif tests
3887 o points to the current vacation draft in the documentation
3888 and documents the missing references header update
3892 o fixes a bug in processing the envelope test (when testing
3893 multiple envelope elements, the last element determined the
3896 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3897 Electronic Mail") by including:
3899 Auto-submitted: auto-generated
3901 in the messages that it generates (bounce messages and others, such as
3902 warnings). In the case of bounce messages for non-SMTP messages, there was
3903 also a typo: it was using "Auto_submitted" (underscore instead of
3904 hyphen). Since every message generated by Exim is necessarily in response
3905 to another message, thes have all been changed to:
3907 Auto-Submitted: auto-replied
3909 in accordance with these statements in the RFC:
3911 The auto-replied keyword:
3913 - SHOULD be used on messages sent in direct response to another
3914 message by an automatic process,
3916 - MUST NOT be used on manually-generated messages,
3918 - MAY be used on Delivery Status Notifications (DSNs) and Message
3919 Disposition Notifications (MDNs),
3921 - MUST NOT be used on messages generated by automatic or periodic
3922 processes, except for messages which are automatic responses to
3925 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3926 to the default Received: header definition.
3928 PH/12 Added log selector acl_warn_skipped (default on).
3930 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3931 variables because (a) they are in the wrong storage pool and (b) even if
3932 they were copied, it wouldn't work properly because of the caching.
3934 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3935 checking when decoding. Apparently there are clients that generate
3936 overlong encoded strings. Why am I not surprised?
3938 PH/15 If the first argument of "${if match_address" was not empty, but did not
3939 contain an "@" character, Exim crashed. Now it writes a panic log message
3940 and treats the condition as false.
3942 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3944 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3945 extension "envelope-auth". The code is finished and in agreement with
3946 other implementations, but there is no documentation so far and in fact,
3947 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3948 not changing the active code.
3950 Print executed "if" and "elsif" statements when debugging is used. This
3951 helps a great deal to understand what a filter does.
3953 Document more things not specified clearly in RFC3028. I had all this
3954 sorted out, when out of a sudden new issues came to my mind. Oops."
3956 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3959 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3960 canonical form (as documented). However, after a host name lookup from
3961 the IP address, check_host() was doing a simple string comparison with
3962 addresses acquired from the DNS when checking that the found name did
3963 have the original IP as one of its addresses. Since any found IPv6
3964 addresses are likely to be in abbreviated form, the comparison could
3965 fail. Luckily, there already exists a function for doing the comparison
3966 by converting both addresses to binary, so now that is used instead of
3967 the text comparison.
3969 PH/20 There was another similar case to PH/19, when a complete host name was
3970 given in a host list; looking up its IP address could give an abbreviated
3971 form, whereas the current host's name might or might not be abbreviated.
3972 The same fix has been applied.
3978 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3979 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3982 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3983 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3985 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3987 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3988 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3989 Also, for "Auto-Submitted", treat anything other than "no" as
3990 non-personal, in accordance with RFC 3834. (Previously it treated
3991 anything starting "auto-" as non-personal.)
3993 TF/01 The control=submission/name=... option had a problem with syntax
3994 errors if the name included a slash character. The /name= option
3995 now slurps the rest of the string, so it can include any characters
3996 but it must come last in the list of options (after /sender_retain
3999 PH/05 Some modifications to the interface to the fake nameserver for the new
4007 TK/01 Added the "success_on_redirect" address verification option. See
4008 NewStuff for rationale and an example.
4010 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4012 PH/02 Patch to exigrep to allow it to work on syslog lines.
4014 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4015 fread() to skip over the body file's header line, because in Cygwin the
4016 header line is locked and is inaccessible.
4018 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4019 co-exist for some time) to make it clear that it is the Exim ID that is
4020 referenced, not the Message-ID: header line.
4022 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4023 string_format() function, because snprintf() does not exist on all
4026 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4027 ${stat: expansion item.
4029 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4030 protocol synchronization error", to keep the pedants happy.
4032 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4033 well as for IRIX systems, when gcc is being used. See the host.c source
4036 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4038 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4041 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4042 transport and to the smtp transport in LMTP mode.
4044 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4046 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4047 run a filter in a subprocess. This could lead to confusion in subsequent
4048 lookups in the parent process. There should also be a search_tidyup() at
4049 the end of the subprocess.
4051 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4052 only if the host matched helo_try_verify_hosts, which caused the
4053 verification to occur when the EHLO/HELO command was issued. The ACL just
4054 tested the remembered result. Now, if a previous verification attempt has
4055 not happened, "verify = helo" does it there and then.
4057 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4059 TK/03 Fix log output including CR from clamd.
4061 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4062 exist provoked a memory error which could cause a segfault.
4064 PH/15 Installed PCRE 6.2
4066 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4068 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4069 of the problem. Specifically, suggested +O2 rather than +O1 for the
4072 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4074 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4075 record, the port was not correctly logged when the outgoing_port log
4076 selector was set (it logged the transort's default port).
4078 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4079 fallback_hosts, and "hosts" in the smtp transport.
4081 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4082 host errors such as "Connection refused".
4084 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4085 authentication with radiusclient 0.4.9:
4087 - Error returned from rc_read_config was caught wrongly
4088 - Username/password not passed on to radius server due to wrong length.
4090 The presumption is that some radiusclient API changes for 4.51/PH/17
4091 were not taken care of correctly. The code is still untested by me (my
4092 Linux distribution still has 0.3.2 of radiusclient), but it was
4093 contributed by a Radius user.
4095 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4096 expanding the "port" option of the smtp transport.
4098 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4099 while reading a MIME header. Thanks to Tom Hughes for a patch.
4101 PH/24 Include config.h inside local_scan.h so that configuration settings are
4104 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4105 that in an ACL for RCPT (for example), you can examine exactly what was
4108 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4109 commands, but it was not correctly comparing the address with the actual
4110 client host address. Thus, it would show the EHLO address in Received:
4111 header lines when this was not necessary.
4113 PH/27 Added the % operator to ${eval:}.
4115 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4116 it should be ignoring failures (because with -C, for example, it has lost
4117 privilege). It wasn't ignoring creation failures other than "already
4120 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4123 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4124 because an input line was too long, either on its own, or by virtue of
4125 too many continuations, the temporary file was not being removed, and the
4126 return code was incorrect.
4128 PH/31 Missing "BOOL" in function definition in filtertest.c.
4130 PH/32 Applied Sieve patches from the maintainer.
4132 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4134 PH/33 Added "verify = not_blind".
4136 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4137 Local/Makefile (with some defaults set). These are used in built scripts
4138 such as exicyclog, but they have never been used in the exim_install
4139 script (though there are many overriding facilities there). I have
4140 arranged that the exim_install script now takes note of these two
4143 PH/35 Installed configuration files for Dragonfly.
4145 PH/36 When a locally submitted message by a trusted user did not contain a
4146 From: header, and the sender address was obtained from -f or from an SMTP
4147 MAIL command, and the trusted user did not use -F to supply a sender
4148 name, $originator_name was incorrectly used when constructing a From:
4149 header. Furthermore, $originator_name was used for submission mode
4150 messages from external hosts without From: headers in a similar way,
4151 which is clearly wrong.
4153 PH/37 Added control=suppress_local_fixups.
4155 PH/38 When log_selector = +received_sender was set, and the addition of the
4156 sender made the log line's construction buffer exactly full, or one byte
4157 less than full, an overflow happened when the terminating "\n" was
4160 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4161 when the result of a list match is failure because a DNS lookup failed.
4163 PH/40 RM_COMMAND is now used in the building process.
4165 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4166 the "build-* directories that it finds.
4168 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4169 address in a domain literal was a prefix of an interface address.
4171 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4172 when verifying a sender address, unless rewrite_headers is false.
4174 PH/44 Wrote a long comment about why errors_to addresses are verified as
4175 recipients, not senders.
4177 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4178 the ratelimit ACL was added.
4180 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4182 PH/46 Added extra information about PostgreSQL errors to the error string.
4184 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4185 avoiding the need to install special zones in a real server. This is
4186 backwards compatible; if it can't find the fake resolver, it drops back.
4187 Thus, both old and new test suites can be run.
4189 TF/02 Added util/ratelimit.pl
4191 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4192 clock is set back in time.
4194 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4195 Candler <B.Candler@pobox.com>.
4197 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4198 OK for addresses that are the result of redirections.
4200 PH/48 A number of further additions for the benefit of the new test suite,
4201 including a fake gethostbyname() that interfaces to the fake DNS resolver
4204 TF/06 The fix for widen_domains has also been applied to qualify_single and
4205 search_parents which are the other dnslookup options that can cause
4208 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4211 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4213 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4214 service" if it encountered an MX record with an empty target hostname.
4215 The message is now "an MX or SRV record indicated no SMTP service".
4217 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4218 if the DNS of the sending site is misconfigured. This is quite a
4219 common situation. This change restores the behaviour of treating a
4220 helo verification defer as a failure.
4222 PH/51 If self=fail was set on a router, the bounce message did not include the
4223 actual error message.
4229 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4231 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4232 command itself ended in error, the underlying message about the transport
4233 filter timeout was being overwritten with the pipe command error. Now the
4234 underlying error message should be appended to the second error message.
4236 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4238 PH/02 Reduce the amount of output that "make" produces by default. Full output
4239 can still be requested.
4241 PH/03 The warning log line about a condition test deferring for a "warn" verb
4242 was being output only once per connection, rather than after each
4243 occurrence (because it was using the same function as for successful
4244 "warn" verbs). This seems wrong, so I have changed it.
4246 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4247 it should not have, which might have caused a crash in the right
4248 circumstances, but probably never did.
4250 PH/04 Installed a modified version of Tony Finch's patch to make submission
4251 mode fix the return path as well as the Sender: header line, and to
4252 add a /name= option so that you can make the user's friendly name appear
4255 TF/03 Added the control = fakedefer ACL modifier.
4257 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4258 Mark Lowes for thorough testing.
4260 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4262 TK/03 Merged latest SRS patch from Miles Wilton.
4264 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4265 with the definition in sysexits.h (which is #included earlier).
4266 Fortunately, Exim does not actually use EX_OK. The code used to try to
4267 preserve the sysexits.h value, by assuming that macro definitions were
4268 scanned for macro replacements. I have been disabused of this notion,
4269 so now the code just undefines EX_OK before #including unistd.h.
4271 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4272 in the smtp transport. When a block could not be written in a single
4273 write() function, the timeout was being re-applied to each part-write.
4274 This seems wrong - if the receiver was accepting one byte at a time it
4275 would take for ever. The timeout is now adjusted when this happens. It
4276 doesn't have to be particularly precise.
4278 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4279 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4281 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4282 without a domain if the check to <postmaster@domain> fails.
4284 SC/01 Eximstats: added -xls and the ability to specify output files
4285 (patch written by Frank Heydlauf).
4287 SC/02 Eximstats: use FileHandles for outputting results.
4289 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4291 SC/04 Eximstats: fixed display of large numbers with -nvr option
4293 SC/05 Eximstats: fixed merging of reports with empty tables.
4295 SC/06 Eximstats: added the -include_original_destination flag
4297 SC/07 Eximstats: removed tabs and trailing whitespace.
4299 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4301 TK/06 MBOX spool code: Add real "From " MBOX separator line
4302 so the .eml file is really in mbox format (even though
4303 most programs do not really care). Patch from Alex Miller.
4305 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4306 The latter is generated from $received_to and is only set if the
4307 message has one envelope recipient. SA can use these headers,
4308 obviously out-of-the-box. Patch from Alex Miller.
4310 PH/08 The ${def test on a variable was returning false if the variable's
4311 value was "0", contrary to what the specification has always said!
4312 The result should be true unless the variable is empty.
4314 PH/09 The syntax error of a character other than { following "${if
4315 def:variable_name" (after optional whitespace) was not being diagnosed.
4316 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4317 accidental colon was present, for example, could give incorrect results.
4319 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4320 result is used (not including appendfile, where other changes are about
4323 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4324 This involved changing a lot of size variables from int to off_t. It
4325 should work with maildirs and everything.
4327 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4328 spamd dying while we are connected to it.
4330 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4333 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4334 function for BDB 4.3.
4336 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4338 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4339 the sources for PCRE have been re-arranged and more files are now
4342 PH/15 The code I had for printing potentially long long variables in PH/11
4343 above was not the best (it lost precision). The length of off_t variables
4344 is now inspected at build time, and an appropriate printing format (%ld
4345 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4346 to be "long long int" or "long int". This is needed for the internal
4347 formatting function string_vformat().
4349 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4350 the configuration file to be ":syslog", then the script "guesses" where
4351 the logs files are, rather than using the compiled in default. In our
4352 case the guess is not the same as the compiled default, so the script
4353 suddenly stopped working when I started to use syslog. The patch checks
4354 to see if log_file_path is "". If so, it attempts to read it from exim
4355 with no configuration file to get the compiled in version, before it
4356 falls back to the previous guessing code."
4358 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4359 implementing BATV in an Exim configuration. See NewStuff for the gory
4362 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4363 Makefile that are specific to HP-UX.
4365 PH/18 If the "use_postmaster" option was set for a recipient callout together
4366 with the "random" option, the postmaster address was used as the MAIL
4367 FROM address for the random test, but not for the subsequent recipient
4368 test. It is now used for both.
4370 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4371 patch removes a few documentation additions to RFC 3028, because the
4372 latest draft now contains them. It adds the new en;ascii-case comparator
4373 and a new error check for 8bit text in MIME parts. Comparator and
4374 require names are now matched exactly. I enabled the subaddress
4375 extension, but it is not well tested yet (read: it works for me)."
4377 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4378 rework some of the code of TK/09 above to avoid the hardwired use of
4379 "%lld" and "long long". Replaced the call to snprintf() with a call to
4382 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4383 records point to non-existent hosts", "retry timeout exceeded", and
4384 "retry time not reached for any host after a long failure period".
4386 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4387 experimental DomainKeys support:
4389 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4390 (2) On an error such as an illegally used "control", the wrong name for
4391 the control was given.
4393 These problems did NOT occur unless DomainKeys support was compiled.
4395 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4397 PH/24 Added ${if match_ip condition.
4399 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4400 databases so that it will be absolutely obvious if a crash occurs in the
4401 DB library. This is a regular occurrence (often caused by mis-matched
4404 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4405 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4406 picked up on a user's system that detects such things. There doesn't seem
4407 to be a gcc warning option for this - only an attribute that has to be
4408 put on the function's prototype. It seems that in Fedora Core 4 they have
4409 set this on a number of new functions. No doubt there will be more in due
4412 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4413 specify a transport. However, if an address that was verified by such a
4414 router was the subject of a callout, Exim crashed because it tried to
4415 read the rcpt_include_affixes from the non-existent transport. Now it
4416 just assumes that the setting of that option is false. This bug was
4417 introduced by 4.51/PH/31.
4419 PH/28 Changed -d+all to exclude +memory, because that information is very
4420 rarely of interest, but it makes the output a lot bigger. People tend to
4421 do -d+all out of habit.
4423 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4424 code in os-type was giving problems when libc.so lives in lib64, like on
4427 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4428 aren't the modern standard, and it seems that some systems' include files
4429 don't always have them. Exim was already checking for some of the newer
4430 ones like T_AAAA, and defining it itself. I've added checks for all the
4431 record types that Exim uses.
4433 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4434 not automatically generating a new one, as it is supposed to. This
4435 prevented TLS from working. If the file did exist, but contained invalid
4436 data, a new version was generated, as expected. It was only the case of a
4437 non-existent file that was broken.
4439 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4440 with a change in libdomainkeys > 0.64.
4442 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4443 from DNS. If the selector record carries the flag, it now has
4444 precedence over the domain-wide flag.
4446 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4448 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4449 the use of a port name that isn't defined in /etc/services) occurred, the
4450 message was deferred as in a normal delivery, and thus remained on the
4451 spool, instead of being failed because of the mua_wrapper setting. This
4452 is now fixed, and I tidied up some of the mua_wrapper messages at the
4455 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4456 the messages in a hash of arrays rather than using individual hashes.
4457 This is a bit cleaner and results in dramatic memory savings, albeit
4458 at a slight CPU cost.
4460 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4461 as requested by Marc Sherman.
4463 SC/10 Eximstats: added histograms for user specified patterns as requested
4466 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4468 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4469 fopen() in the content-scanning modules that did not already have it.
4475 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4476 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4478 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4480 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4482 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4483 to test Sieve filters that use "vacation".
4485 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4486 that changes the way the GnuTLS parameters are stored in the cache file.
4487 The new format can be generated externally. For backward compatibility,
4488 if the data in the cache doesn't make sense, Exim assumes it has read an
4489 old-format file, and it generates new data and writes a new file. This
4490 means that you can't go back to an older release without removing the
4493 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4494 work if there are any delivery delays because "one_time" forces the
4495 parent to be marked "delivered", so its unseen clone is never tried
4496 again. For this reason, Exim now forbids the simultaneous setting of
4499 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4500 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4501 entry if you want to know the details. The fix, however, neglected to
4502 consider the case when local delivery batching is involved. The test for
4503 "previously delivered" was not happening when checking to see if an
4504 address could be batched with a previous (undelivered) one; under
4505 certain circumstances this could lead to multiple deliveries to the same
4508 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4509 in its include files, and this causes problems building Exim.
4511 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4512 header_syntax) but Exim was just ignoring anything given after a slash.
4513 In particular, this caused confusion with an attempt to use "verify =
4514 reverse_host_lookup/defer_ok". An error is now given when options are
4515 supplied for verify items that do not have them. (Maybe reverse_host_
4516 lookup should have a defer_ok option, but that's a different point.)
4518 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4519 defined by RFC 821) to 2048, because there were problems with some AUTH
4520 commands, and RFC 1869 says the size should be increased for extended
4521 SMTP commands that take arguments.
4523 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4526 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4527 "unknown" error; now it says that the functionality isn't in the binary.
4529 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4530 an address' error message when a string expansion fails (syntax or
4531 whatever). Otherwise the password may appear in the log. Following change
4532 PH/42 below, there is no longer a chance of it appearing in a bounce
4535 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4537 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4538 of its addresses was ever tried. (Bugzilla bug #2.)
4540 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4541 the result incorrectly in the debug output. (It correctly added a newline
4542 to what was transported.)
4544 TF/01 Added $received_time.
4546 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4547 commented out examples of how to interface to a virus scanner and to
4548 SpamAssassin. Also added commented examples of av_scanner and
4549 spamd_address settings.
4551 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4552 and controls are allowed in which ACLs. There were a couple of minor
4553 errors. Some of the entries in the conditions table (which is a table of
4554 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4555 negation of where the condition IS allowed.
4557 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4559 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4560 header file does not have a version number, so I've had to invent a new
4561 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4562 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4563 radiusclient), but it was contributed by a Radius user.
4565 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4566 files or directories, for OpenSSL.
4568 PH/19 When an Exim process that is running as root has to create an Exim log
4569 file, it does so in a subprocess that runs as exim:exim so as to get the
4570 ownership right at creation (otherwise, other Exim processes might see
4571 the file with the wrong ownership). There was no test for failure of this
4572 fork() call, which would lead to the process getting stuck as it waited
4573 for a non-existent subprocess. Forks do occasionally fail when resources
4574 run out. I reviewed all the other calls to fork(); they all seem to check
4577 PH/20 When checking for unexpected SMTP input at connect time (before writing
4578 the banner), Exim was not dealing correctly with a non-positive return
4579 from the read() function. If the client had disconnected by this time,
4580 the result was a log entry for a synchronization error with an empty
4581 string after "input=" when read() returned zero. If read() returned -1
4582 (an event I could not check), uninitialized data bytes were printed.
4583 There were reports of junk text (parts of files, etc) appearing after
4586 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4588 PH/22 Added support for macro redefinition, and (re)definition in between
4589 driver and ACL definitions.
4591 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4592 forgetting to use the resulting value; it was using the unexpanded value.
4594 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4595 hadn't been configured. The fix is from Juergen Kreileder, who
4596 understands it better than I do:
4598 "Here's what I see happening with three configured cyrus_sasl
4599 authenticators configured (plain, login, cram-md5):
4601 On startup auth_cyrus_sasl_init() gets called for each of these.
4602 This means three calls to sasl_listmech() without a specified mech_list.
4603 => SASL tests which mechs of all available mechs actually work
4604 => three warnings about OTP not working
4605 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4607 With the patch, sasl_listmech() also gets called three times. But now
4608 SASL's mech_list option is set to the server_mech specified in the the
4609 authenticator. Or in other words, the answer from sasl_listmech()
4610 gets limited to just the mech you're testing for (which is different
4612 => the return list contains just 'plain' or 'login', 'cram-md5' or
4613 nothing depending on the value of ob->server_mech.
4615 I've just tested the patch: Authentication still works fine,
4616 unavailable mechs specified in the exim configuration are still
4617 caught, and the auth.log warnings about OTP are gone."
4619 PH/25 When debugging is enabled, the contents of the command line are added
4620 to the debugging output, even when log_selector=+arguments is not
4623 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4624 answer is "GNU", and only if the return is "GNU/something" is the answer
4627 PH/27 $acl_verify_message is now set immediately after the failure of a
4628 verification in an ACL, and so is available in subsequent modifiers. In
4629 particular, the message can be preserved by coding like this:
4631 warn !verify = sender
4632 set acl_m0 = $acl_verify_message
4634 Previously, $acl_verify_message was set only while expanding "message"
4635 and "log_message" when a very denied access.
4637 PH/28 Modified OS/os.c-Linux with
4639 -#ifndef OS_LOAD_AVERAGE
4640 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4642 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4643 nomenclature these days.)
4645 PH/29 Installed patch from the Sieve maintainer that adds the options
4646 sieve_useraddress and sieve_subaddress to the redirect router.
4648 PH/30 In these circumstances:
4649 . Two addresses routed to the same list of hosts;
4650 . First host does not offer TLS;
4651 . First host accepts first address;
4652 . First host gives temporary error to second address;
4653 . Second host offers TLS and a TLS session is established;
4654 . Second host accepts second address.
4655 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4656 and peerdn, if requested) that were in fact used only for the second
4659 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4660 attention to any local part prefix or suffix that was matched by the
4661 router that accepted the address. It now behaves in the same way as it
4662 does for delivery: the affixes are removed from the local part unless
4663 rcpt_include_affixes is set on the transport.
4665 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4666 timeout during the DATA phase of an incoming message.
4668 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4669 applied a patch sanctioned by the Sieve maintainer.
4671 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4672 the uid or gid is negative. A case of a negative gid caused this to be
4673 noticed. The fix allows for either to be negative.
4675 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4676 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4679 PH/36 The change PH/12 above was broken. Fixed it.
4681 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4682 the grounds that routing the same address twice would always produce the
4683 same answer. This might have been true once, but it is certainly no
4684 longer true now. Routing a child address may depend on the previous
4685 routing that produced that child. Some complicated redirection strategies
4686 went wrong when messages had multiple recipients, and made Exim's
4687 behaviour dependent on the order in which the addresses were given.
4689 I have moved the duplicate checking until after the routing is complete.
4690 Exim scans the addresses that are assigned to local and remote
4691 transports, and removes any duplicates. This means that more work will be
4692 done, as duplicates will always all be routed, but duplicates are
4693 presumably rare, so I don't expect this is of any significance.
4695 For deliveries to pipes, files, and autoreplies, the duplicate checking
4696 still happens during the routing process, since they are not going to be
4699 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4700 It corrects a timeout issue with spamd. This is Ian's comment: "The
4701 background is that sometimes spamd either never reads data from a
4702 connection it has accepted, or it never writes response data. The exiscan
4703 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4704 blindly assumes that writes won't block so it may never time out."
4706 PH/39 Allow G after quota size as well as K and M.
4708 PH/40 The value set for $authenticated_id in an authenticator may not contain
4709 binary zeroes or newlines because the value is written to log lines and
4710 to spool files. There was no check on this. Now the value is run through
4711 the string_printing() function so that such characters are converted to
4712 printable escape sequences.
4714 PH/41 $message_linecount is a new variable that contains the total number of
4715 lines in the message. Compare $body_linecount, which is the count for the
4718 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4719 bounce and delay warning messages, except in certain special cases, which
4722 (a) An SMTP error message from a remote host;
4723 (b) A message specified in a :fail: redirection;
4724 (c) A message specified in a "fail" command in a system filter;
4725 (d) A message specified in a FAIL return from the queryprogram router;
4726 (e) A message specified by the cannot_route_message router option.
4728 In these cases only, Exim does include the error details in bounce and
4729 warning messages. There are also a few cases where bland messages such
4730 as "unrouteable address" or "local delivery error" are given.
4732 PH/43 $value is now also set for the "else" part of a ${run expansion.
4734 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4735 being worked on, but at least Exim now implements the latest version to
4738 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4739 process to complete was treated as a delivery failure, a timeout while
4740 writing the message to the pipe was logged, but erroneously treated as a
4741 successful delivery. Such timeouts include transport filter timeouts. For
4742 consistency with the overall process timeout, these timeouts are now
4743 treated as errors, giving rise to delivery failures by default. However,
4744 there is now a new Boolean option for the pipe transport called
4745 timeout_defer, which, if set TRUE, converts the failures into defers for
4746 both kinds of timeout. A transport filter timeout is now identified in
4749 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4750 systems where "make" and "gmake" are different, calling "gmake" at top
4751 level broke things. I've arranged for the value of $(MAKE) to be passed
4752 from the Makefile to this script so that it can call the same version of
4756 A note about Exim versions 4.44 and 4.50
4757 ----------------------------------------
4759 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4760 changes of various kinds. As a consequence, a big documentation update was
4761 needed. This delayed the release for rather longer than seemed good, especially
4762 in the light of a couple of (minor) security issues. Therefore, the changes
4763 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4764 release. So 4.44 and 4.50 are in effect two different branches that both start
4767 I have left the 4.50 change log unchanged; it contains all the changes since
4768 4.43. The change log for 4.44 is below; many of its items are identical to
4769 those for 4.50. This seems to be the most sensible way to preserve the
4770 historical information.
4776 1. Minor wording change to the doc/README.SIEVE file.
4778 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4779 computation of the current number of files was incorrect.
4781 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4782 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4785 4. Give more explanation in the error message when the command for a transport
4786 filter fails to execute.
4788 5. There are several places where Exim runs a non-Exim command in a
4789 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4790 was being done only for the command run by the queryprogram router. It is
4791 now done for all such subprocesses. The other cases are: ${run, transport
4792 filters, and the commands run by the lmtp and pipe transports.
4794 6. Added CONFIGURE_GROUP build-time option.
4796 7. Some older OS have a limit of 256 on the maximum number of file
4797 descriptors. Exim was using setrlimit() to set 1000 as a large value
4798 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4799 systems. I've change it so that if it can't get 1000, it tries for 256.
4801 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4802 was an oversight, and furthermore, ever since the addition of extra
4803 controls (e.g. 4.43/32), the checks on when to allow different forms of
4804 "control" were broken. There should now be diagnostics for all cases when a
4805 control that does not make sense is encountered.
4807 9. Added the /retain_sender option to "control=submission".
4809 10. $recipients is now available in the predata ACL (oversight).
4811 11. Tidy the search cache before the fork to do a delivery from a message
4812 received from the command line. Otherwise the child will trigger a lookup
4813 failure and thereby defer the delivery if it tries to use (for example) a
4814 cached ldap connection that the parent has called unbind on.
4816 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4817 of $address_data from the recipient verification was clobbered by the
4818 sender verification.
4820 13. The value of address_data from a sender verification is now available in
4821 $sender_address_data in subsequent conditions in the ACL statement.
4823 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4825 15. Added a new option "connect=<time>" to callout options, to set a different
4828 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4829 was its contents. (It was OK if the option was not defined at all.)
4831 17. A "Completed" log line is now written for messages that are removed from
4832 the spool by the -Mrm option.
4834 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4835 information about exactly what failed.
4837 19. Added -dd to debug only the daemon process.
4839 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4840 handles timeouts, both on the server side and network timeouts. Renamed the
4841 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4843 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4844 It is now set to "smtps".
4846 22. $host_address is now set to the target address during the checking of
4847 ignore_target_hosts.
4849 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4850 being passed; this would have caused $sender_host_name to have been used if
4851 matching the list had actually called for a host name (not very likely,
4852 since this list is usually IP addresses). A host name is now passed as
4855 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4856 code that specifies a non-restarting handler (typically sigaction() in
4857 modern systems) in an attempt to fix a rare and obscure crash bug.
4859 25. Narrowed the window for a race in the daemon that could cause it to ignore
4860 SIGCHLD signals. This is not a major problem, because they are used only to
4861 wake it up if nothing else does.
4863 26. A malformed maildirsize file could cause Exim to calculate negative values
4864 for the mailbox size or file count. Odd effects could occur as a result.
4865 The maildirsize information is now recalculated if the size or filecount
4868 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4869 support for a long time. Removed HAVE_SYS_VFS_H.
4871 28. Installed the latest version of exipick from John Jetmore.
4873 29. In an address list, if the pattern was not a regular expression, an empty
4874 subject address (from a bounce message) matched only if the pattern was an
4875 empty string. Non-empty patterns were not even tested. This was the wrong
4876 because it is perfectly reasonable to use an empty address as part of a
4877 database query. An empty address is now tested by patterns that are
4878 lookups. However, all the other forms of pattern expect the subject to
4879 contain a local part and a domain, and therefore, for them, an empty
4880 address still always fails if the pattern is not itself empty.
4882 30. Exim went into a mad DNS loop when attempting to do a callout where the
4883 host was specified on an smtp transport, and looking it up yielded more
4884 than one IP address.
4886 31. Re-factored the code for checking spool and log partition space into a
4887 function that finds that data and another that does the check. The former
4888 is then used to implement four new variables: $spool_space, $log_space,
4889 $spool_inodes, and $log_inodes.
4891 32. The RFC2047 encoding function was originally intended for short strings
4892 such as real names; it was not keeping to the 75-character limit for
4893 encoded words that the RFC imposes. It now respects the limit, and
4894 generates multiple encoded words if necessary. To be on the safe side, I
4895 have increased the buffer size for the ${rfc2047: expansion operator from
4898 33. It is now permitted to omit both strings after an "if" condition; if the
4899 condition is true, the result is "true". As before, when the second string
4900 is omitted, a false condition yields an empty string. This makes it less
4901 cumbersome to write custom ACL and router conditions.
4903 34. Failure to deliver a bounce message always caused it to be frozen, even if
4904 there was an errors_to setting on the router. The errors_to setting is now
4907 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4908 canonical form (fully expanded) before being placed in
4909 $sender_host_address.
4911 36. The table in the code that translates DNS record types into text (T_A to
4912 "A" for instance) was missing entries for NS and CNAME. It is just possible
4913 that this could have caused confusion if both these types were looked up
4914 for the same domain, because the text type is used as part of Exim's
4915 per-process caching. But the chance of anyone hitting this buglet seems
4918 37. The dnsdb lookup has been extended in a number of ways.
4920 (1) There is a new type, "zns", which walks up the domain tree until it
4921 finds some nameserver records. It should be used with care.
4923 (2) There is a new type, "mxh", which is like "mx" except that it returns
4924 just the host names, not the priorities.
4926 (3) It is now possible to give a list of domains (or IP addresses) to be
4927 looked up. The behaviour when one of the lookups defers can be
4928 controlled by a keyword.
4930 (4) It is now possible to specify the separator character for use when
4931 multiple records are returned.
4933 38. The dnslists ACL condition has been extended: it is now possible to supply
4934 a list of IP addresses and/or domains to be looked up in a particular DNS
4937 39. Added log_selector=+queue_time_overall.
4939 40. When running the queue in the test harness, wait just a tad after forking a
4940 delivery process, to get repeatability of debugging output.
4942 41. Include certificate and key file names in error message when GnuTLS fails
4943 to set them up, because the GnuTLS error message doesn't include the name
4944 of the failing file when there is a problem reading it.
4946 42. Allow both -bf and -bF in the same test run.
4948 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4950 44. The "Exiscan patch" is now merged into the mainline Exim source.
4952 45. Sometimes the final signoff response after QUIT could fail to get
4953 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4954 < 0 before doing a fflush(). This bug looks as though it goes back to the
4955 introduction of TLS in release 3.20, but "sometimes" must have been rare
4956 because the tests only now provoked it.
4958 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4959 (this can affect the format of dates).
4961 47. exim_tidydb, when checking for the continued existence of a message for
4962 which it has found a message-specific retry record, was not finding
4963 messages that were in split spool directories. Consequently, it was
4964 deleting retry records that should have stayed in existence.
4966 48. Steve fixed some bugs in eximstats.
4968 49. The SPA authentication driver was not abandoning authentication and moving
4969 on to the next authenticator when an expansion was forced to fail,
4970 contradicting the general specification for all authenticators. Instead it
4971 was generating a temporary error. It now behaves as specified.
4973 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4974 (the order specifies the preference for clients). The order is now AES256,
4975 AES128, 3DES, ARCFOUR128.
4977 51. Small patch to Sieve code - explicitly set From: when generating an
4980 52. Exim crashed if a remote delivery caused a very long error message to be
4981 recorded - for instance if somebody sent an entire SpamAssassin report back
4982 as a large number of 550 error lines. This bug was coincidentally fixed by
4983 increasing the size of one of Exim's internal buffers (big_buffer) that
4984 happened as part of the Exiscan merge. However, to be on the safe side, I
4985 have made the code more robust (and fixed the comments that describe what
4988 53. Now that there can be additional text after "Completed" in log lines (if
4989 the queue_time_overall log selector is set), a one-byte patch to exigrep
4990 was needed to allow it to recognize "Completed" as not the last thing in
4993 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4994 patch that reportedly fixes this has been added. I am not expert enough to
4995 create a test for it. This is what the patch creator wrote:
4997 "I found a little strange behaviour of ldap code when working with
4998 Windows 2003 AD Domain, where users was placed in more than one
4999 Organization Units. When I tried to give exim partial DN, the exit code
5000 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5001 But simultaneously result of request was absolutely normal ldap result,
5002 so I produce this patch..."
5004 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5005 REFERENCE, so I have modified the code to exclude the patch when that macro
5008 55. Some experimental protocols are using DNS PTR records for new purposes. The
5009 keys for these records are domain names, not reversed IP addresses. The
5010 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5011 leaves it alone. Component reversal etc. now happens only for IP addresses.
5014 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5016 57. Double the size of the debug message buffer (to 2048) so that more of very
5017 long debug lines gets shown.
5019 58. The exicyclog utility now does better if the number of log files to keep
5020 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5022 59. Two changes related to the smtp_active_hostname option:
5024 (1) $smtp_active_hostname is now available as a variable.
5025 (2) The default for smtp_banner uses $smtp_active_hostname instead
5026 of $primary_hostname.
5028 60. The host_aton() function is supposed to be passed a string that is known
5029 to be a valid IP address. However, in the case of IPv6 addresses, it was
5030 not checking this. This is a hostage to fortune. Exim now panics and dies
5031 if the condition is not met. A case was found where this could be provoked
5032 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5033 components; fortuitously, this particular loophole had already been fixed
5034 by change 4.50/55 above.
5036 If there are any other similar loopholes, the new check in host_aton()
5037 itself should stop them being exploited. The report I received stated that
5038 data on the command line could provoke the exploit when Exim was running as
5039 exim, but did not say which command line option was involved. All I could
5040 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5041 running as the user.
5044 61. There was a buffer overflow vulnerability in the SPA authentication code
5045 (which came originally from the Samba project). I have added a test to the
5046 spa_base64_to_bits() function which I hope fixes it.
5049 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5050 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5052 63. The daemon start-up calls getloadavg() while still root for those OS that
5053 need the first call to be done as root, but it missed one case: when
5054 deliver_queue_load_max is set with deliver_drop_privilege. This is
5055 necessary for the benefit of the queue runner, because there is no re-exec
5056 when deliver_drop_privilege is set.
5058 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5059 This has been fixed.
5061 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5062 was in use, was not putting the data itself into the right store pool;
5063 consequently, it could be overwritten for a subsequent message in the same
5064 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5067 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5069 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5070 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5071 the function were treating the return as a boolean value, which happened to
5072 work because 0=false and not-0=true, but is not correct code.
5074 68. The host_aton() function was not handling scoped IPv6 addresses (those
5075 with, for example, "%eth0" on the end) correctly.
5077 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5078 negated items (that is, ~something) in unsigned ints. Some compilers
5079 apparently mutter when there is no cast.
5081 70. If an address verification called from an ACL failed, and did not produce a
5082 user-specific message (i.e. there was only a "system" message), nothing was
5083 put in $acl_verify_message. In this situation, it now puts the system
5086 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5087 session; change 4.31/43 added the unwanted input to the log line - except
5088 that it did not do this in the start of session case. It now does.
5090 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5091 This is wrong and can cause the other end to generate a synchronization
5092 error if it is another Exim or anything else that does the synchronization
5093 check. A QUIT command is no longer sent after a timeout.
5095 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5096 during host lookups.
5098 74. The defer_ok option of callout verification was not working if it was used
5099 when verifying addresses in header lines, that is, for this case:
5101 verify = header_sender/callout=defer_ok
5103 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5104 those file descriptors could be used for SMTP connections. If anything
5105 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5106 could be sent to the SMTP client, causing chaos. The daemon now opens
5107 stdin, stdout, and stderr to /dev/null when it puts itself into the
5110 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5111 log by default. The user can override this with suitable Perl magic.
5113 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5114 the log message when discard triggers, was not working for the DATA ACL or
5115 for the non-SMTP ACL.
5117 78. Error message wording change in sieve.c.
5119 79. If smtp_accept_max_per_host was set, the number of connections could be
5120 restricted to fewer than expected, because the daemon was trying to set up
5121 a new connection before checking whether the processes handling previous
5122 connections had finished. The check for completed processes is now done
5123 earlier. On busy systems, this bug wouldn't be noticed because something
5124 else would have woken the daemon, and it would have reaped the completed
5127 80. If a message was submitted locally by a user whose login name contained one
5128 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5129 It caused a spool format error. I have fixed the spool reading code. A
5130 related problem was that the "from" clause in the Received: line became
5131 illegal because of the space(s). It is now covered by ${quote_local_part.
5133 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5136 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5138 83. Patch from Sieve maintainer to fix unterminated string problem in
5139 "vacation" handling.
5141 84. Some minor changes to the Linux configuration files to help with other
5142 OS variants using glibc.
5144 85. One more patch for Sieve to update vacation handling to latest spec.
5147 ----------------------------------------------------
5148 See the note above about the 4.44 and 4.50 releases.
5149 ----------------------------------------------------
5155 1. Change 4.43/35 introduced a bug that caused file counts to be
5156 incorrectly computed when quota_filecount was set in an appendfile
5159 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5160 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5163 3. Give more explanation in the error message when the command for a transport
5164 filter fails to execute.
5166 4. There are several places where Exim runs a non-Exim command in a
5167 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5168 was being done only for the command run by the queryprogram router. It is
5169 now done for all such subprocesses. The other cases are: ${run, transport
5170 filters, and the commands run by the lmtp and pipe transports.
5172 5. Some older OS have a limit of 256 on the maximum number of file
5173 descriptors. Exim was using setrlimit() to set 1000 as a large value
5174 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5175 systems. I've change it so that if it can't get 1000, it tries for 256.
5177 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5178 was an oversight, and furthermore, ever since the addition of extra
5179 controls (e.g. 4.43/32), the checks on when to allow different forms of
5180 "control" were broken. There should now be diagnostics for all cases when a
5181 control that does not make sense is encountered.
5183 7. $recipients is now available in the predata ACL (oversight).
5185 8. Tidy the search cache before the fork to do a delivery from a message
5186 received from the command line. Otherwise the child will trigger a lookup
5187 failure and thereby defer the delivery if it tries to use (for example) a
5188 cached ldap connection that the parent has called unbind on.
5190 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5191 of $address_data from the recipient verification was clobbered by the
5192 sender verification.
5194 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5195 was its contents. (It was OK if the option was not defined at all.)
5197 11. A "Completed" log line is now written for messages that are removed from
5198 the spool by the -Mrm option.
5200 12. $host_address is now set to the target address during the checking of
5201 ignore_target_hosts.
5203 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5204 being passed; this would have caused $sender_host_name to have been used if
5205 matching the list had actually called for a host name (not very likely,
5206 since this list is usually IP addresses). A host name is now passed as
5209 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5210 code that specifies a non-restarting handler (typically sigaction() in
5211 modern systems) in an attempt to fix a rare and obscure crash bug.
5213 15. Narrowed the window for a race in the daemon that could cause it to ignore
5214 SIGCHLD signals. This is not a major problem, because they are used only to
5215 wake it up if nothing else does.
5217 16. A malformed maildirsize file could cause Exim to calculate negative values
5218 for the mailbox size or file count. Odd effects could occur as a result.
5219 The maildirsize information is now recalculated if the size or filecount
5222 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5223 support for a long time. Removed HAVE_SYS_VFS_H.
5225 18. Updated exipick to current release from John Jetmore.
5227 19. Allow an empty sender to be matched against a lookup in an address list.
5228 Previously the only cases considered were a regular expression, or an
5231 20. Exim went into a mad DNS lookup loop when doing a callout where the
5232 host was specified on the transport, if the DNS lookup yielded more than
5235 21. The RFC2047 encoding function was originally intended for short strings
5236 such as real names; it was not keeping to the 75-character limit for
5237 encoded words that the RFC imposes. It now respects the limit, and
5238 generates multiple encoded words if necessary. To be on the safe side, I
5239 have increased the buffer size for the ${rfc2047: expansion operator from
5242 22. Failure to deliver a bounce message always caused it to be frozen, even if
5243 there was an errors_to setting on the router. The errors_to setting is now
5246 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5247 canonical form (fully expanded) before being placed in
5248 $sender_host_address.
5250 24. Updated eximstats to version 1.33
5252 25. Include certificate and key file names in error message when GnuTLS fails
5253 to set them up, because the GnuTLS error message doesn't include the name
5254 of the failing file when there is a problem reading it.
5256 26. Expand error message when OpenSSL has problems setting up cert/key files.
5259 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5260 (this can affect the format of dates).
5262 28. exim_tidydb, when checking for the continued existence of a message for
5263 which it has found a message-specific retry record, was not finding
5264 messages that were in split spool directories. Consequently, it was
5265 deleting retry records that should have stayed in existence.
5267 29. eximstats updated to version 1.35
5268 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5269 1.35 - bugfix such that pie charts by volume are generated correctly
5271 30. The SPA authentication driver was not abandoning authentication and moving
5272 on to the next authenticator when an expansion was forced to fail,
5273 contradicting the general specification for all authenticators. Instead it
5274 was generating a temporary error. It now behaves as specified.
5276 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5277 (the order specifies the preference for clients). The order is now AES256,
5278 AES128, 3DES, ARCFOUR128.
5280 31. Small patch to Sieve code - explicitly set From: when generating an
5283 32. Exim crashed if a remote delivery caused a very long error message to be
5284 recorded - for instance if somebody sent an entire SpamAssassin report back
5285 as a large number of 550 error lines. This bug was coincidentally fixed by
5286 increasing the size of one of Exim's internal buffers (big_buffer) that
5287 happened as part of the Exiscan merge. However, to be on the safe side, I
5288 have made the code more robust (and fixed the comments that describe what
5291 33. Some experimental protocols are using DNS PTR records for new purposes. The
5292 keys for these records are domain names, not reversed IP addresses. The
5293 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5294 leaves it alone. Component reversal etc. now happens only for IP addresses.
5297 34. The host_aton() function is supposed to be passed a string that is known
5298 to be a valid IP address. However, in the case of IPv6 addresses, it was
5299 not checking this. This is a hostage to fortune. Exim now panics and dies
5300 if the condition is not met. A case was found where this could be provoked
5301 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5302 components; fortuitously, this particular loophole had already been fixed
5303 by change 4.50/55 or 4.44/33 above.
5305 If there are any other similar loopholes, the new check in host_aton()
5306 itself should stop them being exploited. The report I received stated that
5307 data on the command line could provoke the exploit when Exim was running as
5308 exim, but did not say which command line option was involved. All I could
5309 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5310 running as the user.
5313 35. There was a buffer overflow vulnerability in the SPA authentication code
5314 (which came originally from the Samba project). I have added a test to the
5315 spa_base64_to_bits() function which I hope fixes it.
5318 36. The daemon start-up calls getloadavg() while still root for those OS that
5319 need the first call to be done as root, but it missed one case: when
5320 deliver_queue_load_max is set with deliver_drop_privilege. This is
5321 necessary for the benefit of the queue runner, because there is no re-exec
5322 when deliver_drop_privilege is set.
5324 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5325 was in use, was not putting the data itself into the right store pool;
5326 consequently, it could be overwritten for a subsequent message in the same
5327 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5330 38. Sometimes the final signoff response after QUIT could fail to get
5331 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5332 < 0 before doing a fflush(). This bug looks as though it goes back to the
5333 introduction of TLS in release 3.20, but "sometimes" must have been rare
5334 because the tests only now provoked it.
5340 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5341 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5342 Now it returns an integer. A number of calls were still expecting a T/F
5343 return. Fortuitously, in all cases, the tests worked in OK situations,
5344 which is the norm. However, things would have gone wrong on any write
5345 failures on the smtp file descriptor. This function is used when sending
5346 messages over SMTP and also when doing verify callouts.
5348 2. When Exim is called to do synchronous delivery of a locally submitted
5349 message (the -odf or -odi options), it no longer closes stderr before doing
5352 3. Implemented the mua_wrapper option.
5354 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5356 5. Implemented the functions header_remove(), header_testname(),
5357 header_add_at_position(), and receive_remove_recipient(), and exported them
5360 6. If an ACL "warn" statement specified the addition of headers, Exim already
5361 inserted X-ACL-Warn: at the start if there was no header name. However, it
5362 was not making this test for the second and subsequent header lines if
5363 there were newlines in the string. This meant that an invalid header could
5364 be inserted if Exim was badly configured.
5366 7. Allow an ACL "warn" statement to add header lines at the start or after all
5367 the Received: headers, as well as at the end.
5369 8. Added the rcpt_4xx retry error code.
5371 9. Added postmaster_mailfrom=xxx to callout verification option.
5373 10. Added mailfrom=xxxx to the callout verification option, for verify=
5376 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5377 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5379 12. Inserted some casts to stop certain compilers warning when using pointer
5380 differences as field lengths or precisions in printf-type calls (mostly
5381 affecting debugging statements).
5383 13. Added optional readline() support for -be (dynamically loaded).
5385 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5386 same clock tick as a message's arrival, so that its received time was the
5387 same as the "first fail" time on the retry record, and that message
5388 remained on the queue past the ultimate address timeout, every queue runner
5389 would try a delivery (because it was past the ultimate address timeout) but
5390 after another failure, the ultimate address timeout, which should have then
5391 bounced the address, did not kick in. This was a "< instead of <=" error;
5392 in most cases the first failure would have been in the next clock tick
5393 after the received time, and all would be well.
5395 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5396 being recognized when the domain list was tested by the match_domain
5397 condition in an expansion string.
5399 16. Added the ${str2b64: operator.
5401 17. Exim was always calling setrlimit() to set a large limit for the number of
5402 processes, without checking whether the existing limit was already
5403 adequate. (It did check for the limit on file descriptors.) Furthermore,
5404 errors from getrlimit() and setrlimit() were being ignored. Now they are
5405 logged to the main and panic logs, but Exim does carry on, to try to do its
5406 job under whatever limits there are.
5408 18. Imported PCRE 5.0.
5410 19. Trivial typo in log message " temporarily refused connection" (the leading
5413 20. If the log selector return_path_on_delivery was set and an address was
5414 redirected to /dev/null, the delivery process crashed because it assumed
5415 that a return path would always be set for a "successful" delivery. In this
5416 case, the whole delivery is bypassed as an optimization, and therefore no
5419 21. Internal re-arrangement: the function for sending a challenge and reading
5420 a response while authentication was assuming a zero-terminated challenge
5421 string. It's now changed to take a pointer and a length, to allow for
5422 binary data in such strings.
5424 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5426 23. Exim was not respecting finduser_retries when seeking the login of the
5427 uid under which it was called; it was always trying 10 times. (The default
5428 setting of finduser_retries is zero.) Also, it was sleeping after the final
5429 failure, which is pointless.
5431 24. Implemented tls_on_connect_ports.
5433 25. Implemented acl_smtp_predata.
5435 26. If the domain in control=submission is set empty, Exim assumes that the
5436 authenticated id is a complete email address when it generates From: or
5437 Sender: header lines.
5439 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5440 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5441 chown and chgrp in /bin and hostname in /usr/bin.
5443 28. Exim was keeping the "process log" file open after each use, just as it
5444 does for the main log. This opens the possibility of it remaining open for
5445 long periods when the USR1 signal hits a daemon. Occasional processlog
5446 errors were reported, that could have been caused by this. Anyway, it seems
5447 much more sensible not to leave this file open at all, so that is what now
5450 29. The long-running daemon process does not normally write to the log once it
5451 has entered its main loop, and it closes the log before doing so. This is
5452 so that log files can straightforwardly be renamed and moved. However,
5453 there are a couple of unusual error situations where the daemon does write
5454 log entries, and I had neglected to close the log afterwards.
5456 30. The text of an SMTP error response that was received during a remote
5457 delivery was being truncated at 512 bytes. This is too short for some of
5458 the long messages that one sometimes sees. I've increased the limit to
5461 31. It is now possible to make retry rules that apply only when a message has a
5462 specific sender, in particular, an empty sender.
5464 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5465 it possible to be selective about when SMTP synchronization is enforced.
5467 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5469 32. Implemented hosts_connection_nolog.
5471 33. Added an ACL for QUIT.
5473 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5476 35. Added mailbox_size and mailbox_filecount to appendfile.
5478 36. Added control = no_multiline_responses to ACLs.
5480 37. There was a bug in the logic of the code that waits for the clock to tick
5481 in the case where the clock went backwards by a substantial amount such
5482 that the microsecond fraction of "now" was more than the microsecond
5483 fraction of "then" (but the whole seconds number was less).
5485 38. Added support for the libradius Radius client library this is found on
5486 FreeBSD (previously only the radiusclient library was supported).
5492 1. When certain lookups returned multiple values in the form name=value, the
5493 quoting of the values was not always being done properly. Specifically:
5494 (a) If the value started with a double quote, but contained no whitespace,
5496 (b) If the value contained whitespace other than a space character (i.e.
5497 tabs or newlines or carriage returns) it was not quoted.
5498 This fix has been applied to the mysql and pgsql lookups by writing a
5499 separate quoting function and calling it from the lookup code. The fix
5500 should probably also be applied to nisplus, ibase and oracle lookups, but
5501 since I cannot test any of those, I have not disturbed their existing code.
5503 2. A hit in the callout cache for a specific address caused a log line with no
5504 reason for rejecting RCPT. Now it says "Previous (cached) callout
5505 verification failure".
5507 3. There was an off-by-one bug in the queryprogram router. An over-long
5508 return line was truncated at 256 instead of 255 characters, thereby
5509 overflowing its buffer with the terminating zero. As well as fixing this, I
5510 have increased the buffer size to 1024 (and made a note to document this).
5512 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5513 when Exim is waiting for an SMTP response from a remote server, Exim
5514 restarts its select() call on the socket, thereby resetting its timeout.
5515 This is not a problem when such interrupts are rare. Somebody set up a cron
5516 job to run exiwhat every 2 minutes, which is less than the normal select()
5517 timeout (5 or 10 minutes). This meant that the select() timeout never
5518 kicked in because it was always reset. I have fixed this by comparing the
5519 time when an interrupt arrives with the time at the start of the first call
5520 to select(). If more time than the timeout has elapsed, the interrupt is
5521 treated as a timeout.
5523 5. Some internal re-factoring in preparation for the addition of Sieve
5524 extensions (by MH). In particular, the "personal" test is moved to a
5525 separate function, and given an option for scanning Cc: and Bcc: (which is
5526 not set for Exim filters).
5528 6. When Exim created an email address using the login of the caller as the
5529 local part (e.g. when creating a From: or Sender: header line), it was not
5530 quoting the local part when it contained special characters such as @.
5532 7. Installed new OpenBSD configuration files.
5534 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5535 try to make them clearer.
5537 9. Callout options, other than the timeout value, were being ignored when
5538 verifying sender addresses in header lines. For example, when using
5540 verify = header_sender/callout=no_cache
5542 the cache was (incorrectly) being used.
5544 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5545 only the Cygwin environment.
5547 11. When return_path_on_delivery was set as a log selector, if different remote
5548 addresses in the same message used different return paths and parallel
5549 remote delivery occurred, the wrong values would sometimes be logged.
5550 (Whenever a remote delivery process finished, the return path value from
5551 the most recently started remote delivery process was logged.)
5553 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5554 lines when AUTH and/or TLS are in use. This is the "received protocol"
5555 field. Exim used to use "asmtp" for authenticated SMTP, without any
5556 indication (in the protocol name) for TLS use. Now it follows the RFC and
5557 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5558 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5559 names appear in log lines as well as in Received: header lines.
5561 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5562 extensions, and comparison tests, and to fix some bugs.
5564 14. Changes to the "personal" filter test:
5566 (1) The test was buggy in that it was just doing the equivalent of
5567 "contains" tests on header lines. For example, if a user's address was
5568 anne@some.where, the "personal" test would incorrectly be true for
5570 To: susanne@some.where
5572 This test is now done by extracting each address from the header in turn,
5573 and checking the entire address. Other tests that are part of "personal"
5574 are now done using regular expressions (for example, to check local parts
5575 of addresses in From: header lines).
5577 (2) The list of non-personal local parts in From: addresses has been
5578 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5579 taken from the Sieve specification recommendations.
5581 (3) If the message contains any header line starting with "List-" it is
5582 treated as non-personal.
5584 (4) The test for "circular" in the Subject: header line has been removed
5585 because it now seems ill-conceived.
5587 15. Minor typos in src/EDITME comments corrected.
5589 16. Installed latest exipick from John Jetmore.
5591 17. If headers_add on a router specified a text string that was too long for
5592 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5593 of string_sprintf() is now avoided.
5595 18. $message_body_size was not set (it was always zero) when running the DATA
5596 ACL and the local_scan() function.
5598 19. For the "mail" command in an Exim filter, no default was being set for
5599 the once_repeat time, causing a random time value to be used if "once" was
5600 specified. (If the value happened to be <= 0, no repeat happened.) The
5601 default is now 0s, meaning "never repeat". The "vacation" command was OK
5602 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5603 (I found it when inspecting the code).
5605 20. There is now an overall timeout for performing a callout verification. It
5606 defaults to 4 times the callout timeout, which applies to individual SMTP
5607 commands during the callout. The overall timeout applies when there is more
5608 than one host that can be tried. The timeout is checked before trying the
5609 next host. This prevents very long delays if there are a large number of
5610 hosts and all are timing out (e.g. when the network connections are timing
5611 out). The value of the overall timeout can be changed by specifying an
5612 additional sub-option for "callout", called "maxwait". For example:
5614 verify = sender/callout=5s,maxwait=20s
5616 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5617 to the end before writing, but this should make it even safer).
5619 22. Exim was forgetting that it had advertised PIPELINING for the second and
5620 subsequent messages on an SMTP connection. It was also not resetting its
5621 memory on STARTTLS and an internal HELO.
5623 23. When Exim logs an SMTP synchronization error within a session, it now
5624 records whether PIPELINING has been advertised or not.
5626 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5627 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5628 rather than long int.
5630 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5632 26. Added the never_mail option to autoreply.
5638 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5639 crash if the getsockname() call failed; this can happen if a connection is
5640 closed very soon after it is established. The problem was simply in the
5641 order in which certain operations were done, causing Exim to try to write
5642 to the SMTP stream before it had set up the file descriptor. The bug has
5643 been fixed by making things happen in the correct order.
5649 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5650 before the connection was closed, thus losing the rejection response.
5652 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5653 some early Solaris releases, but causes trouble in current releases where
5654 socklen_t is defined.
5656 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5659 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5662 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5663 that releasing the top of it at the end releases what was used for sub-
5664 expansions (unless the block got too big). However, discard this block if
5665 the first thing is a variable or header, so that we can use its block when
5666 it is dynamic (useful for very large $message_headers, for example).
5668 6. Lookups now cache *every* query, not just the most recent. A new, separate
5669 store pool is used for this. It can be recovered when all lookup caches are
5670 flushed. Lookups now release memory at the end of their result strings.
5671 This has involved some general refactoring of the lookup sources.
5673 7. Some code has been added to the store_xxx() functions to reduce the amount
5674 of flapping under certain conditions.
5676 8. log_incoming_interface used to affect only the <= reception log lines. Now
5677 it causes the local interface and port to be added to several more SMTP log
5678 lines, for example "SMTP connection from", and rejection lines.
5680 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5682 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5684 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5685 could be overwritten at the end of the current message (or the start of a
5686 new message if it was set in a HELO ACL). The value is now preserved for
5687 the duration of the SMTP connection.
5689 12. If a transport had a headers_rewrite setting, and a matching header line
5690 contained an unqualified address, that address was qualified, even if it
5691 did not match any rewriting rules. The underlying bug was that the values
5692 of the flags that permit the existence of unqualified sender and recipient
5693 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5694 non-local messages, and by -bnq for local messages) were not being
5695 preserved with the message after it was received.
5697 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5698 "next input=" as part of the text comprising the host identity instead of
5699 the correct text. The code was using the same buffer for two different
5700 strings. However, depending on which order the printing function evaluated
5701 its arguments, the bug did not always show up. Under Linux, for example, my
5702 test suite worked just fine.
5704 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5705 doesn't work with some older versions of Perl. It has been changed to "my",
5706 which in any case is probably the better facility to use.
5708 15. A really picky compiler found some instances of statements for creating
5709 error messages that either had too many or two few arguments for the format
5712 16. The size of the buffer for calls to the DNS resolver has been increased
5713 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5714 for addresses that have a lot of PTR records. This alleviates a problem; it
5715 does not fully solve it.
5717 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5718 buffer now truncates the list and logs the incident, which is the same
5719 action as happens when Exim is looking up a host name and its aliases.
5720 Previously in this situation something unpredictable would happen;
5721 sometimes it was "internal error: store_reset failed".
5723 18. If a server dropped the connection unexpectedly when an Exim client was
5724 using GnuTLS and trying to read a response, the client delivery process
5725 crashed while trying to generate an error log message.
5727 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5728 string, for example:
5730 warn message = H1: something\nH2: something
5732 the text was added as a single header line from Exim's point of view
5733 though it ended up OK in the delivered message. However, searching for the
5734 second and subsequent header lines using $h_h2: did not work. This has been
5735 fixed. Similarly, if a system filter added multiple headers in this way,
5736 the routers could not see them.
5738 20. Expanded the error message when iplsearch is called with an invalid key to
5739 suggest using net-iplsearch in a host list.
5741 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5742 ACLs are no longer actually imposed (and a message to that effect is
5745 22. If a "gecos" field in a passwd entry contained escaped characters, in
5746 particular, if it contained a \" sequence, Exim got it wrong when building
5747 a From: or a Sender: header from that name. A second bug also caused
5748 incorrect handling when an unquoted " was present following a character
5749 that needed quoting.
5751 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5752 was not being matched caselessly.
5754 24. Arranged for all hyphens in the exim.8 source to be escaped with
5757 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5758 properly. Recipient callout cache records were still being keyed to include
5759 the sender, even when use_sender was set false. This led to far more
5760 callouts that were necessary. The sender is no longer included in the key
5761 when use_sender is false.
5763 26. Added "control = submission" modifier to ACLs.
5765 27. Added the ${base62d: operator to decode base 62 numbers.
5767 28. dnsdb lookups can now access SRV records.
5769 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5770 the configuration file.
5772 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5773 (-v) mode. This makes the output for a verbose queue run more intelligible.
5775 31. Added a use_postmaster feature to recipient callouts.
5777 32. Added the $body_zerocount variable, containing the number of binary zero
5778 bytes in the message body.
5780 33. The time of last modification of the "new" subdirectory is now used as the
5781 "mailbox time last read" when there is a quota error for a maildir
5784 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5786 35. Added +ignore_unknown as a special item in host lists.
5788 36. Code for decoding IPv6 addresses in host lists is now included, even if
5789 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5790 address was recognized as an IP address, but was then not correctly decoded
5791 into binary, causing unexpected and incorrect effects when compared with
5798 1. Very minor rewording of debugging text in manualroute to say "list of
5799 hosts" instead of "hostlist".
5801 2. If verify=header_syntax was set, and a header line with an unqualified
5802 address (no domain) and a large number of spaces between the end of the
5803 name and the colon was received, the reception process suffered a buffer
5804 overflow, and (when I tested it) crashed. This was caused by some obsolete
5805 code that should have been removed. The fix is to remove it!
5807 3. When running in the test harness, delay a bit after writing a bounce
5808 message to get a bit more predictability in the log output.
5810 4. Added a call to search_tidyup() just before forking a reception process. In
5811 theory, someone could use a lookup in the expansion of smtp_accept_max_
5812 per_host which, without the tidyup, could leave open a database connection.
5814 5. Added the variables $recipient_data and $sender_data which get set from a
5815 lookup success in an ACL "recipients" or "senders" condition, or a router
5816 "senders" option, similar to $domain_data and $local_part_data.
5818 6. Moved the writing of debug_print from before to after the "senders" test
5821 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5822 problems for message scanning, either using a data ACL, or using
5823 local_scan() because the Received: header was not generated till after they
5824 were called (in order to set the time as the time of reception completion).
5825 I have revised the way this works. The header is now generated after the
5826 body is received, but before the ACL or local_scan() are called. After they
5827 are run, the timestamp in the header is updated.
5833 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5834 before starting a queue runner without re-exec. This happened only when
5835 deliver_drop_privilege was set or when the Exim user was set to root. The
5836 effect of the bug was that timeouts during subsequent deliveries caused
5837 crashes instead of being properly handled. The handler is now left at its
5838 default (and expected) setting.
5840 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5841 message, again when deliver_drop_privilege is set or Exim is run as root.
5842 The bug described in (1) was not present in this case, but the tidying up
5843 of the other signals was missing. I have made the two cases consistent.
5845 3. The ignore_target_hosts setting on a manualroute router was being ignored
5846 for hosts that were looked up using the /MX notation.
5848 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5851 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5852 operated on the sender address. After changing the $sender_address to <>
5853 for the sender address verify, Exim was re-instated it as the original
5854 (before rewriting) address, but remembering that it had rewritten it, so it
5855 wasn't rewriting it again. This bug also had the effect of breaking the
5856 sender address verification caching when the sender address was rewritten.
5858 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5859 This has been changed so that if the ip literal address matches
5860 ignore_target_hosts, the router declines.
5862 7. Added expansion conditions match_domain, match_address, and match_local_
5863 part (NOT match_host).
5865 8. The placeholder for the Received: header didn't have a length field set.
5867 9. Added code to Exim itself and to exim_lock to test for a specific race
5868 condition that could lead to file corruption when using MBX delivery. The
5869 issue is with the lockfile that is created in /tmp. If this file is removed
5870 after a process has opened it but before that process has acquired a lock,
5871 there is the potential for a second process to recreate the file and also
5872 acquire a lock. This could lead to two Exim processes writing to the file
5873 at the same time. The added code performs the same test as UW imapd; it
5874 checks after acquiring the lock that its file descriptor still refers to
5875 the same named file.
5877 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5878 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5881 11. Added the smtp_active_hostname option. If used, this will typically be made
5882 to depend on the incoming interface address. Because $interface_address is
5883 not set up until the daemon has forked a reception process, error responses
5884 that can happen earlier (such as "too many connections") no longer contain
5887 12. If an expansion in a condition on a "warn" statement fails because a lookup
5888 defers, the "warn" statement is abandoned, and the next ACL statement is
5889 processed. Previously this caused the whole ACL to be aborted.
5891 13. Added the iplsearch lookup type.
5893 14. Added ident_timeout as a log selector.
5895 15. Added tls_certificate_verified as a log selector.
5897 16. Added a global option tls_require_ciphers (compare the smtp transport
5898 option of the same name). This controls incoming TLS connections.
5900 17. I finally figured out how to make tls_require_ciphers do a similar thing
5901 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5902 before starting the TLS session.
5904 18. Tabs are now shown as \t in -bP output.
5906 19. If the log selector return_path_on_delivery was set, Exim crashed when
5907 bouncing a message because it had too many Received: header lines.
5909 20. If two routers both had headers_remove settings, and the first one included
5910 a superfluous trailing colon, the final name in the first list and the
5911 first name in the second list were incorrectly joined into one item (with a
5912 colon in the middle).
5918 1. Added -C and -D options to the exinext utility, mainly to make it easier
5919 to include in the automated testing, but these could be helpful when
5920 multiple configurations are in use.
5922 2. The exinext utility was not formatting the output nicely when there was
5923 an alternate port involved in the retry record key, nor when there was a
5924 message id as well (for retries that were specific to a specific message
5925 and a specific host). It was also confused by IPv6 addresses, because of
5926 the additional colons they contain. I have fixed the IPv4 problem, and
5927 patched it up to do a reasonable job for IPv6.
5929 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5930 delivery, the log line now contains "pipelined" if PIPELINING was used.
5932 4. An SMTP transport process used to panic and die if the bind() call to set
5933 an explicit outgoing interface failed. This has been changed; it is now
5934 treated in the same way as a connect() failure.
5936 5. A reference to $sender_host_name in the part of a conditional expansion
5937 that was being skipped was still causing a DNS lookup. This no longer
5940 6. The def: expansion condition was not recognizing references to header lines
5941 that used bh_ and bheader_.
5943 7. Added the _cache feature to named lists.
5945 8. The code for checking quota_filecount in the appendfile transport was
5946 allowing one more file than it should have been.
5948 9. For compatibility with Sendmail, the command line option
5956 and sets the incoming protocol and host name (for trusted callers). The
5957 host name and its colon can be omitted when only the protocol is to be set.
5958 Note the Exim already has two private options, -pd and -ps, that refer to
5959 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5960 "s", but I don't think that's a major issue.
5962 10. A number of refactoring changes to the code, none of which should affect
5965 (a) The number of logging options was getting close to filling up the
5966 32-bit word that was used as a bit map. I have split them into two classes:
5967 those that are passed in the argument to log_write(), and those that are
5968 only ever tested independently outside of that function. These are now in
5969 separate 32-bit words, so there is plenty of room for expansion again.
5970 There is no change in the user interface or the logging behaviour.
5972 (b) When building, for example, log lines, the code previously used a
5973 macro that called string_cat() twice, in order to add two strings. This is
5974 not really sufficiently general. Furthermore, there was one instance where
5975 it was actually wrong because one of the argument was used twice, and in
5976 one call a function was used. (As it happened, calling the function twice
5977 did not affect the overall behaviour.) The macro has been replaced by a
5978 function that can join an arbitrary number of extra strings onto a growing
5981 (c) The code for expansion conditions now uses a table and a binary chop
5982 instead of a serial search (which was left over from when there were very
5983 few conditions). Also, it now recognizes conditions like "pam" even when
5984 the relevant support is not compiled in: a suitably worded error message is
5985 given if an attempt is made to use such a condition.
5987 11. Added ${time_interval:xxxxx}.
5989 12. A bug was causing one of the ddress fields not to be passed back correctly
5990 from remote delivery subprocesses. The field in question was not being
5991 subsequently used, so this caused to problems in practice.
5993 13. Added new log selectors queue_time and deliver_time.
5995 14. Might have fixed a bug in maildirsizefile handling that threw up
5996 "unexpected character" debug warnings, and recalculated the data
5997 unnecessarily. In any case, I expanded the warning message to give more
6000 15. Added the message "Restricted characters in address" to the statements in
6001 the default ACL that block characters like @ and % in local parts.
6003 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6004 Three changes have been made:
6006 (a) There was a serious bug; a negative response to MAIL caused the whole
6007 recipient domain to be cached as invalid, thereby blocking all messages
6008 to all local parts at the same domain, from all senders. This bug has
6009 been fixed. The domain is no longer cached after a negative response to
6010 MAIL if the sender used is not empty.
6012 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6015 (c) A new callout option, "use_sender" has been added for people who want
6016 the modified behaviour.
6022 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6025 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6026 indeed breaks things for older releases.
6028 3. Added additional logging to the case where there is a problem reading data
6029 from a filter that is running in a subprocess using a pipe, in order to
6030 try to track down a specific problem.
6032 4. Testing facility fudge: when running in the test harness and attempting
6033 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6034 getting "No route to host". Convert this to a timeout.
6036 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6039 6. Some OS don't have socklen_t but use size_t instead. This affects the
6040 fifth argument of getsockopt() amongst other things. This is now
6041 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6042 can be set for individual OS. I have set it for SunOS5, OSF1, and
6043 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6044 some earlier ones do not.
6046 7. Change 4.30/15 was not doing the test caselessly.
6048 8. The standard form for an IPv6 address literal was being rejected by address
6049 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6050 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6051 this, as well as the form without the "IPv6" on the front (but only when
6052 address literals are enabled, of course).
6054 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6056 10. Exim crashed if a message with an empty sender address specified by -f
6057 encountered a router with an errors_to setting. This could be provoked only
6058 by a command such as
6062 where an empty string was supplied; "<>" did not hit this bug.
6064 11. Installed PCRE release 4.5.
6066 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6067 remained set. It is now erased.
6069 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6070 times from message ids (which are base 36 rather than the normal 62).
6072 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6073 were being counted as actual protocol errors, and logged if the log
6074 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6075 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6076 and DATA following a set of rejected RCPTs do not count as protocol errors.
6077 In other words, Exim assumes they were pipelined, though this may not
6078 actually be the case. Of course, in all cases the client gets an
6079 appropriate error code.
6081 15. If a lookup fails in an ACL condition, a message about the failure may
6082 be available; it is used if testing the ACL cannot continue, because most
6083 such messages specify what the cause of the deferral is. However, some
6084 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6085 that caused an old message to be retained and used if a later statement
6086 caused a defer, replacing the real cause of the deferral.
6088 16. If an IP address had so many PTR records that the DNS lookup buffer
6089 was not large enough to hold them, Exim could crash while trying to process
6090 the truncated data. It now detects and logs this case.
6092 17. Further to 4.21/58, another change has been made: if (and only if) the
6093 first line of a message (the first header line) ends with CRLF, a bare LF
6094 in a subsequent header line has a space inserted after it, so as not to
6095 terminate the header.
6097 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6098 unnecessarily, used atoi() instead of strtol(), and didn't check the
6099 termination when getting file sizes from file names by regex.
6101 19. Completely re-implemented the support for maildirsize files, in the light
6102 of a number of problems with the previous contributed implementation
6103 (4.30/29). In particular:
6105 . If the quota is zero, the maildirsize file is maintained, but no quota is
6108 . If the maildir directory does not exist, it is created before any attempt
6109 to write a maildirsize file.
6111 . The quota value in the file is just a cache; if the quota is changed in
6112 the transport, the new value overrides.
6114 . A regular expression is available for excluding directories from the
6117 20. The autoreply transport checks the characters in options that define the
6118 message's headers; it allows continued headers, but it was checking with
6119 isspace() after an embedded newline instead of explicitly looking for a
6122 21. If all the "regular" hosts to which an address was routed had passed their
6123 expiry times, and had not reached their retry times, the address was
6124 bounced, even if fallback hosts were defined. Now Exim should go on to try
6127 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6128 equivalent code in the SMTP transport. Some hosts send humungous responses
6129 to HELO/EHLO, more than 1024 it seems.
6131 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6132 gives compiler warnings in some environments. I've now done it "properly",
6135 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6136 (because of problems with the built-in one) was declared to return uschar *
6137 instead of char *, causing compiler failure.
6139 25. Fixed a file descriptor leak when processing alias/forward files.
6141 26. Fixed a minor format string issue in dbfn.c.
6143 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6145 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6146 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6149 29. When a sender address is verified, it is cached, to save repeating the test
6150 when there is more than one recipient in a message. However, when the
6151 verification involves a callout, it is possible for different callout
6152 options to be set for different recipients. It is too complicated to keep
6153 track of this in the cache, so now Exim always runs a verification when a
6154 callout is required, relying on the callout cache for the optimization.
6155 The overhead is duplication of the address routing, but this should not be
6158 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6159 to be verified with callout=postmaster, and the main callout worked but the
6160 postmaster check failed, the verification correctly failed. However, if a
6161 subsequent RCPT command asked for sender verification *without* the
6162 postmaster check, incorrect caching caused this verification also to fail,
6165 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6166 it was not caching the DNS options (qualify_single, search_parents) that
6167 were used when the lookup failed. A subsequent lookup with different
6168 options therefore always gave the same answer, though there were cases
6169 where it should not have. (Example: a "domains = !$mx_any" option on a
6170 dnslookup router: the "domains" option is always processed without any
6171 widening, but the router might have qualify_single set.) Now Exim uses the
6172 cached value only when the same options are set.
6174 32. Added John Jetmore's "exipick" utility to the distribution.
6176 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6177 than a timeout (e.g. a certificate is required, and is not provided), an
6178 Exim server now closes the connection immediately. Previously it waited for
6179 the client to close - but if the client is SSL, it seems that they each
6180 wait for each other, leading to a delay before one of them times out.
6182 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6183 maintained 0.8.x compatibility because I don't think many are using it, and
6184 it is clearly obsolete.
6186 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6189 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6190 client certificate was expired. A simple patch fixes this, though I don't
6191 understand the full logic of why the verify callback is called multiple
6194 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6195 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6196 which causes problems with some clients (such as the Certicom SSL Plus
6197 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6198 disables the coutermeasure allowing Eudora to connect."
6200 38. Exim was not checking that a write() to a log file succeeded. This could
6201 lead to Bad Things if a log got too big, in particular if it hit a file
6202 size limit. Exim now panics and dies if it cannot write to a log file, just
6203 as it does if it cannot open a log file.
6205 39. Modified OS/Makefile-Linux so that it now contains
6207 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6209 The two -D definitions ensure that Exim is compiled with large file
6210 support, which makes it possible to handle log files that are bigger than
6213 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6214 instance) a domain was checked against a named list that involved a lookup,
6215 causing $domain_data to be set, then another domain was checked against the
6216 same list, then the first domain was re-checked, the value of $domain_data
6217 after the final check could be wrong. In particular, if the second check
6218 failed, it could be set empty. This bug probably also applied to
6221 41. The strip_trailing_dot option was not being applied to the address given
6222 with the -f command-line option.
6224 42. The code for reading a message's header from the spool was incrementing
6225 $received_count, but never initializing it. This meant that the value was
6226 incorrect (doubled) while delivering a message in the same process in which
6227 it was received. In the most common configuration of Exim, this never
6228 happens - a fresh exec is done - but it can happen when
6229 deliver_drop_privilege is set.
6231 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6232 it now includes up to 150 characters of the unexpected data in the log
6235 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6236 and building data strings. The size of both of these buffers was 10 000
6237 bytes - far larger than anybody would *ever* want, thought I. Needless to
6238 say, somebody hit the limit. I have increased the maximum line length to
6239 20 000 and the maximum data length of concatenated lines to 100 000. I have
6240 also fixed two bugs, because there was no checking on these buffers. Tsk,
6241 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6242 buffer is too small.
6244 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6245 lsearch lookups. Now it does.
6247 46. When parsing a route_list item in a manualroute router, a fixed-length
6248 buffer was used for the list of hosts. I made this 1024 bytes long,
6249 thinking that nobody would ever have a list of hosts that long. Wrong.
6250 Somebody had a whole pile of complicated expansion conditions, and the
6251 string was silently truncated, leading to an expansion error. It turns out
6252 that it is easier to change to an unlimited length (owing to other changes
6253 that have happened since this code was originally written) than to build
6254 structure for giving a limitation error. The length of the item that
6255 expands into the list of hosts is now unlimited.
6257 47. The lsearch lookup could not handle data where the length of text line was
6258 more than 4095 characters. Such lines were truncated, leading to shortened
6259 data being returned. It should now handle lines of any length.
6261 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6262 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6265 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6266 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6267 no longer happens in comment lines. A list of replacements is now placed
6268 at the head of all of the source files, except those whose only change is
6269 to replace PERL_COMMAND in the very first #! line.
6271 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6272 messages on the queue, with a bottom-up merge sort, using code contributed
6273 by Michael Haardt. This should make operations like -bp somewhat faster on
6274 large queues. It won't affect queue runners, except when queue_run_in_order
6277 51. Installed eximstats 1.31 in the distribution.
6279 52. Added support for SRV lookups to the dnslookup router.
6281 53. If an ACL referred to $message_body or $message_body_end, the value was not
6282 reset for any messages that followed in the same SMTP session.
6284 54. The store-handling optimization for building very long strings was not
6285 differentiating between the different store pools. I don't think this
6286 actually made any difference in practice, but I've tidied it.
6288 55. While running the routers to verify a sender address, $sender_address
6289 was still set to the sender address. This is wrong, because when routing to
6290 send a bounce to the sender, it would be empty. Therefore, I have changed
6291 it so that, while verifying a sender address, $sender_address is set to <>.
6292 (There is no change to what happens when verifying a recipient address.)
6294 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6295 target A or AAAA records (if not already returned) without resetting the
6296 qualify_single or search_parents options of the DNS resolver. These are
6297 inappropriate in this case because the targets of MX and SRV records must
6298 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6299 target that, when qualified, matched something in the local domain. These
6300 two options are now turned off when doing these lookups.
6302 57. It seems that at least some releases of Reiserfs (which does not have the
6303 concept of a fixed number of inodes) returns zero and not -1 for the
6304 number of available inodes. This interacted badly with check_spool_inodes,
6305 which assumed that -1 was the "no such thing" setting. What I have done is
6306 to check that the total number of inodes is greater than zero before doing
6307 the test of how many are available.
6309 58. When a "warn" ACL statement has a log_message modifier, the message is
6310 remembered, and not repeated. This is to avoid a lot of repetition when a
6311 message has many recipients that cause the same warning to be written.
6312 However, Exim was preserving the list of already written lines for an
6313 entire SMTP session, which doesn't seem right. The memory is now reset if a
6314 new message is started.
6316 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6317 debugging output unless log_rewrite was also set.
6319 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6320 of (int)(handle) when we know that handle contains (void *)(-1).
6322 61. The Exim daemon panic-logs an error return when it closes the incoming
6323 connection. However "connection reset by peer" seems to be common, and
6324 isn't really an error worthy of noting specially, so that particular error
6327 62. When Exim is trying to find all the local interfaces, it used to panic and
6328 die if the ioctl to get the interface flags failed. However, it seems that
6329 on at least one OS (Solaris 9) it is possible to have an interface that is
6330 included in the list of interfaces, but for which you get a failure error
6331 for this call. This happens when the interface is not "plumbed" into a
6332 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6333 failure of the "get flags" call assumes that the interface is down.
6335 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6336 makes life easier for people who are doing arithmetic on fields extracted
6337 from dates, where you often get leading zeros that should not be
6338 interpreted as octal.
6340 64. Added qualify_domain to the redirect router, to override the global
6343 65. If a pathologically long header line contained very many addresses (the
6344 report of this problem mentioned 10 000) and each of them was rewritten,
6345 Exim could use up a very large amount of memory. (It kept on making new
6346 copies of the header line as it rewrote, and never released the old ones.)
6347 At the expense of a bit more processing, the header rewriting function has
6348 been changed so that it no longer eats memory in this way.
6350 66. The generation of the Received: header has been moved from the time that a
6351 message starts to be received, to the time that it finishes. The timestamp
6352 in the Received: header should now be very close to that of the <= log
6353 line. There are two side-effects of this change:
6355 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6356 logged header lines no longer include the local Received: line, because
6357 it has not yet been created. The same applies to a copy of the message
6358 that is returned to a non-SMTP sender when a message is rejected.
6360 (b) When a filter file is tested using -bf, no additional Received: header
6361 is added to the test message. After some thought, I decided that this
6364 This change does not affect the value of $received_for. It is still set
6365 after address rewriting, but before local_scan() is called.
6367 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6369 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6370 gave an unhelpful panic error message, and a defer error. I have managed to
6371 change this behaviour so that it now rejects any supplied certificate,
6372 which seems right, as the list of acceptable certificates is empty.
6374 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6375 gave an unhelpful defer error. I have not managed to make this reject any
6376 supplied certificates, but the error message it gives is "no certificate
6377 supplied", which is not helpful.
6379 70. exigrep's output now also includes lines that are not associated with any
6380 message, but which match the given pattern. Implemented by a patch from
6381 Martin Sluka, which also tidied up the Perl a bit.
6383 71. Recipient callout verification, like sender verification, was using <> in
6384 the MAIL FROM command. This isn't really the right thing, since the actual
6385 sender may affect whether the remote host accepts the recipient or not. I
6386 have changed it to use the actual sender in the callout; this means that
6387 the cache record is now keyed on a recipient/sender pair, not just the
6388 recipient address. There doesn't seem to be a real danger of callout loops,
6389 since a callout by the remote host to check the sender would use <>.
6390 [SEE ABOVE: changed after hitting problems.]
6392 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6393 temporary errors. However, in the case of such a code being given after
6394 the end of a data transmission (i.e. after ".") Exim was failing to write
6395 a retry record for the message. (Yes, there was some broken host that was
6396 actually sending 8xx at this point.)
6398 73. An unknown lookup type in a host list could cause Exim to panic-die when
6399 the list was checked. (An example that provoked this was putting <; in the
6400 middle of a list instead of at the start.) If this happened during a DATA
6401 ACL check, a -D file could be left lying around. This kind of configuration
6402 error no longer causes Exim to die; instead it causes a defer error. The
6403 incident is still logged to the main and panic logs.
6405 74. Buglet left over from Exim 3 conversion. The message "too many messages
6406 in one connection" was written to the rejectlog but not the mainlog, except
6407 when address rewriting (yes!) was being logged.
6409 75. Added write_rejectlog option.
6411 76. When a system filter was run not as root (that is, when system_filter_user
6412 was set), the values of the $n variables were not being returned to the
6413 main process; thus, they were not subsequently available in the $sn
6416 77. Added +return_path_on_delivery log selector.
6418 78. A connection timeout was being treated differently from recipients deferred
6419 when testing hosts_max_try with a message that was older than the host's
6420 retry timeout. (The host should not be counted, thus allowing all hosts to
6421 be tried at least once before bouncing.) This may have been the cause of an
6422 occasionally reported bug whereby a message would remain on the queue
6423 longer than the retry timeout, but would be bounced if a delivery was
6424 forced. I say "may" because I never totally pinned down the problem;
6425 setting up timeout/retry tests is difficult. See also the next item.
6427 79. The ultimate address timeout was not being applied to errors that involved
6428 a combination of host plus message (for example, a timeout on a MAIL
6429 command). When an address resolved to a number of possible hosts, and they
6430 were not all tried for each delivery (e.g. because of hosts_max_try), a
6431 message could remain on the queue longer than the retry timeout.
6433 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6436 81. Fixed an obscure SMTP outgoing bug which required at least the following
6437 conditions: (a) there was another message waiting for the same server;
6438 (b) the server returned 5xx to all RCPT commands in the first message so
6439 that the message was not completed; (c) the server dropped the connection
6440 or gave a negative response to the RSET that Exim sends to abort the
6441 transaction. The observed case was a dropped connection after DATA that had
6442 been sent in pipelining mode. That is, the server had advertised PIPELINING
6443 but was not implementing it correctly. The effect of the bug was incorrect
6444 behaviour, such as trying another host, and this could lead to a crash.
6450 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6451 and daemon.c were passed as pointers to ints; they should have been
6452 pointers to socklen_t variables (which are typically unsigned ints).
6454 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6457 3. Fixed a really odd bug that affected only the testing scheme; patching a
6458 certain fixed string in the binary changed the value of another string that
6459 happened to be identical to the end of the original first string.
6461 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6462 name", it returns that address as the IP address. On some operating
6463 systems (e.g. Solaris), it also passes back the IP address string as the
6464 "host name". However, on others (e.g. Linux), it passes back an empty
6465 string. Exim wasn't checking for this, and was changing the host name to an
6466 empty string, assuming it had been canonicalized.
6468 5. Although rare, it is permitted to have more than one PTR record for a given
6469 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6470 all the names associated with an address, because they do in Solaris.
6471 However, it seems that they do not in Linux for data that comes from the
6472 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6473 I found this out when I moved to a new Linux workstation and tried to run
6474 the Exim test suite.
6476 To get round this problem I have changed the code so that it now does its
6477 own call to the DNS to look up PTR records when searching for a host name.
6478 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6479 addresses that are only in /etc/hosts are still found.
6481 This behaviour is, however, controlled by an option called host_lookup_
6482 order, which defaults to "bydns:byaddr". If people want to use the other
6483 order, or indeed, just use one or the other means of lookup, they can
6484 specify it in this variable.
6486 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6487 some operating systems, this comes back from gethostbyaddr() as an empty
6488 string, and this is what Exim used to test for. However, it seems that in
6489 other systems, "." is yielded. Exim now tests for this case too.
6491 7. The values of check_spool_space and check_log_space are now held internally
6492 as a number of kilobytes instead of an absolute number of bytes. If a
6493 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6494 kilobyte. This means that much larger values can be stored.
6496 8. Exim monitor: an attempt to get the action menu when not actually pointing
6497 at a message produces an empty menu entitled "No message selected". This
6498 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6499 no entries in it ("Shell widget menu has zero width and/or height"). So I
6500 have added a single, blank menu entry in this case.
6502 9. Added ${quote_local_part.
6504 10. MIME decoding is now applied to the contents of Subject: header lines when
6507 11. Now that a reference to $sender_host_address automatically causes a reverse
6508 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6509 host lookup before query-style lookups in lists that might use this
6510 variable. This has therefore been abolished, and the "net-" prefix is no
6511 longer necessary for query-style lookups.
6513 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6514 have been a typo for LFLAGS, so it has been changed.
6516 13. The install script calls Exim with "-C /dev/null" in order to find the
6517 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6518 to be output. However, since Exim outputs its version number before the
6519 error, it didn't break the script. It just looked ugly. I fixed this by
6520 always allowing "-C /dev/null" if the caller is root.
6522 14. Ignore overlarge ACL variable number when reading spool file - insurance
6523 against a later release with more variables having written the file.
6525 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6526 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6527 this, as well as the form without the "IPv6" on the front.
6529 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6530 OS/Makefile-Darwin file.
6532 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6533 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6535 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6536 variable. However, some OS use ints for inodes. I've added cast to long int
6537 to get rid of the compiler warning.
6539 19. I had forgotten to lock out "/../" in configuration file names when
6540 ALT_CONFIG_PREFIX was set.
6542 20. Routers used for verification do not need to specify transports. However,
6543 if such a router generated a host list, and callout was configured, Exim
6544 crashed, because it could not find a port number from the (non-existent)
6545 transport. It now assumes port 25 in this circumstance.
6547 21. Added the -t option to exigrep.
6549 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6550 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6551 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6552 them is set, LOOKUP_LSEARCH is forced.
6554 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6556 24. Added sender and host information to the "rejected by local_scan()" log
6557 line; previously there was no indication of these.
6559 25. Added .include_if_exists.
6561 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6562 Linux. It turns out that not all file systems support this. Apparently some
6563 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6564 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6565 not supported on the file descriptor, is now ignored when Exim is trying to
6566 sync a directory. This applies only to Linux.
6568 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6570 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6571 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6574 29. Added support for maildirsize files from supplied patch (modified a bit).
6576 30. The use of :fail: followed by an empty string could lead Exim to respond to
6577 sender verification failures with (e.g.):
6579 550 Verification failed for <xxx>
6580 550 Sender verify failed
6582 where the first response line was missing the '-' that indicates it is not
6583 the final line of the response.
6585 31. The loop for finding the name of the user that called Exim had a hardwired
6586 limit of 10; it now uses the value of finduser_retries, which is used for
6587 all other user lookups.
6589 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6592 33. Exim was neglecting to zero errno before one call of strtol() when
6593 expanding a string and expecting an integer value. On some systems this
6594 resulted in spurious "integer overflow" errors. Also, it was casting the
6595 result into an int without checking.
6597 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6598 did not work. The code looks as if it has *never* worked, though it appears
6599 to have been documented since at least release 1.62. I have made it work.
6601 35. The "timeout_DNS" error in retry rules, also documented since at least
6602 1.62, also never worked. As it isn't clear exactly what this means, and
6603 clearly it isn't a major issue, I have abolished the feature by treating it
6604 as "timeout", and writing a warning to the main and panic logs.
6606 36. The display of retry rules for -brt wasn't always showing the error code
6609 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6610 timeout_connect_A, timeout_connect_MX.
6612 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6613 to the empty sender.
6615 39. The daemon was not analysing the content of -oX till after it had closed
6616 stderr and disconnected from the controlling terminal. This meant that any
6617 syntax errors were only noted on the panic log, and the return code from
6618 the command was 0. By re-arranging the code a little, I've made the
6619 decoding happen first, so such errors now appear on stderr, and the return
6620 code is 1. However, the actual setting up of the sockets still happens in
6621 the disconnected process, so errors there are still only recorded on the
6624 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6625 connections (as happens on some IP stacks) was logged at start up time as
6626 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6627 it from "IPv6 and IPv4", which means that two separate sockets are being
6630 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6631 says whether AF_INET or AF_INET6 was passed as an argument.
6633 42. Exiwhat output was messed up when time zones were included in log
6636 43. Exiwhat now gives more information about the daemon's listening ports,
6637 and whether -tls-on-connect was used.
6639 44. The "port" option of the smtp transport is now expanded.
6641 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6642 silently ignored. Now an error message is written to the main and panic
6645 46. There's a new ACL modifier called "logwrite" which writes to a log file
6646 as soon as it is encountered.
6648 47. Added $local_user_uid and $local_user_gid at routing time.
6650 48. Exim crashed when trying to verify a sender address that was being
6653 49. Exim was recognizing only a space character after ".include". It now also
6654 recognizes a tab character.
6656 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6657 extracting the relevant information from the specification. The man page no
6658 longer contains scrambled data for the -d option, and I've added a section
6659 at the front about calling Exim under different names.
6661 51. Added "extra_headers" argument to the "mail" command in filter files.
6663 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6666 53. Installed eximstats 1.29.
6668 54. Added transport_filter_timeout as a generic transport option.
6670 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6671 Cc: header lines. This was required by RFC 822, but it not required by RFC
6674 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6675 incoming messages that did not have them. Now it does so only if the
6676 message originates locally, that is, if there is no associated remote host
6677 address. When Resent- header lines are present, this applies to the Resent-
6678 lines rather than the non-Resent- lines.
6680 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6681 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6683 58. Messages for configuration errors now include the name of the main
6684 configuration file - useful now that there may be more than one file in a
6685 list (.included file names were always shown).
6687 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6688 for those rare installations that do not start the daemon as root or run it
6689 setuid root. I've cut out the call to initgroups() if the daemon is not
6692 60. The Exim user and group can now be bound into the binary as text strings
6693 that are looked up at the start of Exim's processing.
6695 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6697 62. Added $mailstore_basename variable.
6699 63. Installed patch to sieve.c from Michael Haardt.
6701 64. When Exim failed to open the panic log after failing to open the main log,
6702 the original message it was trying to log was written to stderr and debug
6703 output, but if they were not available (the usual case in production), it
6704 was lost. Now it is written to syslog before the two lines that record the
6705 failures to open the logs.
6707 65. Users' Exim filters run in subprocesses under the user's uid. It is
6708 possible for a "deliver" command or an alias in a "personal" command to
6709 provoke an address rewrite. If logging of address rewriting is configured,
6710 this fails because the process is not running as root or exim. There may be
6711 a better way of dealing with this, but for the moment (because 4.30 needs
6712 to be released), I have disabled address rewrite logging when running a
6713 filter in a non-root, non-exim process.
6719 1. The buildconfig auxiliary program wasn't quoting the value set for
6720 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6721 not defined. This bug was masked in 4.22 by the effect that was fixed in
6724 2. Some messages that were rejected after a message id was allocated were
6725 shown as "incomplete" by exigrep. It no longer does this for messages that
6726 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6728 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6729 have allow_domain_literals set, the ID did not get logged in the <= line.
6730 Domain literals are now always recognized in Message-ID: header lines.
6732 4. The first argument for a ${extract expansion item is the key name or field
6733 number. Leading and trailing spaces in this item were not being ignored,
6734 causing some misleading effects.
6736 5. When deliver_drop_privilege was set, single queue runner processes started
6737 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6738 same command in the process it spins off) were not dropping privilege.
6740 6. When the daemon running as "exim" started a queue runner, it always
6741 re-executed Exim in the spun-off process. This is a waste of effort when
6742 deliver_drop_privilege is set. The new process now just calls the
6743 queue-runner function directly.
6749 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6752 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6753 ignored. Though the use of -oP was forcing the writing of a pid file, it
6754 was always written to the default place.
6756 3. If the message "no IP address found for host xxxx" is generated during
6757 incoming verification, it is now followed by identification of the incoming
6758 connection (so you can more easily find what provoked it).
6760 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6762 5. Added some features to "harden" Exim a bit more against certain attacks:
6764 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6765 be put in Local/Makefile. This is like the never_users runtime option,
6766 but it cannot be overridden. The default setting is "root".
6768 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6769 prefix string with which any file named in a -C command line option
6772 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6773 is retained for -C and -D only if the caller of Exim is root. Without
6774 it, the exim user may also use -C and -D and retain privilege.
6776 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6777 command line option is disabled.
6779 6. Macro names set by the -D option must start with an upper case letter, just
6780 like macro names defined in the configuration file.
6782 7. Added "dereference=" facility to LDAP.
6784 8. Two instances of the typo "uknown" in the source files are fixed.
6786 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6787 the Configure-Makefile script screwed up while processing it.
6789 10. Incorporated PCRE 4.4.
6791 11. The SMTP synchronization check was not operating right at the start of an
6792 SMTP session. For example, it could not catch a HELO sent before the client
6793 waited for the greeting. There is now a check for outstanding input at the
6794 point when the greeting is written. Because of the duplex, asynchronous
6795 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6796 way, but not yet received, when the check is performed.
6798 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6799 on TCP/IP sockets, because this apparently causes some broken clients to
6802 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6803 unchanged) from the Cygwin maintainer.
6805 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6806 was supported instead of testing for mbx. Effectively a typo.
6808 15. The spa authenticator server code was not checking that the input it
6809 received was valid base64.
6811 16. The debug output line for the "set" modifier in ACLs was not showing the
6812 name of the variable that was being set.
6814 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6816 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6817 lookup on its own. Something else was needed to trigger the lookup. For
6818 example, a match in host_lookup or the need for a host name in a host list.
6819 Now, if $sender_host_name is referenced and the host name has not yet been
6820 looked up, a lookup is performed. If the lookup fails, the variable remains
6821 empty, and $host_lookup_failed is set to "1".
6823 19. Added "eqi" as a case-independent comparison operator.
6825 20. The saslauthd authentication condition could segfault if neither service
6826 nor realm was specified.
6828 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6829 error message that was logged was misleading, and incoming SMTP
6830 connections were dropped. The message is now more accurate, and temporary
6831 errors are given to SMTP connections.
6833 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6834 (except RSET) with a 421 error, until QUIT is received. However, it was
6835 failing to send a response to QUIT.
6837 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6838 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6839 ACL is now run after the helo_try_verify_hosts code.
6841 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6842 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6843 case-independent, so other case variants are also recognized). Apparently
6844 some systems use these upper case variants.
6846 25. If more than two messages were waiting for the same host, and a transport
6847 filter was specified for the transport, Exim sent two messages over the
6848 same TCP/IP connection, and then failed with "socket operation on non-
6849 socket" when it tried to send the third.
6851 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6853 27. The extern definition of crypt16() in expand.c was not being excluded when
6854 the OS had its own crypt16() function.
6856 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6857 as a preferred synonym for return_size_limit, both as an option and as an
6860 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6862 30. Changed the default configuration ACL to relax the local part checking rule
6863 for addresses that are not in any local domains. For these addresses,
6864 slashes and pipe symbols are allowed within local parts, but the sequence
6865 /../ is explicitly forbidden.
6867 31. SPA server authentication was not clearing the challenge buffer before
6870 32. log_message in a "warn" ACL statement was writing to the reject log as
6871 well as to the main log, which contradicts the documentation and doesn't
6872 seem right (because no rejection is happening). So I have stopped it.
6874 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6875 However, I am unable to do any testing of this.
6877 34. Fixed an infelicity in the appendfile transport. When checking directories
6878 for a mailbox, to see if any needed to be created, it was accidentally
6879 using path names with one or more superfluous leading slashes; tracing
6880 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6882 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6883 contents are added to the log line that is written for every discarded
6884 recipient. (Previously a log_message setting was ignored.)
6886 36. The ${quote: operator now quotes the string if it is empty.
6888 37. The install script runs exim in order to find its version number. If for
6889 some reason other than non-existence or emptiness, which it checks, it
6890 could not run './exim', it was installing it with an empty version number,
6891 i.e. as "exim-". This error state is now caught, and the installation is
6894 38. An argument was missing from the function that creates an error message
6895 when Exim fails to connect to the socket for saslauthd authentication.
6896 This could cause Exim to crash, or give a corrupted message.
6898 39. Added isip, isip4, and isip6 to ${if conditions.
6900 40. The ACL variables $acl_xx are now saved with the message, and can be
6901 accessed later in routers, transports, and filters.
6903 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6904 strings in the file are not string-expanded.
6906 42. If a MAIL command specified a SIZE value that was too large to fit into an
6907 int variable, the check against message_size_limit failed. Such values are
6908 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6909 day this will have to be increased, but I don't think I want to be around
6910 when emails are that large.
6917 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6918 iconv() is not standard in FreeBSD.
6920 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6921 IPv6 enabled. The observed symptom was a segmentation fault on return from
6922 the function os_common_find_running_interfaces() in src/os.c.
6924 3. In the check_special_case() function in daemon.c I had used "errno" as an
6925 argument name, which causes warnings on some systems. This was basically a
6926 typo, since it was named "eno" in the comments!
6928 4. The code that waits for the clock to tick (at a resolution of some fraction
6929 of a second) so as to ensure message-id uniqueness was always waiting for
6930 at least one whole tick, when it could have waited for less. [This is
6931 almost certainly not relevant at current processor speeds, where it is
6932 unlikely to ever wait at all. But we try to future-proof.]
6934 5. The function that sleeps for a time interval that includes fractions of a
6935 second contained a race. It did not block SIGALRM between setting the
6936 timer, and suspending (a couple of lines later). If the interval was short
6937 and the sigsuspend() was delayed until after it had expired, the suspension
6938 never ended. On busy systems this could lead to processes getting stuck for
6941 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6942 process, before it forks any delivery processes. The open lookup caching
6943 mechanism meant that the open file or database connection was passed into
6944 the delivery process. The problem was that delivery processes always tidy
6945 up cached lookup data. This could cause a problem for the next delivery
6946 process started by the queue runner, because the external queue runner
6947 process does not know about the closure. So the next delivery process
6948 still has data in the lookup cache. In the case of a file lookup, there was
6949 no problem because closing a file descriptor in a subprocess doesn't affect
6950 the parent. However, if the lookup was caching a connection to a database,
6951 the connection was closed, and the second delivery process was likely to
6952 see errors such as "PGSQL: query failed: server closed the connection
6953 unexpectedly". The problem has been fixed by closing all cached lookups
6954 in a queue runner before running a delivery process.
6956 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6957 seem to have the "const" qualifier which it has on other OS. I've
6960 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6961 *of the same type* (client or server) with the same public name that an
6962 error should be diagnosed.
6964 9. When Exim looked up a host name for an IP address, but failed to find the
6965 original IP address when looking up the host name (a safety check), it
6966 output the message "<ip address> does not match any IP for NULL", which was
6967 confusing, to say the least. The bug was that the host name should have
6968 appeared instead of "NULL".
6970 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6971 user that is built into the binary, and the -C or -D options is used, root
6972 privilege is dropped before the configuration file is read. In addition,
6973 logging is switched to stderr instead of the normal log files. If the
6974 configuration then re-defines the Exim user, the unprivileged environment
6975 is probably not what is expected, so Exim logs a panic warning message (but
6978 However, if deliver_drop_privilege is set, the unprivileged state may well
6979 be exactly what is intended, so the warning has been cut out in that case,
6980 and Exim is allowed to try to write to its normal log files.
6986 1. smtp_return_error_details was not giving details for temporary sender
6987 or receiver verification errors.
6989 2. Diagnose a configuration error if two authenticators have the same public
6992 3. Exim used not to create the message log file for a message until the first
6993 delivery attempt. This could be confusing when incoming messages were held
6994 for policy or load reasons. The message log file is now created at the time
6995 the message is received, and an initial "Received" line is written to it.
6997 4. The automatically generated man page for command line options had a minor
6998 bug that caused no ill effects; however, a more serious problem was that
6999 the procedure for building the man page automatically didn't always
7000 operate. Consequently, release 4.20 contains an out-of-date version. This
7001 shouldn't happen again.
7003 5. When building Exim with embedded Perl support, the script that builds the
7004 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7005 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7007 6. The freeze_tell option was not being used for messages that were frozen on
7008 arrival, either by an ACL or by local_scan().
7010 7. Added the smtp_incomplete_transaction log selector.
7012 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7013 was accepting AUTH without a new EHLO.
7015 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7016 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7019 10. Logging of TCP/IP connections (when configured) now happens in the main
7020 daemon process instead of the child process, so that the TCP/IP connection
7021 count is more accurate (but it can never be perfect).
7023 11. The use of "drop" in a nested ACL was not being handled correctly in the
7024 outer ACL. Now, if condition failure induced by the nested "drop" causes
7025 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7026 or "require"), the connection is dropped.
7028 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7029 that yield "discard" can now be used with an "accept" or a "discard" verb,
7030 but an error is generated for any others (because I can't see a useful way
7031 to define what should happen).
7033 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7034 are now processed in the same way as lines in the Exim configuration file.
7035 In particular, continuation lines are supported.
7037 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7039 15. Added -ti meaning -t -i.
7041 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7042 domains, and warn by logging if others are found.
7044 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7045 returned from GIFCONF when Exim is trying to find the list of interfaces on
7046 a host. The code in os.c has been modified to copy each ifreq to an aligned
7047 structure in all cases.
7049 Also, in some cases, the returned ifreq's were being copied to a 'struct
7050 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7051 means the last couple of bytes of an IPv6 address could be chopped if the
7052 ifreq contained only a normal sockaddr (14 bytes storage).
7054 18. Named domain lists were not supported in the hosts_treat_as_local option.
7055 An entry such as +xxxx was not recognized, and was treated as a literal
7058 19. Ensure that header lines added by a DATA ACL are included in the reject log
7059 if the ACL subsequently rejects the message.
7061 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7062 MD5 (which is deprecated).
7064 21. When testing a filter file using -bf, Exim was writing a message when it
7065 took the sender from a "From " line in the message, but it was not doing so
7066 when it took $return_path from a Return-Path: header line. It now does.
7068 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7069 with a valid header line field name (a series of printing characters
7070 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7072 23. Changed "disc" in the source to "disk" to conform to the documentation and
7073 the book and for uniformity.
7075 24. Ignore Sendmail's -Ooption=value command line item.
7077 25. When execve() failed while trying to run a command in a pipe transport,
7078 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7079 could be confused with a return value of 69 from the command itself. This
7080 has been changed to 127, the value the shell returns if it is asked to run
7081 a non-existent command. The wording for the related log line suggests a
7082 non-existent command as the problem.
7084 26. If received_header_text expands to an empty string, do not add a Received:
7085 header line to the message. (Well, it adds a token one on the spool, but
7086 marks it "old" so that it doesn't get used or transmitted.)
7088 27. Installed eximstats 1.28 (addition of -nt option).
7090 28. There was no check for failure on the call to getsockname() in the daemon
7091 code. This can fail if there is a shortage of resources on the system, with
7092 ENOMEM, for example. A temporary error is now given on failure.
7094 29. Contrary to the C standard, it seems that in some environments, the
7095 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7096 program. Exim now does this explicitly; it affects the formatting of
7097 timestamps using strftime().
7099 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7100 complaints. I've now initialized all the variables, to avoid this.
7102 32. Header lines added by a system filter were not being "seen" during
7103 transport-time rewrites.
7105 33. The info_callback() function passed to OpenSSL is set up with type void
7106 (*)(SSL *, int, int), as described somewhere. However, when calling the
7107 function (actually a macro) that sets it up, the type void(*)() is
7108 expected. I've put in a cast to prevent warnings from picky compilers.
7110 34. If a DNS black list lookup found a CNAME record, but there were no A
7111 records associated with the domain it pointed at, Exim crashed.
7113 35. If a DNS black list lookup returned more than one A record, Exim ignored
7114 all but the first. It now scans all returned addresses if a particular IP
7115 value is being sought. In this situation, the contents of the
7116 $dnslist_value variable are a list of all the addresses, separated by a
7119 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7120 to accept a host name and all its aliases if the forward lookup for any of
7121 them yielded the IP address of the incoming connection. Now it accepts only
7122 those names whose forward lookup yields the correct IP address. Any other
7123 names are discarded. This closes a loophole whereby a rogue DNS
7124 administrator could create reverse DNS records to break through a
7125 wildcarded host restriction in an ACL.
7127 37. If a user filter or a system filter that ran in a subprocess used any of
7128 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7129 the wrong values were passed to the pipe command ($thisaddress had the
7130 value of $0, $0 had the value of $1, etc). This bug was introduced by
7131 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7133 38. Improved the line breaking for long SMTP error messages from ACLs.
7134 Previously, if there was no break point between 40 and 75 characters, Exim
7135 left the rest of the message alone. Two changes have been made: (a) I've
7136 reduced the minimum length to 35 characters; (b) if it can't find a break
7137 point between 35 and 75 characters, it looks ahead and uses the first one
7138 that it finds. This may give the occasional overlong line, but at least the
7139 remaining text gets split now.
7141 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7142 file descriptors might be low, and that setting 1000 would always raise it.
7143 It turns out that in some environments, the limit is already over 1000 and
7144 that lowering it causes trouble. So now Exim takes care not to decrease it.
7146 40. When delivering a message, the value of $return_path is set to $sender_
7147 address at the start of routing (routers may change the value). By an
7148 oversight, this default was not being set up when an address was tested by
7149 -bt or -bv, which affected the outcome if any router or filter referred to
7152 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7153 log only when "message" or "log_message" are set. However, if one of the
7154 conditions was an address verification, or a call to a nested ACL, the
7155 messages generated by the underlying test were being passed through. This
7156 no longer happens. The underlying message is available in $acl_verify_
7157 message for both "message" and "log_message" expansions, so it can be
7158 passed through if needed.
7160 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7161 new expansion $bh_ to give the encoded byte string without charset
7162 translation. Translation happens only if iconv() is available; HAVE_ICONV
7163 indicates this at build time. HEADERS_CHARSET gives the charset to
7164 translate to; headers_charset can change it in the configuration, and
7165 "headers charset" can change it in an individual filter file.
7167 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7168 that creates RFC 2047 encoded "words" labels them as that charset instead
7169 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7170 expansion operator; (ii) when Exim creates a From: line for a local
7171 message; (iii) when a header line is rewritten to include a "phrase" part.
7173 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7174 buggy, causing it to skip the first lines of messages whose message ID
7175 ended in 'D'. This would not have bitten before Exim release 4.14, because
7176 message IDs were unlikely to end in 'D' before then. The effect was to have
7177 incorrect size information for certain domains.
7179 45. #include "config.h" was missing at the start of the crypt16.c module. This
7180 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7183 46. If there was a timeout during a "random" callout check, Exim treated it as
7184 a failure of the random address, and carried on sending RSET and the real
7185 address. If the delay was just some slowness somewhere, the response to the
7186 original RCPT would be taken as a response to RSET and so on, causing
7187 mayhem of various kinds.
7189 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7190 when I implemented it. It didn't allow for the fact that some option values
7191 may legitimately be negative (e.g. size_addition), and it didn't even do
7192 the right test for positive values.
7194 48. Domain names in DNS records are case-independent. Exim always looks them up
7195 in lower case. Some resolvers return domain names in exactly the case they
7196 appear in the zone file, that is, they may contain uppercase letters. Not
7197 all resolvers do this - some return always lower case. Exim was treating a
7198 change of case by a resolver as a change of domain, similar to a widening
7199 of a domain abbreviation. This triggered its re-routing code and so it was
7200 trying to route what was effectively the same domain again. This normally
7201 caused routing to fail (because the router wouldn't handle the domain
7202 twice). Now Exim checks for this case specially, and just changes the
7203 casing of the domain that it ultimately uses when it transmits the message
7206 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7209 50. If a filter generated a file delivery with a non-absolute name (possible if
7210 no home directory exists for the router), the forbid_file option was not
7213 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7214 the existing equality matching.
7216 52. Exim was using ints instead of ino_t variables in some places where it was
7217 dealing with inode numbers.
7219 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7220 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7221 called TMPDIR, and if it finds it is different, it changes its value.
7223 54. The smtp_printf() function is now made available to local_scan() so
7224 additional output lines can be written before returning. There is also an
7225 smtp_fflush() function to enable the detection of a dropped connection.
7226 The variables smtp_input and smtp_batched_input are exported to
7229 55. Changed the default runtime configuration: the message "Unknown user"
7230 has been removed from the ACL, and instead placed on the localuser router,
7231 using the cannot_route_message feature. This means that any verification
7232 failures that generate their own messages won't get overridden. Similarly,
7233 the "Unrouteable address" message that was in the ACL for unverifiable
7234 relay addresses has also been removed.
7236 56. Added hosts_avoid_esmtp to the smtp transport.
7238 57. The exicyclog script was not checking for the esoteric option
7239 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7240 will work only if exicyclog is run under the appropriate euid.
7242 58. Following a discussion on the list, the rules by which Exim recognises line
7243 endings on incoming messages have been changed. The -dropcr and drop_cr
7244 options are now no-ops, retained only for backwards compatibility. The
7245 following line terminators are recognized: LF CRLF CR. However, special
7246 processing applies to CR:
7248 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7249 nor a local message in the state where . is a terminator.
7251 (ii) If a bare CR is encountered in a header line, an extra space is added
7252 after the line terminator so as not to end the header. The reasoning
7253 behind this is that bare CRs in header lines are most likely either
7254 to be mistakes, or people trying to play silly games.
7256 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7257 was being incorrectly given as 18 bytes larger than it should have been.
7258 This is a VOB (very old bug).
7260 60. This may never have affected anything current, but just in case it has:
7261 When the local host is found other than at the start of a list of hosts,
7262 the local host, those with the same MX, and any that follow, are discarded.
7263 When the list in question was part of a longer list of hosts, the following
7264 hosts (not currently being processed) were also being discarded. This no
7265 longer happens. I'm not sure if this situation could ever has previously
7268 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7271 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7272 header. This is something that is recommended in a new Internet Draft, and
7273 is something that is documented as being done by Sendmail. There are two
7274 possible values. For messages generated by the autoreply transport, Exim
7277 Auto-Submitted: auto-replied
7279 whereas for all other generated messages (e.g. bounces) it adds
7281 Auto-Submitted: auto-generated
7283 63. The "personal" condition in filters now includes a test for the
7284 Auto-Submitted: header. If it contains the string "auto-" the message it
7285 not considered personal.
7287 64. Added rcpt_include_affixes as a generic transport option.
7289 65. Added queue_only_override (default true).
7291 66. Added the syslog_duplication option.
7293 67. If what should have been the first header line of a message consisted of
7294 a space followed by a colon, Exim was mis-interpreting it as a header line.
7295 It isn't of course - it is syntactically invalid and should therefore be
7296 treated as the start of the message body. The misbehaviour could have
7297 caused a number of strange effects, including loss of data in subsequent
7298 header lines, and spool format errors.
7300 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7301 client host had authenticated. This control can now be exercised by an ACL
7302 for more flexibility.
7304 69. By default, callouts do not happen when testing with -bh. There is now a
7305 variant, -bhc, which does actually run the callout code, including
7306 consulting and updating the callout cache.
7308 70. Added support for saslauthd authentication, courtesy of Alexander
7311 71. If statvfs() failed on the spool or log directories while checking their
7312 size for availability, Exim confusingly gave the error "space shortage".
7313 Furthermore, in debugging mode it crashed with a floating point exception.
7314 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7315 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7316 serious problem, Exim now writes to the main and panic logs when this
7317 happens, with details of the failure. It then refuses to accept the
7318 incoming message, giving the message "spool directory problem" or "log
7319 directory problem" with a 421 code for SMTP messages.
7321 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7322 0, 1, and 2 exist, because some OS complain for execs without them (see
7323 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7324 descriptors. However, the code omitted to check that the open succeeded,
7325 causing mysterious errors if for some reason the permissions on /dev/null
7326 got screwed. Now Exim writes a message to the main and panic logs, and
7327 bombs out if it can't open /dev/null.
7329 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7330 interact so that it is all more flexible. It is supposed to remain
7331 backwards compatible. Also added extra_local_interfaces.
7333 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7334 to bomb out with an assertion failure - to the client this appears as a
7335 connection drop. This problem occurs in the part of the code that was taken
7336 from the Samba project. Fortunately, the assertion is in a very simple
7337 function, so I have fixed this by reproducing the function inline in the
7338 one place where it is called, and arranging for authentication to fail
7339 instead of killing the process with assert().
7341 75. The SPA client code was not working when the server requested OEM rather
7342 than Unicode encoding.
7344 76. Added code to make require_files with a specific uid setting more usable in
7345 the case where statting the file as root fails - usually a non-root-mounted
7346 NFS file system. When this happens and the failure is EACCES, Exim now
7347 forks a subprocess and does the per-uid checking as the relevant uid.
7349 77. Added process_log_path.
7351 78. If log_file_path was not explicitly set, a setting of check_log_space or
7352 check_log_inodes was ignored.
7354 79. If a space check for the spool or log partitions fails, the incident is now
7355 logged. Of course, in the latter case the data may get lost...
7357 80. Added the %p formatting code to string_format() so that it can be used to
7358 print addresses in debug_print(). Adjusted all the address printing in the
7359 debugging in store.c to use %p rather than %d.
7361 81. There was a concern that a line of code in smtp_in.c could overflow a
7362 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7363 initially expressed, the concern was not well-founded, because trailing
7364 spaces are removed early. However, if the trailing spaces were followed by
7365 a NULL, they did not get removed, so the overflow was possible. Two fixes
7368 (a) I re-wrote the offending code in a cleaner fashion.
7369 (b) If an incoming SMTP command contains a NULL character, it is rejected
7372 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7373 runs initgroups(), so that if the Exim user is in any additional groups,
7374 they will be used during message reception.
7380 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob