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/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917, OVE-20190718-0006)
17 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
18 buffer overrun for (non-chunking) other transports.
20 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
21 TLS1.3, means that a server rejecting a client certificate is not visible
22 to the client until the first read of encrypted data (typically the
23 response to EHLO). Add detection for that case and treat it as a failed
24 TLS connection attempt, so that the normal retry-in-clear can work (if
27 JB/01 BZg 2375: fix expansions of 822 addresses having comments in local-part
28 and/or domain. Found and fixed by Jason Betts.
34 JH/01 Remove code calling the customisable local_scan function, unless a new
35 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
37 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
38 non-signal-safe functions being used.
40 JH/03 Bug 2269: When presented with a received message having a stupidly large
41 number of DKIM-Signature headers, disable DKIM verification to avoid
42 a resource-consumption attack. The limit is set at twenty.
44 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
45 report of oldest_pass in ${authres } in consequence, and separate out
46 some descriptions of reasons for verification fail.
48 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
49 files in the spool were present and unlocked. A queue-runner could spot
50 them, resulting in a duplicate delivery. Fix that by doing the unlock
51 after the unlink. Investigation by Tim Stewart. Take the opportunity to
52 add more error-checking on spoolfile handling while that code is being
55 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
56 No known attacks, no CVE, this is defensive hardening.
58 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
59 a queue-runner could start a delivery while other operations were ongoing.
60 Cutthrough delivery was a common victim, resulting in duplicate delivery.
61 Found and investigated by Tim Stewart. Fix by using the open message data
62 file handle rather than opening another, and not locally closing it (which
63 releases a lock) for that case, while creating the temporary .eml format
64 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
66 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
67 $sender_verify_failure/$recipient_verify_failure to "random".
69 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
72 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
73 Previously this would segfault.
75 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
78 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
79 like zero, since the resolver should be doing this for us, But we need one
80 as a CNAME but no MX presence gets the CNAME returned; we need to check
81 that doesn't point to an MX to declare it "no MX returned" rather than
82 "error, loop". A new main option is added so the older capability of
83 following some limited number of chain links is maintained.
85 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
87 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
88 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
89 now-deprecated earlier definitions used only specified the range up to TLS
90 1.2 (in the older-version library docs).
92 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
94 JH/15 Rework TLS client-side context management. Stop using a global, and
95 explicitly pass a context around. This enables future use of TLS for
96 connections to service-daemons (eg. malware scanning) while a client smtp
97 connection is using TLS; with cutthrough connections this is quite likely.
99 JH/16 Fix ARC verification to do AS checks in reverse order.
101 JH/17 Support a "tls" option on the ${readsocket } expansion item.
103 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
104 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
105 Previously the "utf8" would be re-prepended for every additional message.
107 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
108 Previously thery were accepted, resulting in issues when attempting to
109 forward messages to a non-supporting MTA.
111 PP/02 Let -n work with printing macros too, not just options.
113 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
114 one parent address was copied, and bogus data was used at delivery-logging
115 time. Either a crash (after delivery) or bogus log data could result.
116 Discovery and analysis by Tim Stewart.
118 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
119 Previously if the string ended mid-character, we did not insert the
120 promised '?' replacement.
122 PP/04 Documentation: current string operators work on bytes, not codepoints.
124 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
125 should pack well giving a smaller memory footprint so better caching and
126 therefore performance. Group the declarations where this can't be done so
127 that the byte-sized flag variables are not interspersed among pointer
128 variables, giving a better chance of good packing by the compiler.
130 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
131 non-null, to avoid issues with sites running BATV. Previously reports were
132 sent with an empty envelope sender so looked like bounces.
134 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
135 The ignore_error flag wasn't being returned from the filter subprocess so
136 was not set for later routers. Investigation and fix by Matthias Kurz.
138 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
139 and a msg:complete for the whole, when a message is manually removed using
140 -Mrm. Developement by Matthias Kurz, hacked on by JH.
142 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
143 a "Gnu special" function, asprintf() in the DB utility binary builds; I
144 hope that is portable enough.
146 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
147 requiring a known-CA anchor certificate; make it now rely entirely on the
148 TLSA as an anchor. Checking the name on the leaf cert against the name
149 on the A-record for the host is still done for TA (but not for EE mode).
151 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
152 included in delivery lines for non-proxied connections, when compiled with
153 SUPPORT_SOCKS and running with proxy logging enabled.
155 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
156 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
157 move the existing event to fire before the normal logging of message
158 failure so that custom logging is bracketed by normal logging.
160 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
161 msg:fail:internal event. Developement by Matthias Kurz.
163 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
164 far too small for todays use of crypto signatures stored there. Go all
165 the way to the max DNS message size of 64kB, even though this might be
166 overmuch for IOT constrained device use.
168 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
169 copy a string over itself. The library routine is documented as not
170 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
172 JH/32 For main options check_spool_space and check_inode_space, where the
173 platform supports 64b integers, support more than the previous 2^31 kB
174 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
175 the previous G, M, k.
177 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
178 $authenticated_fail_id variable on authentication failure. Previously
181 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
182 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
183 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
184 GNUTLS_SEC_PARAM_MEDIUM.
186 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
187 side. Previously we would continue as if no SNI had been received.
189 JH/36 Harden the handling of string-lists. When a list consisted of a sole
190 "<" character, which should be a list-separator specification, we walked
191 off past the nul-terimation.
193 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
194 causes) even when the retry time is not yet met. Previously they were
195 not, meaning that when (say) an account was over-quota and temp-rejecting,
196 and multiple senders' messages were queued, only one sender would get
197 notified on each configured delay_warning cycle.
199 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
201 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
202 cipher-suites, an error can be left on the stack even for a succeeding
203 accept; this results in impossible error messages when a later operation
206 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
207 return error codes indicating retry. Under TLS1.3 this becomes required.
209 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
210 it only wrote the new authenticators, resulting in a lack of tracking of
211 peer changes of ESMTP extensions until the next cache flush.
213 JH/41 Fix the loop reading a message header line to check for integer overflow,
214 and more-often against header_maxsize. Previously a crafted message could
215 induce a crash of the recive process; now the message is cleanly rejected.
217 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
218 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
224 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
225 When redis_servers is set to a list of > 1 element, and the Redis servers
226 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
227 case of MOVED into a DEFER case instead, thus moving the query onto the
228 next server in the list. For a cluster of N elements, all N servers must
229 be defined in redis_servers.
231 GF/02 Catch and remove uninitialized value warning in exiqsumm
232 Check for existence of @ARGV before looking at $ARGV[0]
234 JH/01 Replace the store_release() internal interface with store_newblock(),
235 which internalises the check required to safely use the old one, plus
236 the allocate and data copy operations duplicated in both (!) of the
237 extant use locations.
239 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
240 modifier. This matches the restriction on the commandline.
242 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
243 Previously only the last row was returned.
245 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
246 we assumed that tags in the header were well-formed, and parsed the
247 element content after inspecting only the first char of the tag.
248 Assumptions at that stage could crash the receive process on malformed
251 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
252 While running the DKIM ACL we operate on the Permanent memory pool so that
253 variables created with "set" persist to the DATA ACL. Also (at any time)
254 DNS lookups that fail create cache records using the Permanent pool. But
255 expansions release any allocations made on the current pool - so a dnsdb
256 lookup expansion done in the DKIM ACL releases the memory used for the
257 DNS negative-cache, and bad things result. Solution is to switch to the
258 Main pool for expansions.
259 While we're in that code, add checks on the DNS cache during store_reset,
260 active in the testsuite.
261 Problem spotted, and debugging aided, by Wolfgang Breyha.
263 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
264 When none of the hosts presented to a transport match an already-open
265 connection, close it and proceed with the list. Previously we would
266 queue the message. Spotted by Lena with Yahoo, probably involving
269 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
270 Previously a spurious "250 OK id=" response was appended to the proper
273 JH/08 The "support for" informational output now, which built with Content
274 Scanning support, has a line for the malware scanner interfaces compiled
275 in. Interface can be individually included or not at build time.
277 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
278 by the template makefile "src/EDITME". The "STREAM" support for an older
279 ClamAV interface method is removed.
281 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
282 rows affected is given instead).
284 JH/11 The runtime Berkeley DB library version is now additionally output by
285 "exim -d -bV". Previously only the compile-time version was shown.
287 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
288 SMTP connection. Previously, when one had more receipients than the
289 first, an abortive onward connection was made. Move to full support for
290 multiple onward connections in sequence, handling cutthrough connection
291 for all multi-message initiating connections.
293 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
294 routers. Previously, a multi-recipient message would fail to match the
295 onward-connection opened for the first recipient, and cause its closure.
297 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
298 a timeout on read on a GnuTLS initiating connection, resulting in the
299 initiating connection being dropped. This mattered most when the callout
300 was marked defer_ok. Fix to keep the two timeout-detection methods
303 JH/15 Relax results from ACL control request to enable cutthrough, in
304 unsupported situations, from error to silently (except under debug)
305 ignoring. This covers use with PRDR, frozen messages, queue-only and
308 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
310 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
311 metadata, resulting in a crash in free().
313 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
314 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
315 Broken also in d185889f4, with init system revamp.
317 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
318 Previously we abruptly closed the connection after reading a malware-
319 found indication; now we go on to read the "scan ok" response line,
322 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
323 ACL. Previously, a crash would result.
325 JH/19 Speed up macro lookups during configuration file read, by skipping non-
326 macro text after a replacement (previously it was only once per line) and
327 by skipping builtin macros when searching for an uppercase lead character.
329 JH/20 DANE support moved from Experimental to mainline. The Makefile control
330 for the build is renamed.
332 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
333 was allocated for every new TLS startup, meaning one per message. Fix
334 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
336 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
337 reported the original. Fix to report (as far as possible) the ACL
338 result replacing the original.
340 JH/23 Fix memory leak during multi-message connections using STARTTLS under
341 OpenSSL. Certificate information is loaded for every new TLS startup,
342 and the resources needed to be freed.
344 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
346 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
349 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
350 DATA response info to the (existing) per-recipient response info for
351 the "C=" log element. It can have useful tracking info from the
352 destination system. Patch from Simon Arlott.
354 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
355 length value. Previously this would segfault.
357 HS/02 Support Avast multiline protoocol, this allows passing flags to
358 newer versions of the scanner.
360 JH/28 Ensure that variables possibly set during message acceptance are marked
361 dead before release of memory in the daemon loop. This stops complaints
362 about them when the debug_store option is enabled. Discovered specifically
363 for sender_rate_period, but applies to a whole set of variables.
364 Do the same for the queue-runner and queue-list loops, for variables set
365 from spool message files. Do the same for the SMTP per-message loop, for
366 certain variables indirectly set in ACL operations.
368 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
369 as a multi-recipient message from a mailinglist manager). The coding had
370 an arbitrary cutoff number of characters while checking for more input;
371 enforced by writing a NUL into the buffer. This corrupted long / fast
372 input. The problem was exposed more widely when more pipelineing of SMTP
373 responses was introduced, and one Exim system was feeding another.
374 The symptom is log complaints of SMTP syntax error (NUL chars) on the
375 receiving system, and refused recipients seen by the sending system
376 (propating to people being dropped from mailing lists).
377 Discovered and pinpointed by David Carter.
379 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
380 replaced by the ${authresults } expansion.
382 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
384 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
385 allows proper process termination in container environments.
387 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
388 Previously the "final dot" had a newline after it; ensure it is CR,LF.
390 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
391 and "err_perm", deprecated since 4.83 when the RFC-defined words
392 "temperror" and "permerror" were introduced.
394 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
395 transport-filters or DKIM-signing. The restriction was lost in the
396 consolidation of verify-callout and delivery SMTP handling.
397 Extend the restriction to also cover ARC-signing.
399 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
400 in defer=pass mode supply a 450 to the initiator. Previously the message
403 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
404 tls_require_ciphers is used as before.
406 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
407 "pass_unscanned". Only tmpfails from the scanner are written to
408 the paniclog, as they may require admin intervention (permission
409 denied, license issues). Other scanner errors (like decompression
410 bombs) do not cause a paniclog entry.
412 JH/36 Fix reinitialisation of DKIM logging variable between messages.
413 Previously it was possible to log spurious information in receive log
416 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
417 triggered odd behaviour from Outlook Express clients.
419 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
422 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
423 since the IETF WG has not yet settled on that versus the original
424 "bare" representation.
426 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
427 Previously the millisecond value corrupted the output.
428 Fix also for syslog_pid=no and log_selector +pid, for which the pid
429 corrupted the output.
435 JH/01 Rework error string handling in TLS interface so that the caller in
436 more cases is responsible for logging. This permits library-sourced
437 string to be attached to addresses during delivery, and collapses
438 pairs of long lines into single ones.
440 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
441 during configuration. Wildcards are allowed and expanded.
443 JH/02 Rework error string handling in DKIM to pass more info back to callers.
444 This permits better logging.
446 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
447 do not close it down and have the child transport start it up again on
448 the passed-on TCP connection. Instead, proxy the child (and any
449 subsequent ones) for TLS via a unix-domain socket channel. Logging is
450 affected: the continued delivery log lines do not have any DNSSEC, TLS
451 Certificate or OCSP information. TLS cipher information is still logged.
453 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
454 identical IP addresses on different listening ports. Will also affect
457 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
458 add noisy ifdef guards to special-case this sillyness.
459 Patch from Bernd Kuhls.
461 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
462 than 255 are no longer allowed.
464 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
465 Disable the session-cache too, which might reduce our load. Since we
466 currrectly use a new context for every connection, both as server and
467 client, there is no benefit for these.
468 GnuTLS appears to not support tickets server-side by default (we don't
469 call gnutls_session_ticket_enable_server()) but client side is enabled
470 by default on recent versions (3.1.3 +) unless the PFS priority string
473 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
474 <https://reproducible-builds.org/specs/source-date-epoch/>.
476 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
477 the check for any unsuccessful recipients did not notice the limit, and
478 erroneously found still-pending ones.
480 JH/08 Pipeline CHUNKING command and data together, on kernels that support
481 MSG_MORE. Only in-clear (not on TLS connections).
483 JH/09 Avoid using a temporary file during transport using dkim. Unless a
484 transport-filter is involved we can buffer the headers in memory for
485 creating the signature, and read the spool data file once for the
486 signature and again for transmission.
488 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
489 4.77 as the kernel support then wasn't solid, having issues in 64bit
490 mode. Now, it's been long enough. Add support for FreeBSD also.
492 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
493 case where the routing stage had gathered several addresses to send to
494 a host before calling the transport for the first, we previously failed
495 to close down TLS in the old transport process before passing the TCP
496 connection to the new process. The new one sent a STARTTLS command
497 which naturally failed, giving a failed delivery and bloating the retry
498 database. Investigation and fix prototype from Wolfgang Breyha.
500 JH/12 Fix check on SMTP command input synchronisation. Previously there were
501 false-negatives in the check that the sender had not preempted a response
502 or prompt from Exim (running as a server), due to that code's lack of
503 awareness of the SMTP input buffering.
505 PP/04 Add commandline_checks_require_admin option.
506 Exim drops privileges sanely, various checks such as -be aren't a
507 security problem, as long as you trust local users with access to their
508 own account. When invoked by services which pass untrusted data to
509 Exim, this might be an issue. Set this option in main configuration
510 AND make fixes to the calling application, such as using `--` to stop
513 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
514 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
515 responses to those, into a single TLS record each way (this usually means
516 a single packet). As a side issue, smtp_enforce_sync now works on TLS
519 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
520 affects you only if you're dancing at the edge of the param size limits.
521 If you are, and this message makes sense to you, then: raise the
522 configured limit or use OpenSSL 1.1. Nothing we can do for older
525 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
526 cmdline element to get the documented default one. Previously it was
529 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
530 are unsuable for non-message-specific reasons - eg. connection timeout,
531 banner-time rejection.
533 JH/16 Fix logging of delivery remote port, when specified by router, under
536 PP/06 Repair manualroute's ability to take options in any order, even if one
537 is the name of a transport.
540 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
542 JH/17 Change the list-building routines interface to use the expanding-string
543 triplet model, for better allocation and copying behaviour.
545 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
546 Previously it was constructed the first time a possibly-matching string
547 was met in the configuration file input during startup; now it is done
550 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
551 compatible one, to avoid the (poorly documented) possibility of a config
552 file in the working directory redirecting the DB files, possibly correpting
553 some existing file. CVE-2017-10140 assigned for BDB.
555 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
556 cache-hot. Previously, although the result was properly cached, the
557 initial verify call returned a defer.
559 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
560 the main verify for receipient in uncached-mode.
562 JH/22 Retire historical build files to an "unsupported" subdir. These are
563 defined as "ones for which we have no current evidence of testing".
565 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
566 if present. Previously it was ignored.
568 JH/24 Start using specified-initialisers in C structure init coding. This is
569 a C99 feature (it's 2017, so now considered safe).
571 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
572 if was a fixed-sized field and bitmask ops via macros; it is now more
575 PP/07 GitHub PR 56: Apply MariaDB build fix.
576 Patch provided by Jaroslav Å karvada.
578 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
579 during Coverity cleanups [4.87 JH/47]
580 Diagnosis and fix provided by Michael Fischer v. Mollard.
582 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
583 the right size to place the terminating semicolon on its own folded
584 line, the header hash was calculated to an incorrect value thanks to
585 the (relaxed) space the fold became.
587 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
588 and confused the parent.
590 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
591 which could crash as a result. This could lead to undeliverable messages.
593 JH/28 Logging: "next input sent too soon" now shows where input was truncated
596 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
597 matters on fast-turnover and PID-randomising systems, which were getting
598 out-of-order delivery.
600 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
601 a possibly-overlapping copy. The symptom was that "Remote host closed
602 connection in response to HELO" was logged instead of the actual 4xx
605 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
606 Previously only that bufferd was discarded, resulting in SYMTP command
609 JH/32 DKIM: when a message has multiple signatures matching an identity given
610 in dkim_verify_signers, run the dkim acl once for each. Previously only
611 one run was done. Bug 2189.
613 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
614 "panic the current process" to "deliberately defer". The panic log is
615 still written with the problem list name; the mail and reject logs now
616 get a temp-reject line for the message that was being handled, saying
617 something like "domains check lookup or other defer". The SMTP 451
618 message is still "Temporary local problem".
620 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
621 A crafted sequence of BDAT commands could result in in-use memory beeing
622 freed. CVE-2017-16943.
624 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
625 from SMTP input. Previously it was always done; now only done for DATA
626 and not BDAT commands. CVE-2017-16944.
628 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
629 to the message (such as an overlong header line). Previously this was
630 not done and we did not exit BDAT mode. Followon from the previous item
631 though a different problem.
637 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
638 than -2003 did; needs libidn2 in addition to libidn.
640 JH/02 The path option on a pipe transport is now expanded before use.
642 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
643 Patch provided by "Björn", documentation fix added too.
645 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
646 missing a wire-to-host endian conversion.
648 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
649 close after a BDAT command line could be taken as a following command,
650 giving a synch failure. Fix by only checking for synch immediately
651 before acknowledging the chunk.
653 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
654 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
655 macro. Patches provided by Josh Soref.
657 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
658 Previously we did not; the RFC seems ambiguous and VRFY is not listed
659 by IANA as a service extension. However, John Klensin suggests that we
662 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
663 the dkim code may be unix-mode line endings rather than smtp wire-format
664 CRLF, so prepend a CR to any bare LF.
666 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
667 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
669 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
670 the first were themselves being wrongly included in the feed into dkim
671 processing; with most chunk sizes in use this resulted in an incorrect
672 body hash calculated value.
674 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
675 DKIM signature block, for verification. Although advised against by
676 standards it is specifically not ruled illegal.
678 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
680 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
681 missing a body hash (the bh= tag).
683 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
684 It seems that HAProxy sends the Proxy Protocol information in clear and
685 only then does a TLS startup, so do the same.
687 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
688 TCP connections (such as for Spamd) unless the daemon successfully set
689 Fast Open mode on its listening sockets. This fixes breakage seen on
690 too-old kernels or those not configured for Fast Open, at the cost of
691 requiring both directions being enabled for TFO, and TFO never being used
692 by non-daemon-related Exim processes.
694 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
695 endings, at least on the first header line. Try to canonify any that get
696 past that check, despite the cost.
698 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
699 now limited to an arbitrary five deep, while parsing addresses with the
700 strip_excess_angle_brackets option enabled.
702 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
703 instead leave the unprompted TLS handshake in socket buffer for the
704 TLS library to consume.
706 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
708 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
710 JH/16 Drop variables when they go out of scope. Memory management drops a whole
711 region in one operation, for speed, and this leaves assigned pointers
712 dangling. Add checks run only under the testsuite which checks all
713 variables at a store-reset and panics on a dangling pointer; add code
714 explicitly nulling out all the variables discovered. Fixes one known
715 bug: a transport crash, where a dangling pointer for $sending_ip_address
716 originally assigned in a verify callout, is re-used.
718 PP/06 Drop '.' from @INC in various Perl scripts.
720 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
722 PP/08 Reduce a number of compilation warnings under clang; building with
723 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
724 should be warning-free.
726 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
728 HS/01 Fix portability problems introduced by PP/08 for platforms where
729 realloc(NULL) is not equivalent to malloc() [SunOS et al].
731 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
732 chunk. This allows us to accept broken chunked messages. We need a more
733 general solution here.
735 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
736 already-broken messages in the queue.
738 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
740 JH/19 Fix reference counting bug in routing-generated-address tracking.
746 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
747 supports it and a size is available (ie. the sending peer gave us one).
749 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
750 years of being deprecated). The replacements are the ACLs
751 acl_smtp_mime and acl_not_smtp_mime.
753 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
754 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
755 or even an in-clear connection were permitted. Now, if the host lookup
756 was dnssec and dane was requested then the host is only used if the
757 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
758 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
759 if one fails this test.
760 This means that a poorly-configured remote DNS will make it incommunicado;
761 but it protects against a DNS-interception attack on it.
763 JH/04 Bug 1810: make continued-use of an open smtp transport connection
764 non-noisy when a race steals the message being considered.
766 JH/05 If main configuration option tls_certificate is unset, generate a
767 self-signed certificate for inbound TLS connections.
769 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
770 in rewrites and routers.
772 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
773 and logged a warning sing 4.83; now they are a configuration file error.
775 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
776 (lacking @domain). Apply the same qualification processing as RCPT.
778 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
780 JH/10 Support ${sha256:} applied to a string (as well as the previous
783 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
784 a cutthrough deliver is pending, as we always want to make a connection.
785 This also avoids re-routing the message when later placing the cutthrough
786 connection after a verify cache hit.
787 Do not update it with the verify result either.
789 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
790 when routing results in more than one destination address.
792 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
793 signing (which inhibits the cutthrough capability). Previously only
794 the presence of an option was tested; now an expansion evaluating as
795 empty is permissible (obviously it should depend only on data available
796 when the cutthrough connection is made).
798 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
799 the relevant preceding SMTP command did not note the pipelining mode.
801 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
802 Previously they were not counted.
804 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
805 as one having no matching records. Previously we deferred the message
806 that needed the lookup.
808 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
809 distinguished as "(=".
811 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
812 for missing MX records. Previously it only worked for missing A records.
814 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
816 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
817 after the data-go-ahead and data-ack. Patch from Jason Betts.
819 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
820 even for a "none" policy. Patch from Tony Meyer.
822 JH/22 Fix continued use of a connection for further deliveries. If a port was
823 specified by a router, it must also match for the delivery to be
826 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
827 When one of the recipients of a message was unsuitable for the connection
828 (has no matching addresses), we lost track of needing to mark it
829 deferred. As a result mail would be lost.
831 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
833 JH/25 Decoding ACL controls is now done using a binary search; the source code
834 takes up less space and should be simpler to maintain. Merge the ACL
835 condition decode tables also, with similar effect.
837 JH/26 Fix problem with one_time used on a redirect router which returned the
838 parent address unchanged. A retry would see the parent address marked as
839 delivered, so not attempt the (identical) child. As a result mail would
842 JH/27 Fix a possible security hole, wherein a process operating with the Exim
843 UID can gain a root shell. Credit to http://www.halfdog.net/ for
844 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
847 JH/28 Enable {spool,log} filesystem space and inode checks as default.
848 Main config options check_{log,spool}_{inodes,space} are now
849 100 inodes, 10MB unless set otherwise in the configuration.
851 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
852 Previously it only applied to the main-section connection policy
855 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
857 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
858 by me. Added RFC7919 DH primes as an alternative.
860 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
861 are not in the system include path.
863 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
864 GnuTLS, when a session startup failed (eg because the client disconnected)
865 Exim did stdio operations after fclose. This was exposed by a recent
866 change which nulled out the file handle after the fclose.
868 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
869 signed directly by the cert-signing cert, rather than an intermediate
870 OCSP-signing cert. This is the model used by LetsEncrypt.
872 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
874 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
875 an incoming connection.
877 HS/02 Bug 1802: Do not half-close the connection after sending a request
880 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
881 fallback to "prime256v1".
883 JH/34 SECURITY: Use proper copy of DATA command in error message.
884 Could leak key material. Remotely exploitable. CVE-2016-9963.
890 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
891 and 3.4.4 - once the server is enabled to respond to an OCSP request
892 it does even when not requested, resulting in a stapling non-aware
893 client dropping the TLS connection.
895 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
896 support variable-length bit vectors. No functional change.
898 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
899 The I= interface field on outgoing lines is now after the H= remote
900 host field, same as incoming lines. There is a separate
901 outgoing_interface log selector which allows you to disable the
904 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
905 If not running log_selector +smtp_connection the mainlog would be held
906 open indefinitely after a "too many connections" event, including to a
907 deleted file after a log rotate. Leave the per net connection logging
908 leaving it open for efficiency as that will be quickly detected by the
909 check on the next write.
911 HS/01 Bug 1671: Fix post transport crash.
912 Processing the wait-<transport> messages could crash the delivery
913 process if the message IDs didn't exist for some reason. When
914 using 'split_spool_directory=yes' the construction of the spool
915 file name failed already, exposing the same netto behaviour.
917 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
918 mime_regex ACL conditions.
920 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
921 to DSN fail messages (bounces): remote IP, remote greeting, remote response
922 to HELO, local diagnostic string.
924 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
925 log line to debug. Even when configured with a tls authenticator many
926 client connections are expected to not authenticate in this way, so
927 an authenticate fail is not an error.
929 HS/02 Add the Exim version string to the process info. This way exiwhat
930 gives some more detail about the running daemon.
932 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
933 matter for fast-change records such as DNSBLs.
935 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
936 retry record, even if constant. There may be multiple transports with
937 different interface settings and the retry behaviour needs to be kept
940 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
942 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
944 JH/10 Bug 840: fix log_defer_output option of pipe transport
946 JH/11 Bug 830: use same host for all RCPTS of a message, even under
947 hosts_randomize. This matters a lot when combined with mua_wrapper.
949 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
950 ${quote_pgsql:<string>} operator.
952 JH/13 Bug 1708: avoid misaligned access in cached lookup.
954 JH/14 Change header file name for freeradius-client. Relevant if compiling
955 with Radius support; from the Gentoo tree and checked under Fedora.
957 JH/15 Bug 1712: Introduce $prdr_requested flag variable
959 JH/16 Bug 1714: Permit an empty string as expansion result for transport
960 option transport_filter, meaning no filtering.
962 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
964 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
965 defaults to "*" (all hosts). The variable is now available when not built
966 with TLS, default unset, mainly to enable keeping the testsuite sane.
967 If a server certificate is not supplied (via tls_certificate) an error is
968 logged, and clients will find TLS connections fail on startup. Presumably
969 they will retry in-clear.
970 Packagers of Exim are strongly encouraged to create a server certificate
971 at installation time.
973 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
974 with the $config_file variable.
976 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
977 in transport context, after the attempt, and per-recipient. The latter type
978 is per host attempted. The event data is the error message, and the errno
979 information encodes the lookup type (A vs. MX) used for the (first) host,
980 and the trailing two digits of the smtp 4xx response.
982 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
983 to write to mainlog (or rejectlog, paniclog) in the window between file
984 creation and permissions/ownership being changed. Particularly affects
985 installations where exicyclog is run as root, rather than exim user;
986 result is that the running daemon panics and dies.
988 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
990 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
991 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
992 "pri" and "weight". Note that the previous implicit priority given by the
993 list order is no longer honoured.
995 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
998 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
999 by defining SUPPORT_SOCKS.
1001 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1002 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1003 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1004 variables are renamed to proxy_{local,external}_{address,port}.
1006 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1007 for a build by defining SUPPORT_I18N
1009 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1010 of the query string, and make ${quote_redis:} do that quoting.
1012 JH/29 Move Events support from Experimental to mainline, enabled by default
1013 and removable for a build by defining DISABLE_EVENT.
1015 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1017 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1018 cached by the daemon.
1020 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1021 by defining LOOKUP_REDIS. The libhiredis library is required.
1023 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1024 keys are given for lookup.
1026 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1027 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1028 only supported when built with TLS support. The PolarSSL SHA routines
1029 are still used when the TLS library is too old for convenient support.
1031 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1032 openssl_options), for security. OpenSSL forces this from version 1.1.0
1033 server-side so match that on older versions.
1035 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1036 allocation for $value could be released as the expansion processing
1037 concluded, but leaving the global pointer active for it.
1039 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1040 and to use the domains and local_parts ACL conditions.
1042 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1043 incorrectly not doubled on cutthrough transmission, hence seen as a
1044 body-termination at the receiving system - resulting in truncated mails.
1045 Commonly the sender saw a TCP-level error, and retransmitted the message
1046 via the normal store-and-forward channel. This could result in duplicates
1047 received - but deduplicating mailstores were liable to retain only the
1048 initial truncated version.
1050 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1052 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1054 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1055 we're in there, support oversigning also; bug 1309.
1057 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1059 HS/04 Add support for keep_environment and add_environment options.
1061 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1062 either intentional arithmetic overflow during PRNG, or testing config-
1065 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1066 delivery resulted in actual delivery. Cancel cutthrough before DATA
1069 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1070 rejects a recipient: pass the reject to the originator.
1072 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1073 Many were false-positives and ignorable, but it's worth fixing the
1076 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1077 for the new environment-manipulation done at startup. Move the routines
1078 from being local to tls.c to being global via the os.c file.
1080 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1081 an extract embedded as result-arg for a map, the first arg for extract
1082 is unavailable so we cannot tell if this is a numbered or keyed
1083 extraction. Accept either.
1089 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1092 JH/02 The smtp transport option "multi_domain" is now expanded.
1094 JH/03 The smtp transport now requests PRDR by default, if the server offers
1097 JH/04 Certificate name checking on server certificates, when exim is a client,
1098 is now done by default. The transport option tls_verify_cert_hostnames
1099 can be used to disable this per-host. The build option
1100 EXPERIMENTAL_CERTNAMES is withdrawn.
1102 JH/05 The value of the tls_verify_certificates smtp transport and main options
1103 default to the word "system" to access the system default CA bundle.
1104 For GnuTLS, only version 3.0.20 or later.
1106 JH/06 Verification of the server certificate for a TLS connection is now tried
1107 (but not required) by default. The verification status is now logged by
1108 default, for both outbound TLS and client-certificate supplying inbound
1111 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1114 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1115 Status Notification (bounce) messages are now MIME format per RFC 3464.
1116 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1117 under the control of the dsn_advertise_hosts option, and routers may
1118 have a dsn_lasthop option.
1120 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1121 default, modifiable by a malware= option. The list separator for
1122 the options can now be changed in the usual way. Bug 68.
1124 JH/10 The smtp_receive_timeout main option is now expanded before use.
1126 JH/11 The incoming_interface log option now also enables logging of the
1127 local interface on delivery outgoing connections.
1129 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1130 if the interface and destination host and port all match.
1132 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1135 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1136 Patch from Andrew Lewis.
1138 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1139 now supports optional time-restrictions, weighting, and priority
1140 modifiers per server. Patch originally by <rommer@active.by>.
1142 JH/16 The spamd_address main option now supports a mixed list of local
1143 and remote servers. Remote servers can be IPv6 addresses, and
1144 specify a port-range.
1146 JH/17 Bug 68: The spamd_address main option now supports an optional
1147 timeout value per server.
1149 JH/18 Bug 1581: Router and transport options headers_add/remove can
1150 now have the list separator specified.
1152 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1155 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1158 JH/21 Support for the A6 type of dns record is withdrawn.
1160 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1161 rather than the verbs used.
1163 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1164 from 255 to 1024 chars.
1166 JH/24 Verification callouts now attempt to use TLS by default.
1168 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1169 are generic router options now. The defaults didn't change.
1171 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1172 Original patch from Alexander Shikoff, worked over by JH.
1174 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1175 files if ZCAT_COMMAND is not executable.
1177 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1179 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1181 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1182 Normally benign, it bites when the pair was led to by a CNAME;
1183 modern usage is to not canonicalize the domain to a CNAME target
1184 (and we were inconsistent anyway for A-only vs AAAA+A).
1186 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1188 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1189 when evaluating $sender_host_dnssec.
1191 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1192 $sender_helo_dnssec variable.
1194 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1196 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1198 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1200 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1201 documented as working, but never had. Support all but $spam_report.
1203 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1204 added for tls authenticator.
1206 HS/03 Add perl_taintmode main config option
1212 TL/01 When running the test suite, the README says that variables such as
1213 no_msglog_check are global and can be placed anywhere in a specific
1214 test's script, however it was observed that placement needed to be near
1215 the beginning for it to behave that way. Changed the runtest perl
1216 script to read through the entire script once to detect and set these
1217 variables, reset to the beginning of the script, and then run through
1218 the script parsing/test process like normal.
1220 TL/02 The BSD's have an arc4random API. One of the functions to induce
1221 adding randomness was arc4random_stir(), but it has been removed in
1222 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1223 function when detected.
1225 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1226 cause callback expansion.
1228 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1229 syntax errors in an expansion can be treated as a string instead of
1230 logging or causing an error, due to the internal use of bool_lax
1231 instead of bool when processing it.
1233 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1234 server certificates when making smtp deliveries.
1236 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1238 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1240 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1242 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1243 Merged patch from Sebastian Wiedenroth.
1245 JH/05 Fix results-pipe from transport process. Several recipients, combined
1246 with certificate use, exposed issues where response data items split
1247 over buffer boundaries were not parsed properly. This eventually
1248 resulted in duplicates being sent. This issue only became common enough
1249 to notice due to the introduction of connection certificate information,
1250 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1252 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1253 size buffer was used, resulting in syntax errors when an expansion
1256 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1257 version 3.3.6 or later.
1259 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1260 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1261 both become "event_action", the variables become $event_name, $event_data
1262 and $event_defer_errno. There is a new variable $verify_mode, usable in
1263 routers, transports and related events. The tls:cert event is now also
1264 raised for inbound connections, if the main configuration event_action
1267 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1268 early OCSP support, but no stapling (appears to be less than 1.0.0).
1270 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1271 server certificate names available under the smtp transport option
1272 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1275 JH/10 Time-related extraction expansions from certificates now use the main
1276 option "timezone" setting for output formatting, and are consistent
1277 between OpenSSL and GnuTLS compilations. Bug 1541.
1279 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1280 encoded parameter in the incoming message. Bug 1558.
1282 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1283 include certificate info, eximon was claiming there were spoolfile
1286 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1288 JH/14 Log delivery-related information more consistently, using the sequence
1289 "H=<name> [<ip>]" wherever possible.
1291 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1292 are problematic for Debian distribution, omit them from the release
1295 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1297 JH/16 Fix string representation of time values on 64bit time_t architectures.
1300 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1301 output list separator was used.
1306 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1307 checkers that were complaining about end of non-void function with no
1310 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1311 This was a regression introduced in 4.83 by another bugfix.
1313 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1315 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1316 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1322 TF/01 Correctly close the server side of TLS when forking for delivery.
1324 When a message was received over SMTP with TLS, Exim failed to clear up
1325 the incoming connection properly after forking off the child process to
1326 deliver the message. In some situations the subsequent outgoing
1327 delivery connection happened to have the same fd number as the incoming
1328 connection previously had. Exim would try to use TLS and fail, logging
1329 a "Bad file descriptor" error.
1331 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1332 utilities have not been installed.
1334 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1335 temporary space as the ACL may create new global variables.
1337 TL/01 LDAP support uses per connection or global context settings, depending
1338 upon the detected version of the libraries at build time.
1340 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1341 to extract and use the src ip:port in logging and expansions as if it
1342 were a direct connection from the outside internet. PPv2 support was
1343 updated based on HAProxy spec change in May 2014.
1345 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1347 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1348 Properly escape header and check for NULL return.
1350 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1353 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1355 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1356 characters in header names, implemented as a verify condition.
1357 Contributed by Michael Fischer v. Mollard.
1359 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1360 results permerror and temperror. Previous values are deprecated but
1361 still accepted. In a future release, err_perm and err_temp will be
1362 completely removed, which will be a backward incompatibility if the
1363 ACL tests for either of these two old results. Patch contributed by
1364 user bes-internal on the mailing list.
1366 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1368 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1369 selectors, in both main and reject logs.
1371 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1374 JH/07 Add malware type "sock" for talking to simple daemon.
1376 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1378 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1379 routers/transports under cutthrough routing.
1381 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1382 numbers. Touch up "bool" conditional to keep the same definition.
1384 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1386 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1388 JH/12 Expand items in router/transport headers_add or headers_remove lists
1389 individually rather than the list as a whole. Bug 1452.
1391 Required for reasonable handling of multiple headers_ options when
1392 they may be empty; requires that headers_remove items with embedded
1393 colons must have them doubled (or the list-separator changed).
1395 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1396 view the policy declared in the DMARC record. Currently, $dmarc_status
1397 is a combined value of both the record presence and the result of the
1400 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1402 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1403 dnslookup router and the smtp transport (applying to the forward
1406 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1407 of ldap servers used for a specific lookup. Patch provided by Heiko
1410 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1411 New variable $lookup_dnssec_authenticated for observability.
1413 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1414 Patch submitted by Lars Timman.
1416 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1418 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1419 Requires trusted mode and valid format message id, aborts otherwise.
1420 Patch contributed by Heiko Schlichting.
1422 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1423 certextract with support for various fields. Bug 1358.
1425 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1426 is requested by default, modifiable by smtp transport option
1429 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1430 operate on certificate variables to give certificate fingerprints
1431 Also new ${sha256:cert_variable}.
1433 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1435 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1436 Christian Aistleitner.
1438 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1440 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1441 file. Patch from Wolfgang Breyha.
1443 JH/25 Expand the coverage of the delivery $host and $host_address to
1444 client authenticators run in verify callout. Bug 1476.
1446 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1447 align with daemon_smtp_ports. Bug 72.
1449 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1450 support and error reporting did not work properly.
1452 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1453 and is readable. Patch from Andrew Colin Kissa.
1455 TL/14 Enhance documentation of ${run expansion and how it parses the
1456 commandline after expansion, particularly in the case when an
1457 unquoted variable expansion results in an empty value.
1459 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1461 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1462 by renaming away from it. Addresses GH issue 15, reported by
1465 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1467 TL/15 SECURITY: prevent double expansion in math comparison functions
1468 (can expand unsanitized data). Not remotely exploitable.
1475 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1477 PP/02 Make -n do something, by making it not do something.
1478 When combined with -bP, the name of an option is not output.
1480 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1483 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1484 $sender_host_name and config options to manage this, and basic check
1487 PP/05 DSCP support for outbound connections and control modifier for inbound.
1489 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1490 (Only plugin which currently uses this is kerberos4, which nobody should
1491 be using, but we should make it available and other future plugins might
1492 conceivably use it, even though it would break NAT; stuff *should* be
1493 using channel bindings instead).
1495 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1496 name; added for Sendmail compatibility; requires admin caller.
1497 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1498 just ignore it); requires trusted caller.
1499 Also parse but ignore: -Ac -Am -X<logfile>
1502 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1504 TL/02 Add +smtp_confirmation as a default logging option.
1506 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1507 Patch by Magnus Holmgren from 2007-02-20.
1509 TL/04 Bugzilla 1281 - Spec typo.
1510 Bugzilla 1283 - Spec typo.
1511 Bugzilla 1290 - Spec grammar fixes.
1513 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1515 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1517 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1518 mailing list by by Dmitry Isaikin.
1520 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1522 JH/02 Support "G" suffix to numbers in ${if comparisons.
1524 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1526 NM/01 Bugzilla 1197 - Spec typo
1527 Bugzilla 1196 - Spec examples corrections
1529 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1531 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1532 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1535 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1536 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1538 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1539 "acl {{name}{arg}...}", and optional args on acl condition
1542 JH/05 Permit multiple router/transport headers_add/remove lines.
1544 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1546 JH/07 Avoid using a waiting database for a single-message-only transport.
1547 Performance patch from Paul Fisher. Bugzilla 1262.
1549 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1552 JH/09 Add $headers_added variable, with content from use of ACL modifier
1553 add_header (but not yet added to the message). Bugzilla 199.
1555 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1556 Pulled from Bugzilla 817 by Wolfgang Breyha.
1558 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1560 (nb: this is the same fix as in Exim 4.80.1)
1562 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1565 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1568 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1570 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1571 advertise SMTP AUTH mechanism to us, instead of a generic
1572 protocol violation error. Also, make Exim more robust to bad
1573 data from the Dovecot auth socket.
1575 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1577 When a queue runner is handling a message, Exim first routes the
1578 recipient addresses, during which it prunes them based on the retry
1579 hints database. After that it attempts to deliver the message to
1580 any remaining recipients. It then updates the hints database using
1583 So if a recipient address works intermittently, it can get repeatedly
1584 deferred at routing time. The retry hints record remains fresh so the
1585 address never reaches the final cutoff time.
1587 This is a fairly common occurrence when a user is bumping up against
1588 their storage quota. Exim had some logic in its local delivery code
1589 to deal with this. However it did not apply to per-recipient defers
1590 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1592 This change adds a proper retry rule check during routing so that the
1593 final cutoff time is checked against the message's age. We only do
1594 this check if there is an address retry record and there is not a
1595 domain retry record; this implies that previous attempts to handle
1596 the address had the retry_use_local_parts option turned on. We use
1597 this as an approximation for the destination being like a local
1598 delivery, as in LMTP.
1600 I suspect this new check makes the old local delivery cutoff check
1601 redundant, but I have not verified this so I left the code in place.
1603 TF/02 Correct gecos expansion when From: is a prefix of the username.
1605 Test 0254 submits a message to Exim with the header
1609 When I ran the test suite under the user fanf2, Exim expanded
1610 the header to contain my full name, whereas it should have added
1611 a Resent-Sender: header. It erroneously treats any prefix of the
1612 username as equal to the username.
1614 This change corrects that bug.
1616 GF/01 DCC debug and logging tidyup
1617 Error conditions log to paniclog rather than rejectlog.
1618 Debug lines prefixed by "DCC: " to remove any ambiguity.
1620 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1622 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1623 Bug spotted by Jeremy Harris; was flawed since initial commit.
1624 Would have resulted in OCSP responses post-SNI triggering an Exim
1625 NULL dereference and crash.
1627 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1629 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1630 Bug detection, analysis and fix by Samuel Thibault.
1631 Bugzilla 1331, Debian bug #698092.
1633 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1635 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1636 Server implementation by Todd Lyons, client by JH.
1637 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1638 config variable "prdr_enable" controls whether the server
1639 advertises the facility. If the client requests PRDR a new
1640 acl_data_smtp_prdr ACL is called once for each recipient, after
1641 the body content is received and before the acl_smtp_data ACL.
1642 The client is controlled by both of: a hosts_try_prdr option
1643 on the smtp transport, and the server advertisement.
1644 Default client logging of deliveries and rejections involving
1645 PRDR are flagged with the string "PRDR".
1647 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1648 fclose(). Diagnosis by Todd Lyons.
1650 PP/17 Update configure.default to handle IPv6 localhost better.
1651 Patch by Alain Williams (plus minor tweaks).
1654 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1655 This is now consistent with GnuTLS, and is now documented: the
1656 previous undocumented portable approach to treating the option as
1657 unset was to force an expansion failure. That still works, and
1658 an empty string is now equivalent.
1660 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1661 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1662 not performing validation itself.
1664 PP/20 Added force_command boolean option to pipe transport.
1665 Patch from Nick Koston, of cPanel Inc.
1667 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1670 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1672 PP/21 Fix eximon continuous updating with timestamped log-files.
1673 Broken in a format-string cleanup in 4.80, missed when I repaired the
1674 other false fix of the same issue.
1675 Report and fix from Heiko Schlichting.
1678 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1679 Report from Prashanth Katuri.
1681 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1682 It's SecureTransport, so affects any MacOS clients which use the
1683 system-integrated TLS libraries, including email clients.
1685 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1686 using a MIME ACL for non-SMTP local injection.
1687 Report and assistance in diagnosis by Warren Baker.
1689 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1691 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1693 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1694 last id that failed so it may be referenced in subsequent ACL's.
1696 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1699 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1700 ldap library initialization, allowing self-signed CA's to be
1701 used. Also properly sets require_cert option later in code by
1702 using NULL (global ldap config) instead of ldap handle (per
1703 session). Bug diagnosis and testing by alxgomz.
1705 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1706 the src/util/ subdirectory.
1708 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1709 renamed to Transport Post Delivery Action by Jeremy Harris, as
1712 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1713 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1714 redis_servers = needs to be configured which will be used by the redis
1715 lookup. Patch from Warren Baker, of The Packet Hub.
1717 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1719 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1720 hostname or reverse DNS when processing a host list. Used suggestions
1721 from multiple comments on this bug.
1723 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1725 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1726 Missed a few lines, added it to make the runtest require no keyboard
1729 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1730 contains upper case chars. Make router use caseful_local_part.
1732 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1733 support when GnuTLS has been built with p11-kit.
1739 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1741 This, or similar/improved, will also be change PP/11 of 4.82.
1747 PP/01 Handle short writes when writing local log-files.
1748 In practice, only affects FreeBSD (8 onwards).
1749 Bugzilla 1053, with thanks to Dmitry Isaikin.
1751 NM/01 Bugzilla 949 - Documentation tweak
1753 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1756 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1758 PP/02 Implemented gsasl authenticator.
1760 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1762 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1763 `pkg-config foo` for cflags/libs.
1765 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1766 with rest of GSASL and with heimdal_gssapi.
1768 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1769 `pkg-config foo` for cflags/libs for the TLS implementation.
1771 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1772 properties get this fed in as external SSF. A number of robustness
1773 and debugging improvements to the cyrus_sasl authenticator.
1775 PP/08 cyrus_sasl server now expands the server_realm option.
1777 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1778 Patch by Jeremy Harris.
1780 PP/10 Added dbmjz lookup type.
1782 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1784 PP/12 MAIL args handles TAB as well as SP, for better interop with
1785 non-compliant senders.
1786 Analysis and variant patch by Todd Lyons.
1788 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1789 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1790 Patch from Dirk Mueller <dmueller@suse.com>
1792 PP/13 tls_peerdn now print-escaped for spool files.
1793 Observed some $tls_peerdn in wild which contained \n, which resulted
1794 in spool file corruption.
1796 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1797 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1798 or write after TLS renegotiation, which otherwise led to messages
1801 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1802 as a tracking header (ie: a signed header comes before the signature).
1803 Patch from Wolfgang Breyha.
1805 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1806 comma-sep list; embedded commas doubled.
1808 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1810 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1812 Report and patch from Dmitry Banschikov.
1814 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1815 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1816 failures. We appear to now support TLS1.1+ with Exim.
1818 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1819 lets Exim select keys and certificates based upon TLS SNI from client.
1820 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1821 before an outbound SMTP session. New log_selector, +tls_sni.
1823 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1824 NULL dereference. Report and patch from Alun Jones.
1826 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1827 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1828 Not seeing resolver debug output on NetBSD, but suspect this is a
1829 resolver implementation change.
1831 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1832 Left warnings. Added "eximon gdb" invocation mode.
1834 PP/21 Defaulting "accept_8bitmime" to true, not false.
1836 PP/22 Added -bw for inetd wait mode support.
1838 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1839 locate the relevant includes and libraries. Made this the default.
1841 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1842 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1844 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1845 This may cause build issues on older platforms.
1847 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1848 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1849 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1850 Added SNI support via GnuTLS too.
1851 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1853 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1855 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1856 Applied second patch from Janne, implementing suggestion to default
1857 multiple-strings-in-record handling to match SPF spec.
1859 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1861 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1862 read-only, out of scope).
1863 Patch from Wolfgang Breyha, report from Stuart Northfield.
1865 PP/29 Fix three issues highlighted by clang analyser static analysis.
1866 Only crash-plausible issue would require the Cambridge-specific
1867 iplookup router and a misconfiguration.
1868 Report from Marcin Mirosław.
1870 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1872 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1873 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1874 As part of this, removing so much warning spew let me fix some minor
1875 real issues in debug logging.
1877 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1878 assignment on my part. Fixed.
1880 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1881 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1882 Janne Snabb (who went above and beyond: thank you).
1884 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1885 string otherwise requires a connection and a bunch more work and it's
1886 relatively easy to get wrong. Should also expose TLS library linkage
1889 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1890 64-bit ${eval} (JH/03).
1892 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1893 GNU libc to support some of the 64-bit stuff, should not lead to
1894 conflicts. Defined before os.h is pulled in, so if a given platform
1895 needs to override this, it can.
1897 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1898 protection layer was required, which is not implemented.
1899 Bugzilla 1254, patch from Wolfgang Breyha.
1901 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1902 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1903 tls_dhparam take prime identifiers. Also unbreak combination of
1904 OpenSSL+DH_params+TLSSNI.
1906 PP/39 Disable SSLv2 by default in OpenSSL support.
1912 PP/01 Solaris build fix for Oracle's LDAP libraries.
1913 Bugzilla 1109, patch from Stephen Usher.
1915 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1917 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1920 TF/02 Fix a couple more cases where we did not log the error message
1921 when unlink() failed. See also change 4.74-TF/03.
1923 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1924 lock up or crash if it happened to be inside a call to libc when it
1925 got a SIGUSR1 from exiwhat.
1927 The SIGUSR1 handler appends the current process status to the process
1928 log which is later printed by exiwhat. It used to use the general
1929 purpose logging code to do this, but several functions it calls are
1930 not safe for signals.
1932 The new output code in the SIGUSR1 handler is specific to the process
1933 log, and simple enough that it's easy to inspect for signal safety.
1934 Removing some special cases also simplifies the general logging code.
1935 Removing the spurious timestamps from the process log simplifies
1938 TF/04 Improved ratelimit ACL condition.
1940 The /noupdate option has been deprecated in favour of /readonly which
1941 has clearer semantics. The /leaky, /strict, and /readonly update modes
1942 are mutually exclusive. The update mode is no longer included in the
1943 database key; it just determines when the database is updated. (This
1944 means that when you upgrade Exim will forget old rate measurements.)
1946 Exim now checks that the per_* options are used with an update mode that
1947 makes sense for the current ACL. For example, when Exim is processing a
1948 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1949 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1950 must specify per_mail/readonly. If you omit the update mode it defaults to
1951 /leaky where that makes sense (as before) or /readonly where required.
1953 The /noupdate option is now undocumented but still supported for
1954 backwards compatibility. It is equivalent to /readonly except that in
1955 ACLs where /readonly is required you may specify /leaky/noupdate or
1956 /strict/noupdate which are treated the same as /readonly.
1958 A useful new feature is the /count= option. This is a generalization
1959 of the per_byte option, so that you can measure the throughput of other
1960 aggregate values. For example, the per_byte option is now equivalent
1961 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1963 The per_rcpt option has been generalized using the /count= mechanism
1964 (though it's more complicated than the per_byte equivalence). When it is
1965 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1966 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1967 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1968 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1969 Note that using per_rcpt with a non-readonly update mode in more than
1970 one ACL will cause the recipients to be double-counted. (The per_mail
1971 and per_byte options don't have this problem.)
1973 The handling of very low rates has changed slightly. If the computed rate
1974 is less than the event's count (usually one) then this event is the first
1975 after a long gap. In this case the rate is set to the same as this event's
1976 count, so that the first message of a spam run is counted properly.
1978 The major new feature is a mechanism for counting the rate of unique
1979 events. The new per_addr option counts the number of different
1980 recipients that someone has sent messages to in the last time period. It
1981 behaves like per_rcpt if all the recipient addresses are different, but
1982 duplicate recipient addresses do not increase the measured rate. Like
1983 the /count= option this is a general mechanism, so the per_addr option
1984 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1985 example, measure the rate that a client uses different sender addresses
1986 with the options per_mail/unique=$sender_address. There are further
1987 details in the main documentation.
1989 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1991 TF/06 Removed a few PCRE remnants.
1993 TF/07 Automatically extract Exim's version number from tags in the git
1994 repository when doing development or release builds.
1996 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1997 Bugzilla 879. Patch from Paul Fisher.
1999 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2000 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2003 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2005 PP/05 Variable $av_failed, true if the AV scanner deferred.
2006 Bugzilla 1078. Patch from John Horne.
2008 PP/06 Stop make process more reliably on build failure.
2009 Bugzilla 1087. Patch from Heiko Schlittermann.
2011 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2012 Bugzilla 1089. Patch from Heiko Schlittermann.
2014 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2015 Bugzilla 1131. Patch from Holger Weiß.
2017 PP/09 Handle IPv6 addresses with SPF.
2018 Bugzilla 860. Patch from Wolfgang Breyha.
2020 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2022 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2025 PP/11 match_* no longer expand right-hand-side by default.
2026 New compile-time build option, EXPAND_LISTMATCH_RHS.
2027 New expansion conditions, "inlist", "inlisti".
2029 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2031 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2033 PP/14 fix log_write() format string regression from TF/03.
2034 Bugzilla 1152. Patch from Dmitry Isaikin.
2040 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2042 PP/02 Harmonised TLS library version reporting; only show if debugging.
2043 Layout now matches that introduced for other libraries in 4.74 PP/03.
2045 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2047 PP/04 New "dns_use_edns0" global option.
2049 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2052 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2053 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2055 TK/01 Updated PolarSSL code to 0.14.2.
2056 Bugzilla 1097. Patch from Andreas Metzler.
2058 PP/07 Catch divide-by-zero in ${eval:...}.
2059 Fixes bugzilla 1102.
2061 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2064 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2065 format-string attack -- SECURITY: remote arbitrary code execution.
2067 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2068 time unintentionally subject to list matching rules, letting the header
2069 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2070 arbitrary string expansion). This allowed for information disclosure.
2072 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2073 INT_MIN/-1 -- value coerced to INT_MAX.
2079 NM/01 Workaround for PCRE version dependency in version reporting
2082 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2083 This fixes portability to compilers other than gcc, notably
2084 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2086 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2087 makefiles for portability to HP-UX and POSIX correctness.
2089 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2090 Also via indented variable definition in the Makefile.
2091 (Debugging by Oliver Heesakkers).
2093 PP/02 Restore caching of spamd results with expanded spamd_address.
2094 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2096 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2097 Improves build reliability. Fix from: Frank Elsner
2099 NM/02 Fix wide character breakage in the rfc2047 coding
2100 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2102 NM/03 Allow underscore in dnslist lookups
2103 Fixes bug 1026. Patch from Graeme Fowler
2105 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2106 Code patches from Adam Ciarcinski of NetBSD.
2108 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2111 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2112 is logged, to avoid truncation. Patch from John Horne.
2114 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2115 Patch from Jakob Hirsch.
2117 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2118 SQL string expansion failure details.
2119 Patch from Andrey Oktyabrski.
2121 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2122 Patch from Simon Arlott.
2124 PP/09 New lookups functionality failed to compile on old gcc which rejects
2125 extern declarations in function scope.
2126 Patch from Oliver Fleischmann
2128 PP/10 Use sig_atomic_t for flags set from signal handlers.
2129 Check getgroups() return and improve debugging.
2130 Fixed developed for diagnosis in bug 927 (which turned out to be
2133 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2134 Patch from Mark Zealey.
2136 PP/12 Bugzilla 1056: Improved spamd server selection.
2137 Patch from Mark Zealey.
2139 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2140 Based on patch from Heiko Schlittermann.
2142 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2143 Patch from Uwe Doering, sign-off by Michael Haardt.
2145 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2146 variable declaration deep within a block. Bug and patch from
2149 PP/15 lookups-Makefile IRIX compatibility coercion.
2151 PP/16 Make DISABLE_DKIM build knob functional.
2153 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2154 Patch by Simon Arlott
2156 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2157 variable argument macros. Our copy now differs from upstream.
2163 TF/01 Failure to get a lock on a hints database can have serious
2164 consequences so log it to the panic log.
2166 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2167 controlled using the smtp_confirmation log selector.
2169 TF/03 Include the error message when we fail to unlink a spool file.
2171 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2172 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2173 for maintaining out-of-tree patches for some time.
2175 PP/01 Bugzilla 139: Documentation and portability issues.
2176 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2177 Handle per-OS dynamic-module compilation flags.
2179 PP/02 Let /dev/null have normal permissions.
2180 The 4.73 fixes were a little too stringent and complained about the
2181 permissions on /dev/null. Exempt it from some checks.
2182 Reported by Andreas M. Kirchwitz.
2184 PP/03 Report version information for many libraries, including
2185 Exim version information for dynamically loaded libraries. Created
2186 version.h, now support a version extension string for distributors
2187 who patch heavily. Dynamic module ABI change.
2189 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2190 privilege escalation vulnerability whereby the Exim run-time user
2191 can cause root to append content of the attacker's choosing to
2194 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2197 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2198 If dropping privileges for untrusted macros, we disabled normal logging
2199 on the basis that it would fail; for the Exim run-time user, this is not
2200 the case, and it resulted in successful deliveries going unlogged.
2201 Fixed. Reported by Andreas Metzler.
2207 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2208 only prepend for the Resent-* case. Fixes regression introduced in
2209 Exim 4.70 by NM/22 for Bugzilla 607.
2211 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2212 increasing numbers of administrators be bitten by this.
2214 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2216 PP/03 Bugzilla 994: added openssl_options main configuration option.
2218 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2220 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2222 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2224 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2225 without a peer certificate, leading to a segfault because of an
2226 assumption that peers always have certificates. Be a little more
2227 paranoid. Problem reported by Martin Tscholak.
2229 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2230 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2231 NB: ClamAV planning to remove STREAM in "middle of 2010".
2232 CL also introduces -bmalware, various -d+acl logging additions and
2233 more caution in buffer sizes.
2235 PP/09 Implemented reverse_ip expansion operator.
2237 PP/10 Bugzilla 937: provide a "debug" ACL control.
2239 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2241 PP/12 Bugzilla 973: Implement --version.
2243 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2245 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2247 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2249 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2250 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2251 ignore trailing whitespace.
2253 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2255 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2258 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2259 Notification from Dr Andrew Aitchison.
2261 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2262 ExtendedDetectionInfo response format.
2263 Notification from John Horne.
2265 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2268 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2269 XSL and documented dependency on system catalogs, with examples of how
2272 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2275 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2276 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2277 configuration file which is writeable by the Exim user or group.
2279 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2280 of configuration files to cover files specified with the -C option if
2281 they are going to be used with root privileges, not just the default
2284 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2285 option (effectively making it always true).
2287 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2288 files to be used while preserving root privileges.
2290 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2291 that rogue child processes cannot use them.
2293 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2294 run-time user, instead of root.
2296 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2297 Exim run-time user without dropping privileges.
2299 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2300 result string, instead of calling string_vformat() twice with the same
2303 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2304 for other users. Others should always drop root privileges if they use
2305 -C on the command line, even for a whitelisted configure file.
2307 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2309 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2315 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2316 $data_path, and $header_path variables; fixed documentation bugs and
2319 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2320 exipick to access non-standard spools, including the "frozen" queue
2323 NM/01 Bugzilla 965: Support mysql stored procedures.
2324 Patch from Alain Williams
2326 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2328 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2329 Patch from Andreas Metzler
2331 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2332 Patch from Kirill Miazine
2334 NM/05 Bugzilla 671: Added umask to procmail example.
2336 JJ/03 installed exipick 20100323.0, fixing doc bug
2338 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2339 directory. Notification and patch from Dan Rosenberg.
2341 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2343 TK/02 Improve log output when DKIM signing operation fails.
2345 MH/01 Treat the transport option dkim_domain as a colon separated
2346 list, not as a single string, and sign the message with each element,
2347 omitting multiple occurences of the same signer.
2349 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2350 Bugzilla 985, 986. Patch by Simon Arlott
2352 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2353 Patch by Simon Arlott
2355 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2356 MBX locking. Notification from Dan Rosenberg.
2362 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2364 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2366 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2368 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2370 NM/04 Bugzilla 915: Flag broken perl installation during build.
2376 TK/01 Added patch by Johannes Berg that expands the main option
2377 "spamd_address" if it starts with a dollar sign.
2379 TK/02 Write list of recipients to X-Envelope-Sender header when building
2380 the mbox-format spool file for content scanning (suggested by Jakob
2383 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2384 (http://www.dcc-servers.net/) support via dccifd. Activated by
2385 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2387 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2388 by Mark Daniel Reidel <mr@df.eu>.
2390 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2391 When building exim an external PCRE library is now needed -
2392 PCRE is a system library on the majority of modern systems.
2393 See entry on PCRE_LIBS in EDITME file.
2395 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2396 conversation. Added nologin parameter to request.
2397 Patch contributed by Kirill Miazine.
2399 TF/01 Do not log submission mode rewrites if they do not change the address.
2401 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2403 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2404 log files in place. Contributed by Roberto Lima.
2406 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2408 TF/03 Bugzilla 615: When checking the local_parts router precondition
2409 after a local_part_suffix or local_part_prefix option, Exim now
2410 does not use the address's named list lookup cache, since this
2411 contains cached lookups for the whole local part.
2413 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2414 Robert Millan. Documentation is in experimental-spec.txt.
2416 TF/04 Bugzilla 668: Fix parallel build (make -j).
2418 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2420 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2421 Patch provided by Jan Srzednicki.
2423 TF/05 Leading white space used to be stripped from $spam_report which
2424 wrecked the formatting. Now it is preserved.
2426 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2427 that they are available at delivery time.
2429 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2431 TF/08 TLS error reporting now respects the incoming_interface and
2432 incoming_port log selectors.
2434 TF/09 Produce a more useful error message if an SMTP transport's hosts
2435 setting expands to an empty string.
2437 NM/06 Bugzilla 744: EXPN did not work under TLS.
2438 Patch provided by Phil Pennock.
2440 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2441 Patch provided by Richard Godbee.
2443 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2444 acl_smtp_notquit, added index entry.
2446 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2447 Patch provided by Eugene Bujak.
2449 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2450 accept(). Patch provided by Maxim Dounin.
2452 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2453 Patch provided by Phil Pennock.
2455 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2457 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2458 Patch provided by Brad "anomie" Jorsch.
2460 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2461 Patch provided by Dean Brooks.
2463 TK/05 Add native DKIM support (does not depend on external libraries).
2465 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2466 Patch provided by Graeme Fowler.
2468 NM/16 Bugzilla 851: Documentation example syntax fix.
2470 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2472 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2475 NM/19 Bugzilla 745: TLS version reporting.
2476 Patch provided by Phil Pennock.
2478 NM/20 Bugzilla 167: bool: condition support.
2479 Patch provided by Phil Pennock.
2481 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2482 clients. Patch provided by Phil Pennock.
2484 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2485 Patch provided by Brad "anomie" Jorsch.
2487 NM/23 Bugzilla 687: Fix misparses in eximstats.
2488 Patch provided by Heiko Schlittermann.
2490 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2491 Patch provided by Heiko Schlittermann.
2493 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2494 plus update to original patch.
2496 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2498 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2499 Patch provided by David Brownlee.
2501 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2503 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2505 NM/30 Bugzilla 888: TLS documentation bugfixes.
2507 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2509 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2510 Unlike the original bugzilla I have changed all shell scripts in src tree.
2512 NM/33 Bugzilla 898: Transport filter timeout fix.
2513 Patch by Todd Rinaldo.
2515 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2516 Patch by Serge Demonchaux.
2518 NM/35 Bugzilla 39: Base64 decode bug fixes.
2519 Patch by Jakob Hirsch.
2521 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2523 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2525 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2527 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2533 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2534 ALT-N's libdkim that I have put here:
2535 http://duncanthrax.net/exim-experimental/
2537 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2538 were called 'true' and it seems that C99 defines that as a reserved
2539 keyword to be used with 'bool' variable types. That means you could
2540 not include C99-style headers which use bools without triggering
2541 build errors in sieve.c.
2543 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2544 as mailq or other aliases. Changed the --help handling significantly
2545 to do whats expected. exim_usage() emits usage/help information.
2547 SC/01 Added the -bylocaldomain option to eximstats.
2549 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2551 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2553 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2559 PH/01 Another patch from the Sieve maintainer.
2561 PH/02 When an IPv6 address is converted to a string for single-key lookup
2562 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2563 dots are used instead of colons so that keys in lsearch files need not
2564 contain colons. This was done some time before quoting was made available
2565 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2566 (notated using the quote facility) so as to distinguish them from IPv4
2567 keys. This meant that lookups for IP addresses in host lists did not work
2568 for iplsearch lookups.
2570 This has been fixed by arranging for IPv6 addresses to be expressed with
2571 colons if the lookup type is iplsearch. This is not incompatible, because
2572 previously such lookups could never work.
2574 The situation is now rather anomalous, since one *can* have colons in
2575 ordinary lsearch keys. However, making the change in all cases is
2576 incompatible and would probably break a number of configurations.
2578 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2581 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2582 conversion specification without a maximum field width, thereby enabling
2583 a rogue spamd server to cause a buffer overflow. While nobody in their
2584 right mind would setup Exim to query an untrusted spamd server, an
2585 attacker that gains access to a server running spamd could potentially
2586 exploit this vulnerability to run arbitrary code as the Exim user.
2588 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2589 $primary_hostname instead of what libspf2 thinks the hosts name is.
2591 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2592 a directory entry by the name of the lookup key. Previously, if a
2593 symlink pointed to a non-existing file or a file in a directory that
2594 Exim lacked permissions to read, a lookup for a key matching that
2595 symlink would fail. Now it is enough that a matching directory entry
2596 exists, symlink or not. (Bugzilla 503.)
2598 PH/03 The body_linecount and body_zerocount variables are now exported in the
2601 PH/04 Added the $dnslist_matched variable.
2603 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2604 This means they are set thereafter only if the connection becomes
2607 PH/06 Added the client_condition to authenticators so that some can be skipped
2608 by clients under certain conditions.
2610 PH/07 The error message for a badly-placed control=no_multiline_responses left
2611 "_responses" off the end of the name.
2613 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2615 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2616 (without spaces) instead of just copying the configuration text.
2618 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2620 PH/11 Added $max_received_linelength.
2622 PH/12 Added +ignore_defer and +include_defer to host lists.
2624 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2625 way in which PCRE > 7.0 is built.
2627 PH/14 Implemented queue_only_load_latch.
2629 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2630 MAIL command. The effect was to mangle the value on 64-bit systems.
2632 PH/16 Another patch from the Sieve maintainer.
2634 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2636 PH/18 If a system quota error occurred while trying to create the file for
2637 a maildir delivery, the message "Mailbox is full" was not appended to the
2638 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2639 only to a quota excession during the actual writing of the file.
2641 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2642 characters?) which causes problems in log lines. The DN values are now
2643 passed through string_printing() before being added to log lines.
2645 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2646 and InterBase are left for another time.)
2648 PH/21 Added message_body_newlines option.
2650 PH/22 Guard against possible overflow in moan_check_errorcopy().
2652 PH/23 POSIX allows open() to be a macro; guard against that.
2654 PH/24 If the recipient of an error message contained an @ in the local part
2655 (suitably quoted, of course), incorrect values were put in $domain and
2656 $local_part during the evaluation of errors_copy.
2662 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2663 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2666 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2667 issue a MAIL command.
2669 PH/02 In an ACL statement such as
2671 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2673 if a client was not listed at all, or was listed with a value other than
2674 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2675 the condition was not true (as it should be), so access was not denied.
2676 The bug was that the ! inversion was incorrectly passed on to the second
2677 item. This has been fixed.
2679 PH/03 Added additional dnslists conditions == and =& which are different from
2680 = and & when the dns lookup returns more than one IP address.
2682 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2683 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2685 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2686 FSYNC, which compiles an option called disable_fsync that allows for
2687 bypassing fsync(). The documentation is heavily laced with warnings.
2689 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2691 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2692 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2693 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2694 including adding "make clean"; (3) Added -fPIC when compiling the test
2695 dynamically loaded module, to get rid of a warning.
2697 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2698 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2699 = 0s. The bug is otherwise harmless.
2701 PH/07 There was a bug in the dovecot authenticator such that the value of
2702 $auth1 could be overwritten, and so not correctly preserved, after a
2703 successful authentication. This usually meant that the value preserved by
2704 the server_setid option was incorrect.
2706 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2708 PH/09 Installed PCRE release 7.0.
2710 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2711 run for batched SMTP input. It is now run at the start of every message
2712 in the batch. While fixing this I discovered that the process information
2713 (output by running exiwhat) was not always getting set for -bs and -bS
2714 input. This is fixed, and it now also says "batched" for BSMTP.
2716 PH/11 Added control=no_pipelining.
2718 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2719 patch, slightly modified), and move the expansion of helo_data till after
2720 the connection is made in the smtp transport (so it can use these
2723 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2725 PH/14 Added log_selector = +pid.
2727 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2729 PH/16 Add ${if forany and ${if forall.
2731 PH/17 Added dsn_from option to vary the From: line in DSNs.
2733 PH/18 Flush SMTP output before performing a callout, unless control =
2734 no_callout_flush is set.
2736 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2737 was true (the default) a successful delivery failed to delete the retry
2738 item, thus causing premature timeout of the address. The bug is now
2741 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2743 PH/21 Long custom messages for fakedefer and fakereject are now split up
2744 into multiline responses in the same way that messages for "deny" and
2745 other ACL rejections are.
2747 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2748 with slight modification.
2750 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2751 draft, changing the syntax and factoring some duplicate code".
2753 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2754 for deliveries of the second and subsequent messages over the same SMTP
2757 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2758 ${reduce, with only minor "tidies".
2760 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2762 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2763 expansion side effects.
2765 PH/27 When a message times out after an over-quota error from an Exim-imposed
2766 quota, the bounce message says "mailbox is full". This message was not
2767 being given when it was a system quota that was exceeded. It now should
2770 MH/03 Made $recipients available in local_scan(). local_scan() already has
2771 better access to the recipient list through recipients_list[], but
2772 $recipients can be useful in postmaster-provided expansion strings.
2774 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2775 in the case of a MAIL command with additional options following the
2776 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2777 were accidentally chopped off.
2779 PH/29 SMTP synchronization checks are implemented when a command is read -
2780 there is a check that no more input is waiting when there shouldn't be
2781 any. However, for some commands, a delay in an ACL can mean that it is
2782 some time before the response is written. In this time, more input might
2783 arrive, invalidly. So now there are extra checks after an ACL has run for
2784 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2785 pipelining has not been advertised.
2787 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2789 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2790 returned in the SMTP response when smtp_return_error_details was false.
2791 This has been fixed.
2793 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2794 instead of the C I/O that was originally supplied, because problems were
2795 reported on Solaris.
2797 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2798 Exim which did not show up earlier: it was assuming that a call to
2799 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2800 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2801 was a macro that became an assignment, so it seemed to work. This has
2802 changed to a proper function call with a void return, hence the compile
2803 error. Exim's code has been fixed.
2805 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2808 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2810 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2812 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2813 because it turns out to be too big or there is a timeout) there is no
2814 "Completed" line in the log. When some messages of this type were
2815 selected by exigrep, they were listed as "not completed". Others were
2816 picked up by some special patterns. I have improved the selection
2817 criteria to be more general.
2819 PH/38 The host_find_failed option in the manualroute router can now be set
2820 to "ignore", to completely ignore a host whose IP address cannot be
2821 found. If all hosts are ignored, the behaviour is controlled by the new
2822 host_all_ignored option.
2824 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2825 homing or because of multiple MX records with /mx) generated more than
2826 one IP address, and the following item turned out to be the local host,
2827 all the secondary addresses of the first item were incorrectly removed
2828 from the list, along with the local host and any following hosts (which
2829 is what is supposed to happen).
2831 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2832 whoever called Exim into the -H file. In the case of the daemon it was
2833 behaving confusingly. When first started, it used values for whoever
2834 started the daemon, but after a SIGHUP it used the Exim user (because it
2835 calls itself on a restart). I have changed the code so that it now always
2838 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2839 message are rejected with the same error (e.g. no authentication or bad
2840 sender address), and a DATA command is nevertheless sent (as can happen
2841 with PIPELINING or a stupid MUA), the error message that was given to the
2842 RCPT commands is included in the rejection of the DATA command. This is
2843 intended to be helpful for MUAs that show only the final error to their
2846 PH/42 Another patch from the Sieve maintainer.
2848 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2849 Eximstats - Fixed some broken HTML links and added missing column headers
2851 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2852 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2854 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2856 PH/43 Yet another patch from the Sieve maintainer.
2858 PH/44 I found a way to check for a TCP/IP connection going away before sending
2859 the response to the final '.' that terminates a message, but only in the
2860 case where the client has not sent further data following the '.'
2861 (unfortunately, this is allowed). However, in many cases there won't be
2862 any further data because there won't be any more messages to send. A call
2863 to select() can be used: if it shows that the input is "ready", there is
2864 either input waiting, or the socket has been closed. An attempt to read
2865 the next input character can distinguish the two cases. Previously, Exim
2866 would have sent an OK response which the client would never have see.
2867 This could lead to message repetition. This fix should cure that, at
2868 least in a lot of common cases.
2870 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2871 advertised in response to EHLO.
2877 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2878 fixed by 4.65/MH/01 (is this a record?) are fixed:
2880 (i) An empty string was always treated as zero by the numeric comparison
2881 operators. This behaviour has been restored.
2883 (ii) It is documented that the numeric comparison operators always treat
2884 their arguments as decimal numbers. This was broken in that numbers
2885 starting with 0 were being interpreted as octal.
2887 While fixing these problems I realized that there was another issue that
2888 hadn't been noticed. Values of message_size_limit (both the global option
2889 and the transport option) were treated as octal if they started with 0.
2890 The documentation was vague. These values are now always treated as
2891 decimal, and I will make that clear in the documentation.
2897 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2898 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2901 MH/01 Don't check that the operands of numeric comparison operators are
2902 integers when their expansion is in "skipping" mode (fixes bug
2903 introduced by 4.64-PH/07).
2905 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2906 child addresses, Exim now panics and dies. Previously, because the count
2907 is held in a short int, deliveries were likely to be lost. As such a
2908 large number of recipients for a single message is ridiculous
2909 (performance will be very, very poor), I have chosen to impose a limit
2910 rather than extend the field.
2916 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2917 leftover -K file (the existence of which was triggered by #402).
2918 While we were at it, introduced process PID as part of the -K
2919 filename. This should rule out race conditions when creating
2922 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2923 processing considerably. Previous code took too long for large mails,
2924 triggering a timeout which in turn triggers #401.
2926 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2927 in the DK code in transports.c. sendfile() is not really portable,
2928 hence the _LINUX specificness.
2930 TF/01 In the add_headers option to the mail command in an Exim filter,
2931 there was a bug that Exim would claim a syntax error in any
2932 header after the first one which had an odd number of characters
2935 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2936 callout verification, Exim cached a "reject" for the entire domain. This
2937 is correct for most verifications, but it is not correct for a recipient
2938 verification with use_sender or use_postmaster set, because in that case
2939 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2940 case of MAIL FROM:<> rejection from other early rejections (e.g.
2941 rejection of HELO). When verifying a recipient using a non-null MAIL
2942 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2943 Whatever the result of the callout, the value of the domain cache is
2944 left unchanged (for any other kind of callout, getting as far as trying
2945 RCPT means that the domain itself is ok).
2947 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2950 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2951 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2952 session, the daemon ignores ECONNECT errors and logs others; it now
2953 ignores EPIPE as well.
2955 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2956 (quoted-printable decoding).
2958 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2959 later the small subsequent patch to fix an introduced bug.
2961 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2963 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2965 PH/08 An error is now given if message_size_limit is specified negative.
2967 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2968 to be given (somewhat) arbitrary names.
2970 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2973 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2974 miscellaneous code fixes
2976 PH/10 Added the log_reject_target ACL modifier to specify where to log
2979 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2980 hostname. This is wrong, because it relates to the incoming message (and
2981 probably the interface on which it is arriving) and not to the outgoing
2982 callout (which could be using a different interface). This has been
2983 changed to use the value of the helo_data option from the smtp transport
2984 instead - this is what is used when a message is actually being sent. If
2985 there is no remote transport (possible with a router that sets up host
2986 addresses), $smtp_active_hostname is used.
2988 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2989 tweaks were necessary in order to get it to work (see also 21 below):
2990 (a) The code assumed that strncpy() returns a negative number on buffer
2991 overflow, which isn't the case. Replaced with Exim's string_format()
2993 (b) There were several signed/unsigned issues. I just did the minimum
2994 hacking in of casts. There is scope for a larger refactoring.
2995 (c) The code used strcasecmp() which is not a standard C function.
2996 Replaced with Exim's strcmpic() function.
2997 (d) The code set only $1; it now sets $auth1 as well.
2998 (e) A simple test gave the error "authentication client didn't specify
2999 service in request". It would seem that Dovecot has changed its
3000 interface. Fortunately there's a specification; I followed it and
3001 changed what the client sends and it appears to be working now.
3003 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3006 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3007 address A is aliased to B and C, where B exists and C does not. Without
3008 -v the output is "A verified" because verification stops after a
3009 successful redirection if more than one address is generated. However,
3010 with -v the child addresses are also verified. Exim was outputting "A
3011 failed to verify" and then showing the successful verification for C,
3012 with its parentage. It now outputs "B failed to verify", showing B's
3013 parentage before showing the successful verification of C.
3015 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3016 look up a TXT record in a specific list after matching in a combined
3019 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3020 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3021 they consult the DNS. I had assumed they would set it the way they
3022 wanted; and indeed my experiments on Linux seem to show that in some
3023 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3024 To be on the safe side, however, I have now made the interface to
3025 host_find_byname() similar to host_find_bydns(), with an argument
3026 containing the DNS resolver options. The host_find_byname() function now
3027 sets these options at its start, just as host_find_bydns() does. The smtp
3028 transport options dns_qualify_single and dns_search_parents are passed to
3029 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3030 of host_find_byname() use the default settings of RES_DEFNAMES
3031 (qualify_single) but not RES_DNSRCH (search_parents).
3033 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3034 spool_read_header() do less string testing, by means of a preliminary
3035 switch on the second character of optional "-foo" lines. (This is
3036 overdue, caused by the large number of possibilities that now exist.
3037 Originally there were few.) While I was there, I also converted the
3038 str(n)cmp tests so they don't re-test the leading "-" and the first
3039 character, in the hope this might squeeze out yet more improvement.
3041 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3042 flag allowing group syntax was set by the header_syntax check but not
3043 turned off, possible causing trouble later; (2) The flag was not being
3044 set at all for the header_verify test, causing "group"-style headers to
3045 be rejected. I have now set it in this case, and also caused header_
3046 verify to ignore an empty address taken from a group. While doing this, I
3047 came across some other cases where the code for allowing group syntax
3048 while scanning a header line wasn't quite right (mostly, not resetting
3049 the flag correctly in the right place). These bugs could have caused
3050 trouble for malformed header lines. I hope it is now all correct.
3052 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3053 with the "reply" argument non-NULL. The code, however (which originally
3054 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3055 but it didn't always do it. This confused somebody who was copying the
3056 code for some other use. I have removed all the tests.
3058 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3059 feature that was used to support insecure browsers during the U.S. crypto
3060 embargo. It requires special client support, and Exim is probably the
3061 only MTA that supported it -- and would never use it because real RSA is
3062 always available. This code has been removed, because it had the bad
3063 effect of slowing Exim down by computing (never used) parameters for the
3064 RSA_EXPORT functionality.
3066 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3067 authenticator to fail if there's a tab character in the incoming data
3068 (there should never be unless someone is messing about, as it's supposed
3069 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3070 if the connection is using TLS or if the remote IP is the same as the
3071 local IP, and the "valid-client-cert option" if a client certificate has
3074 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3075 authenticators. This can be used for authorization after authentication
3076 succeeds. (In the case of plaintext, it servers for both authentication
3079 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3080 if any retry times were supplied.
3082 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3083 connection, where there is no client IP address to check. In this
3084 situation, the verify now always succeeds.
3086 PH/25 Applied John Jetmore's -Mset patch.
3088 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3090 PH/27 In a string expansion for a processed (not raw) header when multiple
3091 headers of the same name were present, leading whitespace was being
3092 removed from all of them, but trailing whitespace was being removed only
3093 from the last one. Now trailing whitespace is removed from each header
3094 before concatenation. Completely empty headers in a concatenation (as
3095 before) are ignored.
3097 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3098 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3100 PH/29 [Removed. This was a change that I later backed out, and forgot to
3101 correct the ChangeLog entry (that I had efficiently created) before
3102 committing the later change.]
3104 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3105 address errors (4xx response to RCPT) over the same connection as other
3106 messages routed to the same hosts. Such deliveries are always "forced",
3107 so retry times are not inspected. This resulted in far too many retries
3108 for the affected addresses. The effect occurred only when there were more
3109 hosts than the hosts_max_try setting in the smtp transport when it had
3110 the 4xx errors. Those hosts that it had tried were not added to the list
3111 of hosts for which the message was waiting, so if all were tried, there
3112 was no problem. Two fixes have been applied:
3114 (i) If there are any address or message errors in an SMTP delivery, none
3115 of the hosts (tried or untried) are now added to the list of hosts
3116 for which the message is waiting, so the message should not be a
3117 candidate for sending over the same connection that was used for a
3118 successful delivery of some other message. This seems entirely
3119 reasonable: after all the message is NOT "waiting for some host".
3120 This is so "obvious" that I'm not sure why it wasn't done
3121 previously. Hope I haven't missed anything, but it can't do any
3122 harm, as the worst effect is to miss an optimization.
3124 (ii) If, despite (i), such a delivery is accidentally attempted, the
3125 routing retry time is respected, so at least it doesn't keep
3126 hammering the server.
3128 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3129 in ${readsocket because some servers need this prod.
3131 PH/32 Added some extra debug output when updating a wait-xxx database.
3133 PH/33 The hint "could be header name not terminated by colon", which has been
3134 given for certain expansion errors for a long time, was not being given
3135 for the ${if def:h_colon_omitted{... case.
3137 PH/34 The spec says: "With one important exception, whenever a domain list is
3138 being scanned, $domain contains the subject domain." There was at least
3139 one case where this was not true.
3141 PH/35 The error "getsockname() failed: connection reset by peer" was being
3142 written to the panic log as well as the main log, but it isn't really
3143 panic-worthy as it just means the connection died rather early on. I have
3144 removed the panic log writing for the ECONNRESET error when getsockname()
3147 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3148 runs only) independently of the message's sender address. This meant
3149 that, if the 4xx error was in fact related to the sender, a different
3150 message to the same recipient with a different sender could confuse
3151 things. In particular, this can happen when sending to a greylisting
3152 server, but other circumstances could also provoke similar problems.
3153 I have changed the default so that the retry time for these errors is now
3154 based a combination of the sender and recipient addresses. This change
3155 can be overridden by setting address_retry_include_sender=false in the
3158 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3159 remote server are returned as part of bounce messages. This was not
3160 happening for LMTP over a pipe (the lmtp transport), but now it is the
3161 same for both kinds of LMTP.
3163 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3164 in header lines that were in fact CNAMEs. This is no longer the case.
3166 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3167 and queue runs started by the daemon processed all messages. This has
3168 been fixed so that -R and -S can now usefully be given with -q<time>.
3170 PH/40 Import PCRE release 6.7 (fixes some bugs).
3172 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3174 PH/42 Give an error if -q is specified more than once.
3176 PH/43 Renamed the variables $interface_address and $interface_port as
3177 $received_ip_address and $received_port, to make it clear that these
3178 values apply to message reception, and not to the outgoing interface when
3179 a message is delivered. (The old names remain recognized, of course.)
3181 PH/44 There was no timeout on the connect() call when using a Unix domain
3182 socket in the ${readsocket expansion. There now is.
3184 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3185 be meaningful with "accept".
3187 SC/01 Eximstats V1.43
3188 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3190 SC/02 Eximstats V1.44
3191 Use a glob alias rather than an array ref in the generated
3192 parser. This improves both readability and performance.
3194 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3195 Collect SpamAssassin and rejection statistics.
3196 Don't display local sender or destination tables unless
3197 there is data to show.
3198 Added average volumes into the top table text output.
3200 SC/04 Eximstats V1.46
3201 Collect data on the number of addresses (recipients)
3202 as well as the number of messages.
3204 SC/05 Eximstats V1.47
3205 Added 'Message too big' to the list of mail rejection
3206 reasons (thanks to Marco Gaiarin).
3208 SC/06 Eximstats V1.48
3209 Mainlog lines which have GMT offsets and are too short to
3210 have a flag are now skipped.
3212 SC/07 Eximstats V1.49 (Alain Williams)
3213 Added the -emptyok flag.
3215 SC/08 Eximstats V1.50
3216 Fixes for obtaining the IP address from reject messages.
3218 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3219 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3220 whitespace changes from 4.64-PH/27
3222 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3225 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3226 are found, allow negative numbers in numeric criteria)
3228 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3230 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3231 to match changes made in 4.64-PH/43
3233 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3235 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3236 "domain" is actually an IP address, and force a failure. This locks out
3237 those revolvers/nameservers that support "A-for-A" lookups, in
3238 contravention of the specifications.
3240 PH/48 When a host name was looked up from an IP address, and the subsequent
3241 forward lookup of the name timed out, the host name was left in
3242 $sender_host_name, contrary to the specification.
3244 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3245 restricted to single-key lookups, Exim was not diagnosing an error if
3246 * or *@ was used with a query-style lookup.
3248 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3250 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3251 long ago, but noone interested enough thought of it. Let's just say that
3252 the "1.1" means that there are some new functions that weren't there at
3253 some point in the past.
3255 PH/51 Error processing for expansion failure of helo_data from an smtp
3256 transport during callout processing was broken.
3258 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3259 tested/used via the -bh/-bhc/-bs options.
3261 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3262 bug, fixed in subsequent PCRE releases).
3264 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3265 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3267 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3273 SC/01 Use a glob alias rather than an array ref in eximstats generated
3274 parser. This improves both readability and performance.
3276 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3277 Don't display local sender or destination tables in eximstats unless
3278 there is data to show.
3279 Added average volumes into the eximstats top table text output.
3281 SC/03 Collect data on the number of addresses (recipients) as well
3282 as the number of messages in eximstats.
3284 TF/01 Correct an error in the documentation for the redirect router. Exim
3285 does (usually) call initgroups() when daemonizing.
3287 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3288 with consistent privilege compared to when running as a daemon.
3290 TF/03 Note in the spec that $authenticated_id is not set for local
3291 submissions from trusted users.
3293 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3294 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3296 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3297 by adding some example configuration directives to the default
3298 configuration file. A little bit of work is required to uncomment the
3299 directives and define how usernames and passwords are checked, but
3300 there is now a framework to start from.
3302 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3303 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3304 without this. I don't know how relevant this is to other LDAP libraries.
3306 PH/02 Add the verb name to the "unknown ACL verb" error.
3308 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3310 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3312 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3313 directory not expanded when it should be if an expanded home directory
3314 was set for the address (which is overridden by the transport).
3316 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3319 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3320 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3321 because it is too late at that time, and has no effect.
3323 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3324 security issue with \' (bugzilla #107). I could not use the
3325 PQescapeStringConn() function, because it needs a PGconn value as one of
3328 PH/08 When testing addresses using -bt, indicate those final addresses that
3329 are duplicates that would not cause an additional delivery. At least one
3330 person was confused, thinking that -bt output corresponded to deliveries.
3331 (Suppressing duplicates isn't a good idea as you lose the information
3332 about possibly different redirections that led to the duplicates.)
3334 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3335 systems where poll() doesn't work, in particular OS X.
3337 PH/10 Added more information to debugging output for retry time not reached.
3339 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3340 operations in malware.c.
3342 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3345 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3346 syslog_duplication set false, log lines that would normally be written
3347 both the the main log and to the reject log were not written to syslog at
3350 PH/14 In the default configuration, change the use of "message" in ACL warn
3351 statements to "add_header".
3353 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3354 not followed by a command (e.g. "seen endif").
3356 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3357 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3360 PH/17 Added extra conditions to the default value of delay_warning_condition
3364 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3365 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3366 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3369 The Auto-Submitted: and various List- headers are standardised, whereas I
3370 don't think Precedence: ever was.
3372 PH/18 Refactored debugging code in route_finduser() to show more information,
3373 in particular, the error code if getpwnam() issues one.
3375 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3376 This is apparently needed in addition to the PH/07 change above to avoid
3377 any possible encoding problems.
3379 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3380 but not after initializing Perl.
3382 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3383 output them only if debugging. By default they are written stderr,
3384 apparently, which is not desirable.
3386 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3389 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3392 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3394 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3395 authenticated or an ident call has been made. Suppress the default
3396 values for $authenticated_id and $authenticated_sender (but permit -oMai
3397 and -oMas) when testing with -bh.
3399 PH/24 Re-jigged the order of the tests in the default configuration so that the
3400 tests for valid domains and recipients precede the DNS black list and CSA
3401 tests, on the grounds that those ones are more expensive.
3403 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3404 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3405 This bug exists in every version of Exim that I still have, right back to
3408 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3409 However, an attempt to turn on case-sensitivity in a regex key by
3410 including (?-i) didn't work because the subject string was already
3411 lowercased, and the effects were non-intuitive. It turns out that a
3412 one-line patch can be used to allow (?-i) to work as expected.
3418 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3419 other effects) broke the use of negated acl sub-conditions.
3421 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3424 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3425 "Deny" causes Exim to reject the incoming connection with a 554 error.
3426 Unfortunately, if there is a major crisis, such as a disk failure,
3427 tcp-wrappers gives "deny", whereas what one would like would be some
3428 kind of temporary error. A kludge has been added to help with this.
3429 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3430 554 error is used if errno is still zero or contains ENOENT (which occurs
3431 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3434 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3436 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3437 errors. Otherwise a message that provokes a temporary error (when other
3438 messages do not) can cause a whole host to time out.
3440 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3441 addresses were routed directly to files or pipes from a redirect router.
3442 File deliveries just didn't batch; pipe deliveries might have suffered
3445 PH/06 A failure to get a lock for a hints database would erroneously always say
3446 "Failed to get write lock", even when it was really a read lock.
3448 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3449 of 0600. This has been changed to use the value of the lockfile_mode
3450 option (which defaults to 0600).
3452 PH/08 Applied small patch from the Sieve maintainer.
3454 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3455 folder from quota calculations, a direct delivery into this folder messed
3456 up the contents of the maildirsize file. This was because the regex was
3457 used only to exclude .Trash (or whatever) when the size of the mailbox
3458 was calculated. There was no check that a delivery was happening into an
3459 excluded directory. This bug has been fixed by ignoring all quota
3460 processing for deliveries into excluded directories.
3462 PH/10 Added the maildirfolder_create_regex option to appendfile.
3468 PH/01 The code for finding all the local interface addresses on a FreeBSD
3469 system running IPv6 was broken. This may well have applied to all BSD
3470 systems, as well as to others that have similar system calls. The broken
3471 code found IPv4 interfaces correctly, but gave incorrect values for the
3472 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3473 that it would not match correctly against @[] and not recognize the IPv6
3476 PH/02 The ipliteral router was not recognizing addresses of the form user@
3477 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3479 PH/03 Added disable_ipv6.
3481 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3482 decoded form, because it is most often used to construct To: headers
3483 lines in autoreplies, and the decoded form may well be syntactically
3484 invalid. However, $reply_address has leading white space removed, and all
3485 newlines turned into spaces so that the autoreply transport does not
3488 PH/05 If group was specified without a user on a router, and no group or user
3489 was specified on a transport, the group from the router was ignored.
3491 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3492 for visible compile-time settings that can be used to change these
3493 numbers, for those that want even more. Backwards compatibility with old
3494 spool files has been maintained. However, going back to a previous Exim
3495 release will lost any variables that are in spool files.
3497 PH/07 Two small changes when running in the test harness: increase delay when
3498 passing a TCP/IP connection to a new process, in case the original
3499 process has to generate a bounce, and remove special handling of
3500 127.0.0.2 (sic), which is no longer necessary.
3502 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3503 be the same on different OS.
3505 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3508 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3509 whether --show-vars was specified or not
3511 JJ/02 exipick: Added support for new ACL variable spool format introduced
3514 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3515 syntactically invalid From: or Reply-to: line, and a filter used this to
3516 generate an autoreply, and therefore failed to obtain an address for the
3517 autoreply, Exim could try to deliver to a non-existent relative file
3518 name, causing unrelated and misleading errors. What now happens is that
3519 it logs this as a hard delivery error, but does not attempt to create a
3522 PH/11 The exinext utility has a -C option for testing purposes, but although
3523 the given file was scanned by exinext itself; it wasn't being passed on
3524 when Exim was called.
3526 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3527 an end-of-file indication when reading a command response.
3529 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3530 compiled. In many other places in Exim, IPv6 addresses are always
3531 recognized, so I have changed this. It also means that IPv4 domain
3532 literals of the form [IPV4:n.n.n.n] are now always recognized.
3534 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3535 used if the router is not running as root, for example, when verifying at
3536 ACL time, or when using -bh. The debugging output from this situation was
3537 non-existent - all you got was a failure to exec. I have made two
3540 (a) Failures to set uid/gid, the current directory, or a process leader
3541 in a subprocess such as that created by queryprogram now generate
3542 suitable debugging output when -d is set.
3544 (b) The queryprogram router detects when it is not running as root,
3545 outputs suitable debugging information if -d is set, and then runs
3546 the subprocess without attempting to change uid/gid.
3548 PH/15 Minor change to Makefile for building test_host (undocumented testing
3551 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3552 additional section of a DNS packet that returns MX or SRV records.
3553 Instead, it always explicitly searches for A/AAAA records. This avoids
3554 major problems that occur when a DNS server includes only records of one
3555 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3556 fixed another bug: if SRV records were looked up and the corresponding
3557 address records were *not* found in the additional section, the port
3558 values from the SRV records were lost.
3560 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3561 using the correct key (the original address) when searching the retry
3562 rules in order to find which one to use for generating the retry hint.
3564 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3565 adding the default one. Similarly, if it contains a Reply-To: header, the
3566 errors_reply_to option, if set, is not used.
3568 PH/19 When calculating a retry time, Exim used to measure the "time since
3569 failure" by looking at the "first failed" field in the retry record. Now
3570 it does not use this if it is later than than the arrival time of the
3571 message. Instead it uses the arrival time. This makes for better
3572 behaviour in cases where some deliveries succeed, thus re-setting the
3573 "first failed" field. An example is a quota failure for a huge message
3574 when small messages continue to be delivered. Without this change, the
3575 "time since failure" will always be short, possible causing more frequent
3576 delivery attempts for the huge message than are intended.
3577 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3579 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3580 $1, $2, $3) because the numerical variables can be reset during some
3581 expansion items (e.g. "match"), thereby losing the authentication data.
3583 PH/21 Make -bV show the size of off_t variables so that the test suite can
3584 decide whether to run tests for quotas > 2G.
3586 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3587 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3588 filecount value is greater than 2G or if a quota value is greater than 2G
3589 on a system where the size of off_t is not greater than 4, a panic error
3592 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3593 never match. The debug and -bh output now contains an explicit error
3594 message indicating a malformed IPv4 address or mask.
3596 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3597 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3598 PH/23 above applies.
3600 PH/25 Do not write to syslog when running in the test harness. The only
3601 occasion when this arises is a failure to open the main or panic logs
3602 (for which there is an explicit test).
3604 PH/26 Added the /no_tell option to "control=freeze".
3606 PH/27 If a host name lookup failed very early in a connection, for example, if
3607 the IP address matched host_lookup and the reverse lookup yielded a name
3608 that did not have a forward lookup, an error message of the form "no IP
3609 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3610 could be logged. Now it outputs the IP address instead of "NULL".
3612 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3613 allows the sender and the authenticated sender to be set when
3614 submitting a message from within Exim. Since child_open_exim() is
3615 documented for local_scan(), the new function should be too.
3617 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3618 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3619 results in an empty string is now treated as unset.
3621 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3623 PH/31 Added sender_verify_fail logging option.
3625 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3626 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3627 tidied the source and removed it altogether.
3629 PH/33 When a queue run was abandoned because the load average was too high, a
3630 log line was always written; now it is written only if the queue_run log
3631 selector is set. In addition, the log line for abandonment now contains
3632 information about the queue run such as the pid. This is always present
3633 in "start" and "stop" lines but was omitted from the "abandon" line.
3635 PH/34 Omit spaces between a header name and the colon in the error message that
3636 is given when verify = headers_syntax fails (if there are lots of them,
3637 the message gets confusing).
3639 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3640 names, as there are now some PTR records that contain slashes. This check
3641 is only to protect against broken name servers that fall over on strange
3642 characters, so the fact that it applies to all lookups doesn't matter.
3644 PH/36 Now that the new test suite is complete, we can remove some of the
3645 special code in Exim that was needed for the old test suite. For example,
3646 sorting DNS records because real resolvers return them in an arbitrary
3647 order. The new test suite's fake resolver always returns records in the
3650 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3651 bounces) except when queue_only is set, to avoid logging races between
3652 the different processes.
3654 PH/38 Panic-die if .include specifies a non-absolute path.
3656 PH/39 A tweak to the "H" retry rule from its user.
3658 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3659 a label. They prevented compilation on older perls.
3661 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3662 a warning to be raised on newish perls.
3664 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3665 on queue. Changes to match documented behaviour of showing count of
3666 messages matching specified criteria.
3668 PH/40 Changed the default ident timeout from 30s to 5s.
3670 PH/41 Added support for the use of login_cap features, on those BSD systems
3671 that have them, for controlling the resources used by pipe deliveries.
3673 PH/42 The content-scanning code uses fopen() to create files in which to put
3674 message data. Previously it was not paying any attention to the mode of
3675 the files. Exim runs with umask(0) because the rest of the code creates
3676 files with open(), and sets the required mode explicitly. Thus, these
3677 files were ending up world-writeable. This was not a big issue, because,
3678 being within the spool directory, they were not world-accessible. I have
3679 created a function called modefopen, which takes an additional mode
3680 argument. It sets umask(777), creates the file, chmods it to the required
3681 mode, then resets the umask. All the relevant calls to fopen() in the
3682 content scanning code have been changed to use this function.
3684 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3685 to 24 hours. This avoids potential overflow problems when processing G
3686 and H retry rules. I suspect nobody ever tinkers with this value.
3688 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3690 PH/45 When the plaintext authenticator is running as a client, the server's
3691 challenges are checked to ensure they are valid base64 strings. By
3692 default, the authentication attempt is cancelled if an invalid string is
3693 received. Setting client_ignore_invalid_base64 true ignores these errors.
3694 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3695 they are received. Thus, the responses can be made to depend on the
3696 challenges. If an invalid string is ignored, an empty string is placed in
3699 PH/46 Messages that are created by the autoreply transport now contains a
3700 References: header, in accordance with RFCs 2822 and 3834.
3702 PH/47 Added authenticated_sender_force to the smtp transport.
3704 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3706 PH/49 Installed latest patch from the Sieve maintainer.
3708 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3709 was also set, the appendfile transport was unnecessarily scanning a
3710 directory of message files (e.g. for maildir delivery) to find the count
3711 of files (along with the size), even though it did not need this
3712 information. It now does the scan only if it needs to find either the
3713 size of the count of files.
3715 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3717 PH/52 Two bugs concerned with error handling when the smtp transport is
3720 (i) Exim was not creating retry information for temporary errors given
3721 for individual recipients after the DATA command when the smtp transport
3722 was used in LMTP mode. This meant that they could be retried too
3723 frequently, and not timed out correctly.
3725 (ii) Exim was setting the flag that allows error details to be returned
3726 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3727 recipients that were returned after the DATA command.
3729 PH/53 This is related to PH/52, but is more general: for any failing address,
3730 when detailed error information was permitted to be returned to the
3731 sender, but the error was temporary, then after the final timeout, only
3732 "retry timeout exceeded" was returned. Now it returns the full error as
3733 well as "retry timeout exceeded".
3735 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3736 do this, and (what is worse) MTAs that accept it.
3738 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3739 will now be deprecated.
3741 PH/56 New os.c-cygwin from the Cygwin maintainer.
3743 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3744 formats (previously only available in exim formats via -bpr, -bpru,
3745 and -bpra. Now also available in native and exiqgrep formats)
3747 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3748 with very large, slow to parse queues
3750 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3752 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3754 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3755 responses to authentication challenges, though it was showing the
3756 challenges; (ii) I've removed the CR characters from the debug output for
3759 PH/58 Allow for the insertion of a newline as well as a space when a string
3760 is turned into more than one encoded-word during RFC 2047 encoding. The
3761 Sieve code now uses this.
3763 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3764 data_4xx, lost_connection, tls_required.
3766 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3767 message was being sent as an SMTP response.
3769 PH/61 Add -l and -k options to exicyclog.
3771 PH/62 When verifying, if an address was redirected to one new address, so that
3772 verification continued, and the new address failed or deferred after
3773 having set something in $address_data, the value of $address_data was not
3774 passed back to the ACL. This was different to the case when no
3775 redirection occurred. The value is now passed back in both cases.
3777 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3778 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3779 use login_cap.h, so on its own it isn't the distinguishing feature. The
3780 new name refers directly to the setclassresources() function.
3782 PH/65 Added configuration files for NetBSD3.
3784 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3786 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3787 is preferred over IPv4.
3789 PH/68 The bounce_return_message and bounce_return_body options were not being
3790 honoured for bounces generated during the reception of non-SMTP messages.
3791 In particular, this applied to messages rejected by the ACL. This bug has
3792 been fixed. However, if bounce_return_message is true and bounce_return_
3793 body is false, the headers that are returned for a non-SMTP message
3794 include only those that have been read before the error was detected.
3795 (In the case of an ACL rejection, they have all been read.)
3797 PH/69 The HTML version of the specification is now built in a directory called
3798 spec_html instead of spec.html, because the latter looks like a path with
3799 a MIME-type, and this confuses some software.
3801 PH/70 Catch two compiler warnings in sieve.c.
3803 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3804 function verify_get_ident() calls ip_connect() to connect a socket, but
3805 if the "connect()" function timed out, ip_connect() used to close the
3806 socket. However, verify_get_ident() also closes the socket later, and in
3807 between Exim writes to the log, which may get opened at this point. When
3808 the socket was closed in ip_connect(), the log could get the same file
3809 descriptor number as the socket. This naturally causes chaos. The fix is
3810 not to close the socket in ip_connect(); the socket should be closed by
3811 the function that creates it. There was only one place in the code where
3812 this was missing, in the iplookup router, which I don't think anybody now
3813 uses, but I've fixed it anyway.
3815 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3816 well as to direct DNS lookups. Otherwise the handling of names in host
3817 lists is inconsistent and therefore confusing.
3823 PH/01 Two changes to the default runtime configuration:
3825 (1) Move the checks for relay_from_hosts and authenticated clients from
3826 after to before the (commented out) DNS black list checks.
3828 (2) Add control=submission to the relay_from_hosts and authenticated
3829 clients checks, on the grounds that messages accepted by these
3830 statements are most likely to be submissions.
3832 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3834 (1) Generate an error if the third argument for the ${prvs expansion is
3837 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3840 (3) Reset the variables that are obtained from the first argument of
3841 ${prvscheck and used in the second argument before leaving the code,
3842 because their memory is reclaimed, so using them afterwards may do
3845 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3846 one (it's much easier than Tom thought :-).
3848 (5) Because of (4), we can now allow for the use of $prvscheck_result
3849 inside the third argument.
3851 PH/03 For some reason, the default setting of PATH when running a command from
3852 a pipe transport was just "/usr/bin". I have changed it to
3855 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3856 anything to be listed in the output from -bV.
3858 PH/05 When a filter generated an autoreply, the entire To: header line was
3859 quoted in the delivery log line, like this:
3861 => >A.N.Other <ano@some.domain> <original@ddress> ...
3863 This has been changed so that it extracts the operative address. There
3864 may be more than one such address. If so, they are comma-separated, like
3867 => >ano@some.domain,ona@other.domain <original@ddress> ...
3869 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3870 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3871 not being looked up in the rDNS to get a host name, Exim was showing the
3872 IP address twice in Received: lines, even though the IP addresses were
3873 identical. For example:
3875 Received: from [1.2.3.4] (helo=[1.2.3.4])
3877 However, if the real host name was known, it was omitting the HELO data
3878 if it matched the actual IP address. This has been tidied up so that it
3879 doesn't show the same IP address twice.
3881 PH/07 When both +timestamp and +memory debugging was on, the value given by
3882 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3883 called by the debug printing, thereby overwriting the timestamp buffer.
3884 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3886 PH/08 When the original message was included in an autoreply transport, it
3887 always said "this is a copy of the message, including all the headers",
3888 even if body_only or headers_only was set. It now gives an appropriate
3891 PH/09 Applied a patch from the Sieve maintainer which:
3893 o fixes some comments
3894 o adds the (disabled) notify extension core
3895 o adds some debug output for the result of if/elsif tests
3896 o points to the current vacation draft in the documentation
3897 and documents the missing references header update
3901 o fixes a bug in processing the envelope test (when testing
3902 multiple envelope elements, the last element determined the
3905 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3906 Electronic Mail") by including:
3908 Auto-submitted: auto-generated
3910 in the messages that it generates (bounce messages and others, such as
3911 warnings). In the case of bounce messages for non-SMTP messages, there was
3912 also a typo: it was using "Auto_submitted" (underscore instead of
3913 hyphen). Since every message generated by Exim is necessarily in response
3914 to another message, thes have all been changed to:
3916 Auto-Submitted: auto-replied
3918 in accordance with these statements in the RFC:
3920 The auto-replied keyword:
3922 - SHOULD be used on messages sent in direct response to another
3923 message by an automatic process,
3925 - MUST NOT be used on manually-generated messages,
3927 - MAY be used on Delivery Status Notifications (DSNs) and Message
3928 Disposition Notifications (MDNs),
3930 - MUST NOT be used on messages generated by automatic or periodic
3931 processes, except for messages which are automatic responses to
3934 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3935 to the default Received: header definition.
3937 PH/12 Added log selector acl_warn_skipped (default on).
3939 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3940 variables because (a) they are in the wrong storage pool and (b) even if
3941 they were copied, it wouldn't work properly because of the caching.
3943 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3944 checking when decoding. Apparently there are clients that generate
3945 overlong encoded strings. Why am I not surprised?
3947 PH/15 If the first argument of "${if match_address" was not empty, but did not
3948 contain an "@" character, Exim crashed. Now it writes a panic log message
3949 and treats the condition as false.
3951 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3953 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3954 extension "envelope-auth". The code is finished and in agreement with
3955 other implementations, but there is no documentation so far and in fact,
3956 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3957 not changing the active code.
3959 Print executed "if" and "elsif" statements when debugging is used. This
3960 helps a great deal to understand what a filter does.
3962 Document more things not specified clearly in RFC3028. I had all this
3963 sorted out, when out of a sudden new issues came to my mind. Oops."
3965 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3968 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3969 canonical form (as documented). However, after a host name lookup from
3970 the IP address, check_host() was doing a simple string comparison with
3971 addresses acquired from the DNS when checking that the found name did
3972 have the original IP as one of its addresses. Since any found IPv6
3973 addresses are likely to be in abbreviated form, the comparison could
3974 fail. Luckily, there already exists a function for doing the comparison
3975 by converting both addresses to binary, so now that is used instead of
3976 the text comparison.
3978 PH/20 There was another similar case to PH/19, when a complete host name was
3979 given in a host list; looking up its IP address could give an abbreviated
3980 form, whereas the current host's name might or might not be abbreviated.
3981 The same fix has been applied.
3987 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3988 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3991 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3992 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3994 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3996 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3997 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3998 Also, for "Auto-Submitted", treat anything other than "no" as
3999 non-personal, in accordance with RFC 3834. (Previously it treated
4000 anything starting "auto-" as non-personal.)
4002 TF/01 The control=submission/name=... option had a problem with syntax
4003 errors if the name included a slash character. The /name= option
4004 now slurps the rest of the string, so it can include any characters
4005 but it must come last in the list of options (after /sender_retain
4008 PH/05 Some modifications to the interface to the fake nameserver for the new
4016 TK/01 Added the "success_on_redirect" address verification option. See
4017 NewStuff for rationale and an example.
4019 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4021 PH/02 Patch to exigrep to allow it to work on syslog lines.
4023 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4024 fread() to skip over the body file's header line, because in Cygwin the
4025 header line is locked and is inaccessible.
4027 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4028 co-exist for some time) to make it clear that it is the Exim ID that is
4029 referenced, not the Message-ID: header line.
4031 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4032 string_format() function, because snprintf() does not exist on all
4035 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4036 ${stat: expansion item.
4038 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4039 protocol synchronization error", to keep the pedants happy.
4041 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4042 well as for IRIX systems, when gcc is being used. See the host.c source
4045 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4047 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4050 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4051 transport and to the smtp transport in LMTP mode.
4053 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4055 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4056 run a filter in a subprocess. This could lead to confusion in subsequent
4057 lookups in the parent process. There should also be a search_tidyup() at
4058 the end of the subprocess.
4060 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4061 only if the host matched helo_try_verify_hosts, which caused the
4062 verification to occur when the EHLO/HELO command was issued. The ACL just
4063 tested the remembered result. Now, if a previous verification attempt has
4064 not happened, "verify = helo" does it there and then.
4066 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4068 TK/03 Fix log output including CR from clamd.
4070 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4071 exist provoked a memory error which could cause a segfault.
4073 PH/15 Installed PCRE 6.2
4075 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4077 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4078 of the problem. Specifically, suggested +O2 rather than +O1 for the
4081 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4083 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4084 record, the port was not correctly logged when the outgoing_port log
4085 selector was set (it logged the transort's default port).
4087 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4088 fallback_hosts, and "hosts" in the smtp transport.
4090 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4091 host errors such as "Connection refused".
4093 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4094 authentication with radiusclient 0.4.9:
4096 - Error returned from rc_read_config was caught wrongly
4097 - Username/password not passed on to radius server due to wrong length.
4099 The presumption is that some radiusclient API changes for 4.51/PH/17
4100 were not taken care of correctly. The code is still untested by me (my
4101 Linux distribution still has 0.3.2 of radiusclient), but it was
4102 contributed by a Radius user.
4104 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4105 expanding the "port" option of the smtp transport.
4107 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4108 while reading a MIME header. Thanks to Tom Hughes for a patch.
4110 PH/24 Include config.h inside local_scan.h so that configuration settings are
4113 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4114 that in an ACL for RCPT (for example), you can examine exactly what was
4117 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4118 commands, but it was not correctly comparing the address with the actual
4119 client host address. Thus, it would show the EHLO address in Received:
4120 header lines when this was not necessary.
4122 PH/27 Added the % operator to ${eval:}.
4124 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4125 it should be ignoring failures (because with -C, for example, it has lost
4126 privilege). It wasn't ignoring creation failures other than "already
4129 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4132 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4133 because an input line was too long, either on its own, or by virtue of
4134 too many continuations, the temporary file was not being removed, and the
4135 return code was incorrect.
4137 PH/31 Missing "BOOL" in function definition in filtertest.c.
4139 PH/32 Applied Sieve patches from the maintainer.
4141 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4143 PH/33 Added "verify = not_blind".
4145 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4146 Local/Makefile (with some defaults set). These are used in built scripts
4147 such as exicyclog, but they have never been used in the exim_install
4148 script (though there are many overriding facilities there). I have
4149 arranged that the exim_install script now takes note of these two
4152 PH/35 Installed configuration files for Dragonfly.
4154 PH/36 When a locally submitted message by a trusted user did not contain a
4155 From: header, and the sender address was obtained from -f or from an SMTP
4156 MAIL command, and the trusted user did not use -F to supply a sender
4157 name, $originator_name was incorrectly used when constructing a From:
4158 header. Furthermore, $originator_name was used for submission mode
4159 messages from external hosts without From: headers in a similar way,
4160 which is clearly wrong.
4162 PH/37 Added control=suppress_local_fixups.
4164 PH/38 When log_selector = +received_sender was set, and the addition of the
4165 sender made the log line's construction buffer exactly full, or one byte
4166 less than full, an overflow happened when the terminating "\n" was
4169 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4170 when the result of a list match is failure because a DNS lookup failed.
4172 PH/40 RM_COMMAND is now used in the building process.
4174 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4175 the "build-* directories that it finds.
4177 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4178 address in a domain literal was a prefix of an interface address.
4180 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4181 when verifying a sender address, unless rewrite_headers is false.
4183 PH/44 Wrote a long comment about why errors_to addresses are verified as
4184 recipients, not senders.
4186 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4187 the ratelimit ACL was added.
4189 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4191 PH/46 Added extra information about PostgreSQL errors to the error string.
4193 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4194 avoiding the need to install special zones in a real server. This is
4195 backwards compatible; if it can't find the fake resolver, it drops back.
4196 Thus, both old and new test suites can be run.
4198 TF/02 Added util/ratelimit.pl
4200 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4201 clock is set back in time.
4203 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4204 Candler <B.Candler@pobox.com>.
4206 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4207 OK for addresses that are the result of redirections.
4209 PH/48 A number of further additions for the benefit of the new test suite,
4210 including a fake gethostbyname() that interfaces to the fake DNS resolver
4213 TF/06 The fix for widen_domains has also been applied to qualify_single and
4214 search_parents which are the other dnslookup options that can cause
4217 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4220 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4222 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4223 service" if it encountered an MX record with an empty target hostname.
4224 The message is now "an MX or SRV record indicated no SMTP service".
4226 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4227 if the DNS of the sending site is misconfigured. This is quite a
4228 common situation. This change restores the behaviour of treating a
4229 helo verification defer as a failure.
4231 PH/51 If self=fail was set on a router, the bounce message did not include the
4232 actual error message.
4238 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4240 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4241 command itself ended in error, the underlying message about the transport
4242 filter timeout was being overwritten with the pipe command error. Now the
4243 underlying error message should be appended to the second error message.
4245 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4247 PH/02 Reduce the amount of output that "make" produces by default. Full output
4248 can still be requested.
4250 PH/03 The warning log line about a condition test deferring for a "warn" verb
4251 was being output only once per connection, rather than after each
4252 occurrence (because it was using the same function as for successful
4253 "warn" verbs). This seems wrong, so I have changed it.
4255 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4256 it should not have, which might have caused a crash in the right
4257 circumstances, but probably never did.
4259 PH/04 Installed a modified version of Tony Finch's patch to make submission
4260 mode fix the return path as well as the Sender: header line, and to
4261 add a /name= option so that you can make the user's friendly name appear
4264 TF/03 Added the control = fakedefer ACL modifier.
4266 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4267 Mark Lowes for thorough testing.
4269 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4271 TK/03 Merged latest SRS patch from Miles Wilton.
4273 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4274 with the definition in sysexits.h (which is #included earlier).
4275 Fortunately, Exim does not actually use EX_OK. The code used to try to
4276 preserve the sysexits.h value, by assuming that macro definitions were
4277 scanned for macro replacements. I have been disabused of this notion,
4278 so now the code just undefines EX_OK before #including unistd.h.
4280 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4281 in the smtp transport. When a block could not be written in a single
4282 write() function, the timeout was being re-applied to each part-write.
4283 This seems wrong - if the receiver was accepting one byte at a time it
4284 would take for ever. The timeout is now adjusted when this happens. It
4285 doesn't have to be particularly precise.
4287 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4288 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4290 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4291 without a domain if the check to <postmaster@domain> fails.
4293 SC/01 Eximstats: added -xls and the ability to specify output files
4294 (patch written by Frank Heydlauf).
4296 SC/02 Eximstats: use FileHandles for outputting results.
4298 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4300 SC/04 Eximstats: fixed display of large numbers with -nvr option
4302 SC/05 Eximstats: fixed merging of reports with empty tables.
4304 SC/06 Eximstats: added the -include_original_destination flag
4306 SC/07 Eximstats: removed tabs and trailing whitespace.
4308 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4310 TK/06 MBOX spool code: Add real "From " MBOX separator line
4311 so the .eml file is really in mbox format (even though
4312 most programs do not really care). Patch from Alex Miller.
4314 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4315 The latter is generated from $received_to and is only set if the
4316 message has one envelope recipient. SA can use these headers,
4317 obviously out-of-the-box. Patch from Alex Miller.
4319 PH/08 The ${def test on a variable was returning false if the variable's
4320 value was "0", contrary to what the specification has always said!
4321 The result should be true unless the variable is empty.
4323 PH/09 The syntax error of a character other than { following "${if
4324 def:variable_name" (after optional whitespace) was not being diagnosed.
4325 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4326 accidental colon was present, for example, could give incorrect results.
4328 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4329 result is used (not including appendfile, where other changes are about
4332 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4333 This involved changing a lot of size variables from int to off_t. It
4334 should work with maildirs and everything.
4336 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4337 spamd dying while we are connected to it.
4339 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4342 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4343 function for BDB 4.3.
4345 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4347 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4348 the sources for PCRE have been re-arranged and more files are now
4351 PH/15 The code I had for printing potentially long long variables in PH/11
4352 above was not the best (it lost precision). The length of off_t variables
4353 is now inspected at build time, and an appropriate printing format (%ld
4354 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4355 to be "long long int" or "long int". This is needed for the internal
4356 formatting function string_vformat().
4358 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4359 the configuration file to be ":syslog", then the script "guesses" where
4360 the logs files are, rather than using the compiled in default. In our
4361 case the guess is not the same as the compiled default, so the script
4362 suddenly stopped working when I started to use syslog. The patch checks
4363 to see if log_file_path is "". If so, it attempts to read it from exim
4364 with no configuration file to get the compiled in version, before it
4365 falls back to the previous guessing code."
4367 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4368 implementing BATV in an Exim configuration. See NewStuff for the gory
4371 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4372 Makefile that are specific to HP-UX.
4374 PH/18 If the "use_postmaster" option was set for a recipient callout together
4375 with the "random" option, the postmaster address was used as the MAIL
4376 FROM address for the random test, but not for the subsequent recipient
4377 test. It is now used for both.
4379 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4380 patch removes a few documentation additions to RFC 3028, because the
4381 latest draft now contains them. It adds the new en;ascii-case comparator
4382 and a new error check for 8bit text in MIME parts. Comparator and
4383 require names are now matched exactly. I enabled the subaddress
4384 extension, but it is not well tested yet (read: it works for me)."
4386 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4387 rework some of the code of TK/09 above to avoid the hardwired use of
4388 "%lld" and "long long". Replaced the call to snprintf() with a call to
4391 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4392 records point to non-existent hosts", "retry timeout exceeded", and
4393 "retry time not reached for any host after a long failure period".
4395 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4396 experimental DomainKeys support:
4398 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4399 (2) On an error such as an illegally used "control", the wrong name for
4400 the control was given.
4402 These problems did NOT occur unless DomainKeys support was compiled.
4404 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4406 PH/24 Added ${if match_ip condition.
4408 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4409 databases so that it will be absolutely obvious if a crash occurs in the
4410 DB library. This is a regular occurrence (often caused by mis-matched
4413 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4414 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4415 picked up on a user's system that detects such things. There doesn't seem
4416 to be a gcc warning option for this - only an attribute that has to be
4417 put on the function's prototype. It seems that in Fedora Core 4 they have
4418 set this on a number of new functions. No doubt there will be more in due
4421 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4422 specify a transport. However, if an address that was verified by such a
4423 router was the subject of a callout, Exim crashed because it tried to
4424 read the rcpt_include_affixes from the non-existent transport. Now it
4425 just assumes that the setting of that option is false. This bug was
4426 introduced by 4.51/PH/31.
4428 PH/28 Changed -d+all to exclude +memory, because that information is very
4429 rarely of interest, but it makes the output a lot bigger. People tend to
4430 do -d+all out of habit.
4432 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4433 code in os-type was giving problems when libc.so lives in lib64, like on
4436 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4437 aren't the modern standard, and it seems that some systems' include files
4438 don't always have them. Exim was already checking for some of the newer
4439 ones like T_AAAA, and defining it itself. I've added checks for all the
4440 record types that Exim uses.
4442 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4443 not automatically generating a new one, as it is supposed to. This
4444 prevented TLS from working. If the file did exist, but contained invalid
4445 data, a new version was generated, as expected. It was only the case of a
4446 non-existent file that was broken.
4448 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4449 with a change in libdomainkeys > 0.64.
4451 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4452 from DNS. If the selector record carries the flag, it now has
4453 precedence over the domain-wide flag.
4455 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4457 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4458 the use of a port name that isn't defined in /etc/services) occurred, the
4459 message was deferred as in a normal delivery, and thus remained on the
4460 spool, instead of being failed because of the mua_wrapper setting. This
4461 is now fixed, and I tidied up some of the mua_wrapper messages at the
4464 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4465 the messages in a hash of arrays rather than using individual hashes.
4466 This is a bit cleaner and results in dramatic memory savings, albeit
4467 at a slight CPU cost.
4469 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4470 as requested by Marc Sherman.
4472 SC/10 Eximstats: added histograms for user specified patterns as requested
4475 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4477 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4478 fopen() in the content-scanning modules that did not already have it.
4484 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4485 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4487 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4489 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4491 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4492 to test Sieve filters that use "vacation".
4494 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4495 that changes the way the GnuTLS parameters are stored in the cache file.
4496 The new format can be generated externally. For backward compatibility,
4497 if the data in the cache doesn't make sense, Exim assumes it has read an
4498 old-format file, and it generates new data and writes a new file. This
4499 means that you can't go back to an older release without removing the
4502 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4503 work if there are any delivery delays because "one_time" forces the
4504 parent to be marked "delivered", so its unseen clone is never tried
4505 again. For this reason, Exim now forbids the simultaneous setting of
4508 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4509 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4510 entry if you want to know the details. The fix, however, neglected to
4511 consider the case when local delivery batching is involved. The test for
4512 "previously delivered" was not happening when checking to see if an
4513 address could be batched with a previous (undelivered) one; under
4514 certain circumstances this could lead to multiple deliveries to the same
4517 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4518 in its include files, and this causes problems building Exim.
4520 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4521 header_syntax) but Exim was just ignoring anything given after a slash.
4522 In particular, this caused confusion with an attempt to use "verify =
4523 reverse_host_lookup/defer_ok". An error is now given when options are
4524 supplied for verify items that do not have them. (Maybe reverse_host_
4525 lookup should have a defer_ok option, but that's a different point.)
4527 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4528 defined by RFC 821) to 2048, because there were problems with some AUTH
4529 commands, and RFC 1869 says the size should be increased for extended
4530 SMTP commands that take arguments.
4532 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4535 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4536 "unknown" error; now it says that the functionality isn't in the binary.
4538 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4539 an address' error message when a string expansion fails (syntax or
4540 whatever). Otherwise the password may appear in the log. Following change
4541 PH/42 below, there is no longer a chance of it appearing in a bounce
4544 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4546 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4547 of its addresses was ever tried. (Bugzilla bug #2.)
4549 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4550 the result incorrectly in the debug output. (It correctly added a newline
4551 to what was transported.)
4553 TF/01 Added $received_time.
4555 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4556 commented out examples of how to interface to a virus scanner and to
4557 SpamAssassin. Also added commented examples of av_scanner and
4558 spamd_address settings.
4560 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4561 and controls are allowed in which ACLs. There were a couple of minor
4562 errors. Some of the entries in the conditions table (which is a table of
4563 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4564 negation of where the condition IS allowed.
4566 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4568 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4569 header file does not have a version number, so I've had to invent a new
4570 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4571 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4572 radiusclient), but it was contributed by a Radius user.
4574 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4575 files or directories, for OpenSSL.
4577 PH/19 When an Exim process that is running as root has to create an Exim log
4578 file, it does so in a subprocess that runs as exim:exim so as to get the
4579 ownership right at creation (otherwise, other Exim processes might see
4580 the file with the wrong ownership). There was no test for failure of this
4581 fork() call, which would lead to the process getting stuck as it waited
4582 for a non-existent subprocess. Forks do occasionally fail when resources
4583 run out. I reviewed all the other calls to fork(); they all seem to check
4586 PH/20 When checking for unexpected SMTP input at connect time (before writing
4587 the banner), Exim was not dealing correctly with a non-positive return
4588 from the read() function. If the client had disconnected by this time,
4589 the result was a log entry for a synchronization error with an empty
4590 string after "input=" when read() returned zero. If read() returned -1
4591 (an event I could not check), uninitialized data bytes were printed.
4592 There were reports of junk text (parts of files, etc) appearing after
4595 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4597 PH/22 Added support for macro redefinition, and (re)definition in between
4598 driver and ACL definitions.
4600 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4601 forgetting to use the resulting value; it was using the unexpanded value.
4603 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4604 hadn't been configured. The fix is from Juergen Kreileder, who
4605 understands it better than I do:
4607 "Here's what I see happening with three configured cyrus_sasl
4608 authenticators configured (plain, login, cram-md5):
4610 On startup auth_cyrus_sasl_init() gets called for each of these.
4611 This means three calls to sasl_listmech() without a specified mech_list.
4612 => SASL tests which mechs of all available mechs actually work
4613 => three warnings about OTP not working
4614 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4616 With the patch, sasl_listmech() also gets called three times. But now
4617 SASL's mech_list option is set to the server_mech specified in the the
4618 authenticator. Or in other words, the answer from sasl_listmech()
4619 gets limited to just the mech you're testing for (which is different
4621 => the return list contains just 'plain' or 'login', 'cram-md5' or
4622 nothing depending on the value of ob->server_mech.
4624 I've just tested the patch: Authentication still works fine,
4625 unavailable mechs specified in the exim configuration are still
4626 caught, and the auth.log warnings about OTP are gone."
4628 PH/25 When debugging is enabled, the contents of the command line are added
4629 to the debugging output, even when log_selector=+arguments is not
4632 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4633 answer is "GNU", and only if the return is "GNU/something" is the answer
4636 PH/27 $acl_verify_message is now set immediately after the failure of a
4637 verification in an ACL, and so is available in subsequent modifiers. In
4638 particular, the message can be preserved by coding like this:
4640 warn !verify = sender
4641 set acl_m0 = $acl_verify_message
4643 Previously, $acl_verify_message was set only while expanding "message"
4644 and "log_message" when a very denied access.
4646 PH/28 Modified OS/os.c-Linux with
4648 -#ifndef OS_LOAD_AVERAGE
4649 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4651 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4652 nomenclature these days.)
4654 PH/29 Installed patch from the Sieve maintainer that adds the options
4655 sieve_useraddress and sieve_subaddress to the redirect router.
4657 PH/30 In these circumstances:
4658 . Two addresses routed to the same list of hosts;
4659 . First host does not offer TLS;
4660 . First host accepts first address;
4661 . First host gives temporary error to second address;
4662 . Second host offers TLS and a TLS session is established;
4663 . Second host accepts second address.
4664 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4665 and peerdn, if requested) that were in fact used only for the second
4668 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4669 attention to any local part prefix or suffix that was matched by the
4670 router that accepted the address. It now behaves in the same way as it
4671 does for delivery: the affixes are removed from the local part unless
4672 rcpt_include_affixes is set on the transport.
4674 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4675 timeout during the DATA phase of an incoming message.
4677 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4678 applied a patch sanctioned by the Sieve maintainer.
4680 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4681 the uid or gid is negative. A case of a negative gid caused this to be
4682 noticed. The fix allows for either to be negative.
4684 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4685 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4688 PH/36 The change PH/12 above was broken. Fixed it.
4690 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4691 the grounds that routing the same address twice would always produce the
4692 same answer. This might have been true once, but it is certainly no
4693 longer true now. Routing a child address may depend on the previous
4694 routing that produced that child. Some complicated redirection strategies
4695 went wrong when messages had multiple recipients, and made Exim's
4696 behaviour dependent on the order in which the addresses were given.
4698 I have moved the duplicate checking until after the routing is complete.
4699 Exim scans the addresses that are assigned to local and remote
4700 transports, and removes any duplicates. This means that more work will be
4701 done, as duplicates will always all be routed, but duplicates are
4702 presumably rare, so I don't expect this is of any significance.
4704 For deliveries to pipes, files, and autoreplies, the duplicate checking
4705 still happens during the routing process, since they are not going to be
4708 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4709 It corrects a timeout issue with spamd. This is Ian's comment: "The
4710 background is that sometimes spamd either never reads data from a
4711 connection it has accepted, or it never writes response data. The exiscan
4712 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4713 blindly assumes that writes won't block so it may never time out."
4715 PH/39 Allow G after quota size as well as K and M.
4717 PH/40 The value set for $authenticated_id in an authenticator may not contain
4718 binary zeroes or newlines because the value is written to log lines and
4719 to spool files. There was no check on this. Now the value is run through
4720 the string_printing() function so that such characters are converted to
4721 printable escape sequences.
4723 PH/41 $message_linecount is a new variable that contains the total number of
4724 lines in the message. Compare $body_linecount, which is the count for the
4727 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4728 bounce and delay warning messages, except in certain special cases, which
4731 (a) An SMTP error message from a remote host;
4732 (b) A message specified in a :fail: redirection;
4733 (c) A message specified in a "fail" command in a system filter;
4734 (d) A message specified in a FAIL return from the queryprogram router;
4735 (e) A message specified by the cannot_route_message router option.
4737 In these cases only, Exim does include the error details in bounce and
4738 warning messages. There are also a few cases where bland messages such
4739 as "unrouteable address" or "local delivery error" are given.
4741 PH/43 $value is now also set for the "else" part of a ${run expansion.
4743 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4744 being worked on, but at least Exim now implements the latest version to
4747 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4748 process to complete was treated as a delivery failure, a timeout while
4749 writing the message to the pipe was logged, but erroneously treated as a
4750 successful delivery. Such timeouts include transport filter timeouts. For
4751 consistency with the overall process timeout, these timeouts are now
4752 treated as errors, giving rise to delivery failures by default. However,
4753 there is now a new Boolean option for the pipe transport called
4754 timeout_defer, which, if set TRUE, converts the failures into defers for
4755 both kinds of timeout. A transport filter timeout is now identified in
4758 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4759 systems where "make" and "gmake" are different, calling "gmake" at top
4760 level broke things. I've arranged for the value of $(MAKE) to be passed
4761 from the Makefile to this script so that it can call the same version of
4765 A note about Exim versions 4.44 and 4.50
4766 ----------------------------------------
4768 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4769 changes of various kinds. As a consequence, a big documentation update was
4770 needed. This delayed the release for rather longer than seemed good, especially
4771 in the light of a couple of (minor) security issues. Therefore, the changes
4772 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4773 release. So 4.44 and 4.50 are in effect two different branches that both start
4776 I have left the 4.50 change log unchanged; it contains all the changes since
4777 4.43. The change log for 4.44 is below; many of its items are identical to
4778 those for 4.50. This seems to be the most sensible way to preserve the
4779 historical information.
4785 1. Minor wording change to the doc/README.SIEVE file.
4787 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4788 computation of the current number of files was incorrect.
4790 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4791 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4794 4. Give more explanation in the error message when the command for a transport
4795 filter fails to execute.
4797 5. There are several places where Exim runs a non-Exim command in a
4798 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4799 was being done only for the command run by the queryprogram router. It is
4800 now done for all such subprocesses. The other cases are: ${run, transport
4801 filters, and the commands run by the lmtp and pipe transports.
4803 6. Added CONFIGURE_GROUP build-time option.
4805 7. Some older OS have a limit of 256 on the maximum number of file
4806 descriptors. Exim was using setrlimit() to set 1000 as a large value
4807 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4808 systems. I've change it so that if it can't get 1000, it tries for 256.
4810 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4811 was an oversight, and furthermore, ever since the addition of extra
4812 controls (e.g. 4.43/32), the checks on when to allow different forms of
4813 "control" were broken. There should now be diagnostics for all cases when a
4814 control that does not make sense is encountered.
4816 9. Added the /retain_sender option to "control=submission".
4818 10. $recipients is now available in the predata ACL (oversight).
4820 11. Tidy the search cache before the fork to do a delivery from a message
4821 received from the command line. Otherwise the child will trigger a lookup
4822 failure and thereby defer the delivery if it tries to use (for example) a
4823 cached ldap connection that the parent has called unbind on.
4825 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4826 of $address_data from the recipient verification was clobbered by the
4827 sender verification.
4829 13. The value of address_data from a sender verification is now available in
4830 $sender_address_data in subsequent conditions in the ACL statement.
4832 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4834 15. Added a new option "connect=<time>" to callout options, to set a different
4837 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4838 was its contents. (It was OK if the option was not defined at all.)
4840 17. A "Completed" log line is now written for messages that are removed from
4841 the spool by the -Mrm option.
4843 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4844 information about exactly what failed.
4846 19. Added -dd to debug only the daemon process.
4848 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4849 handles timeouts, both on the server side and network timeouts. Renamed the
4850 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4852 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4853 It is now set to "smtps".
4855 22. $host_address is now set to the target address during the checking of
4856 ignore_target_hosts.
4858 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4859 being passed; this would have caused $sender_host_name to have been used if
4860 matching the list had actually called for a host name (not very likely,
4861 since this list is usually IP addresses). A host name is now passed as
4864 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4865 code that specifies a non-restarting handler (typically sigaction() in
4866 modern systems) in an attempt to fix a rare and obscure crash bug.
4868 25. Narrowed the window for a race in the daemon that could cause it to ignore
4869 SIGCHLD signals. This is not a major problem, because they are used only to
4870 wake it up if nothing else does.
4872 26. A malformed maildirsize file could cause Exim to calculate negative values
4873 for the mailbox size or file count. Odd effects could occur as a result.
4874 The maildirsize information is now recalculated if the size or filecount
4877 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4878 support for a long time. Removed HAVE_SYS_VFS_H.
4880 28. Installed the latest version of exipick from John Jetmore.
4882 29. In an address list, if the pattern was not a regular expression, an empty
4883 subject address (from a bounce message) matched only if the pattern was an
4884 empty string. Non-empty patterns were not even tested. This was the wrong
4885 because it is perfectly reasonable to use an empty address as part of a
4886 database query. An empty address is now tested by patterns that are
4887 lookups. However, all the other forms of pattern expect the subject to
4888 contain a local part and a domain, and therefore, for them, an empty
4889 address still always fails if the pattern is not itself empty.
4891 30. Exim went into a mad DNS loop when attempting to do a callout where the
4892 host was specified on an smtp transport, and looking it up yielded more
4893 than one IP address.
4895 31. Re-factored the code for checking spool and log partition space into a
4896 function that finds that data and another that does the check. The former
4897 is then used to implement four new variables: $spool_space, $log_space,
4898 $spool_inodes, and $log_inodes.
4900 32. The RFC2047 encoding function was originally intended for short strings
4901 such as real names; it was not keeping to the 75-character limit for
4902 encoded words that the RFC imposes. It now respects the limit, and
4903 generates multiple encoded words if necessary. To be on the safe side, I
4904 have increased the buffer size for the ${rfc2047: expansion operator from
4907 33. It is now permitted to omit both strings after an "if" condition; if the
4908 condition is true, the result is "true". As before, when the second string
4909 is omitted, a false condition yields an empty string. This makes it less
4910 cumbersome to write custom ACL and router conditions.
4912 34. Failure to deliver a bounce message always caused it to be frozen, even if
4913 there was an errors_to setting on the router. The errors_to setting is now
4916 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4917 canonical form (fully expanded) before being placed in
4918 $sender_host_address.
4920 36. The table in the code that translates DNS record types into text (T_A to
4921 "A" for instance) was missing entries for NS and CNAME. It is just possible
4922 that this could have caused confusion if both these types were looked up
4923 for the same domain, because the text type is used as part of Exim's
4924 per-process caching. But the chance of anyone hitting this buglet seems
4927 37. The dnsdb lookup has been extended in a number of ways.
4929 (1) There is a new type, "zns", which walks up the domain tree until it
4930 finds some nameserver records. It should be used with care.
4932 (2) There is a new type, "mxh", which is like "mx" except that it returns
4933 just the host names, not the priorities.
4935 (3) It is now possible to give a list of domains (or IP addresses) to be
4936 looked up. The behaviour when one of the lookups defers can be
4937 controlled by a keyword.
4939 (4) It is now possible to specify the separator character for use when
4940 multiple records are returned.
4942 38. The dnslists ACL condition has been extended: it is now possible to supply
4943 a list of IP addresses and/or domains to be looked up in a particular DNS
4946 39. Added log_selector=+queue_time_overall.
4948 40. When running the queue in the test harness, wait just a tad after forking a
4949 delivery process, to get repeatability of debugging output.
4951 41. Include certificate and key file names in error message when GnuTLS fails
4952 to set them up, because the GnuTLS error message doesn't include the name
4953 of the failing file when there is a problem reading it.
4955 42. Allow both -bf and -bF in the same test run.
4957 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4959 44. The "Exiscan patch" is now merged into the mainline Exim source.
4961 45. Sometimes the final signoff response after QUIT could fail to get
4962 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4963 < 0 before doing a fflush(). This bug looks as though it goes back to the
4964 introduction of TLS in release 3.20, but "sometimes" must have been rare
4965 because the tests only now provoked it.
4967 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4968 (this can affect the format of dates).
4970 47. exim_tidydb, when checking for the continued existence of a message for
4971 which it has found a message-specific retry record, was not finding
4972 messages that were in split spool directories. Consequently, it was
4973 deleting retry records that should have stayed in existence.
4975 48. Steve fixed some bugs in eximstats.
4977 49. The SPA authentication driver was not abandoning authentication and moving
4978 on to the next authenticator when an expansion was forced to fail,
4979 contradicting the general specification for all authenticators. Instead it
4980 was generating a temporary error. It now behaves as specified.
4982 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4983 (the order specifies the preference for clients). The order is now AES256,
4984 AES128, 3DES, ARCFOUR128.
4986 51. Small patch to Sieve code - explicitly set From: when generating an
4989 52. Exim crashed if a remote delivery caused a very long error message to be
4990 recorded - for instance if somebody sent an entire SpamAssassin report back
4991 as a large number of 550 error lines. This bug was coincidentally fixed by
4992 increasing the size of one of Exim's internal buffers (big_buffer) that
4993 happened as part of the Exiscan merge. However, to be on the safe side, I
4994 have made the code more robust (and fixed the comments that describe what
4997 53. Now that there can be additional text after "Completed" in log lines (if
4998 the queue_time_overall log selector is set), a one-byte patch to exigrep
4999 was needed to allow it to recognize "Completed" as not the last thing in
5002 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5003 patch that reportedly fixes this has been added. I am not expert enough to
5004 create a test for it. This is what the patch creator wrote:
5006 "I found a little strange behaviour of ldap code when working with
5007 Windows 2003 AD Domain, where users was placed in more than one
5008 Organization Units. When I tried to give exim partial DN, the exit code
5009 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5010 But simultaneously result of request was absolutely normal ldap result,
5011 so I produce this patch..."
5013 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5014 REFERENCE, so I have modified the code to exclude the patch when that macro
5017 55. Some experimental protocols are using DNS PTR records for new purposes. The
5018 keys for these records are domain names, not reversed IP addresses. The
5019 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5020 leaves it alone. Component reversal etc. now happens only for IP addresses.
5023 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5025 57. Double the size of the debug message buffer (to 2048) so that more of very
5026 long debug lines gets shown.
5028 58. The exicyclog utility now does better if the number of log files to keep
5029 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5031 59. Two changes related to the smtp_active_hostname option:
5033 (1) $smtp_active_hostname is now available as a variable.
5034 (2) The default for smtp_banner uses $smtp_active_hostname instead
5035 of $primary_hostname.
5037 60. The host_aton() function is supposed to be passed a string that is known
5038 to be a valid IP address. However, in the case of IPv6 addresses, it was
5039 not checking this. This is a hostage to fortune. Exim now panics and dies
5040 if the condition is not met. A case was found where this could be provoked
5041 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5042 components; fortuitously, this particular loophole had already been fixed
5043 by change 4.50/55 above.
5045 If there are any other similar loopholes, the new check in host_aton()
5046 itself should stop them being exploited. The report I received stated that
5047 data on the command line could provoke the exploit when Exim was running as
5048 exim, but did not say which command line option was involved. All I could
5049 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5050 running as the user.
5053 61. There was a buffer overflow vulnerability in the SPA authentication code
5054 (which came originally from the Samba project). I have added a test to the
5055 spa_base64_to_bits() function which I hope fixes it.
5058 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5059 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5061 63. The daemon start-up calls getloadavg() while still root for those OS that
5062 need the first call to be done as root, but it missed one case: when
5063 deliver_queue_load_max is set with deliver_drop_privilege. This is
5064 necessary for the benefit of the queue runner, because there is no re-exec
5065 when deliver_drop_privilege is set.
5067 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5068 This has been fixed.
5070 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5071 was in use, was not putting the data itself into the right store pool;
5072 consequently, it could be overwritten for a subsequent message in the same
5073 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5076 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5078 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5079 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5080 the function were treating the return as a boolean value, which happened to
5081 work because 0=false and not-0=true, but is not correct code.
5083 68. The host_aton() function was not handling scoped IPv6 addresses (those
5084 with, for example, "%eth0" on the end) correctly.
5086 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5087 negated items (that is, ~something) in unsigned ints. Some compilers
5088 apparently mutter when there is no cast.
5090 70. If an address verification called from an ACL failed, and did not produce a
5091 user-specific message (i.e. there was only a "system" message), nothing was
5092 put in $acl_verify_message. In this situation, it now puts the system
5095 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5096 session; change 4.31/43 added the unwanted input to the log line - except
5097 that it did not do this in the start of session case. It now does.
5099 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5100 This is wrong and can cause the other end to generate a synchronization
5101 error if it is another Exim or anything else that does the synchronization
5102 check. A QUIT command is no longer sent after a timeout.
5104 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5105 during host lookups.
5107 74. The defer_ok option of callout verification was not working if it was used
5108 when verifying addresses in header lines, that is, for this case:
5110 verify = header_sender/callout=defer_ok
5112 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5113 those file descriptors could be used for SMTP connections. If anything
5114 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5115 could be sent to the SMTP client, causing chaos. The daemon now opens
5116 stdin, stdout, and stderr to /dev/null when it puts itself into the
5119 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5120 log by default. The user can override this with suitable Perl magic.
5122 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5123 the log message when discard triggers, was not working for the DATA ACL or
5124 for the non-SMTP ACL.
5126 78. Error message wording change in sieve.c.
5128 79. If smtp_accept_max_per_host was set, the number of connections could be
5129 restricted to fewer than expected, because the daemon was trying to set up
5130 a new connection before checking whether the processes handling previous
5131 connections had finished. The check for completed processes is now done
5132 earlier. On busy systems, this bug wouldn't be noticed because something
5133 else would have woken the daemon, and it would have reaped the completed
5136 80. If a message was submitted locally by a user whose login name contained one
5137 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5138 It caused a spool format error. I have fixed the spool reading code. A
5139 related problem was that the "from" clause in the Received: line became
5140 illegal because of the space(s). It is now covered by ${quote_local_part.
5142 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5145 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5147 83. Patch from Sieve maintainer to fix unterminated string problem in
5148 "vacation" handling.
5150 84. Some minor changes to the Linux configuration files to help with other
5151 OS variants using glibc.
5153 85. One more patch for Sieve to update vacation handling to latest spec.
5156 ----------------------------------------------------
5157 See the note above about the 4.44 and 4.50 releases.
5158 ----------------------------------------------------
5164 1. Change 4.43/35 introduced a bug that caused file counts to be
5165 incorrectly computed when quota_filecount was set in an appendfile
5168 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5169 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5172 3. Give more explanation in the error message when the command for a transport
5173 filter fails to execute.
5175 4. There are several places where Exim runs a non-Exim command in a
5176 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5177 was being done only for the command run by the queryprogram router. It is
5178 now done for all such subprocesses. The other cases are: ${run, transport
5179 filters, and the commands run by the lmtp and pipe transports.
5181 5. Some older OS have a limit of 256 on the maximum number of file
5182 descriptors. Exim was using setrlimit() to set 1000 as a large value
5183 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5184 systems. I've change it so that if it can't get 1000, it tries for 256.
5186 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5187 was an oversight, and furthermore, ever since the addition of extra
5188 controls (e.g. 4.43/32), the checks on when to allow different forms of
5189 "control" were broken. There should now be diagnostics for all cases when a
5190 control that does not make sense is encountered.
5192 7. $recipients is now available in the predata ACL (oversight).
5194 8. Tidy the search cache before the fork to do a delivery from a message
5195 received from the command line. Otherwise the child will trigger a lookup
5196 failure and thereby defer the delivery if it tries to use (for example) a
5197 cached ldap connection that the parent has called unbind on.
5199 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5200 of $address_data from the recipient verification was clobbered by the
5201 sender verification.
5203 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5204 was its contents. (It was OK if the option was not defined at all.)
5206 11. A "Completed" log line is now written for messages that are removed from
5207 the spool by the -Mrm option.
5209 12. $host_address is now set to the target address during the checking of
5210 ignore_target_hosts.
5212 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5213 being passed; this would have caused $sender_host_name to have been used if
5214 matching the list had actually called for a host name (not very likely,
5215 since this list is usually IP addresses). A host name is now passed as
5218 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5219 code that specifies a non-restarting handler (typically sigaction() in
5220 modern systems) in an attempt to fix a rare and obscure crash bug.
5222 15. Narrowed the window for a race in the daemon that could cause it to ignore
5223 SIGCHLD signals. This is not a major problem, because they are used only to
5224 wake it up if nothing else does.
5226 16. A malformed maildirsize file could cause Exim to calculate negative values
5227 for the mailbox size or file count. Odd effects could occur as a result.
5228 The maildirsize information is now recalculated if the size or filecount
5231 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5232 support for a long time. Removed HAVE_SYS_VFS_H.
5234 18. Updated exipick to current release from John Jetmore.
5236 19. Allow an empty sender to be matched against a lookup in an address list.
5237 Previously the only cases considered were a regular expression, or an
5240 20. Exim went into a mad DNS lookup loop when doing a callout where the
5241 host was specified on the transport, if the DNS lookup yielded more than
5244 21. The RFC2047 encoding function was originally intended for short strings
5245 such as real names; it was not keeping to the 75-character limit for
5246 encoded words that the RFC imposes. It now respects the limit, and
5247 generates multiple encoded words if necessary. To be on the safe side, I
5248 have increased the buffer size for the ${rfc2047: expansion operator from
5251 22. Failure to deliver a bounce message always caused it to be frozen, even if
5252 there was an errors_to setting on the router. The errors_to setting is now
5255 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5256 canonical form (fully expanded) before being placed in
5257 $sender_host_address.
5259 24. Updated eximstats to version 1.33
5261 25. Include certificate and key file names in error message when GnuTLS fails
5262 to set them up, because the GnuTLS error message doesn't include the name
5263 of the failing file when there is a problem reading it.
5265 26. Expand error message when OpenSSL has problems setting up cert/key files.
5268 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5269 (this can affect the format of dates).
5271 28. exim_tidydb, when checking for the continued existence of a message for
5272 which it has found a message-specific retry record, was not finding
5273 messages that were in split spool directories. Consequently, it was
5274 deleting retry records that should have stayed in existence.
5276 29. eximstats updated to version 1.35
5277 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5278 1.35 - bugfix such that pie charts by volume are generated correctly
5280 30. The SPA authentication driver was not abandoning authentication and moving
5281 on to the next authenticator when an expansion was forced to fail,
5282 contradicting the general specification for all authenticators. Instead it
5283 was generating a temporary error. It now behaves as specified.
5285 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5286 (the order specifies the preference for clients). The order is now AES256,
5287 AES128, 3DES, ARCFOUR128.
5289 31. Small patch to Sieve code - explicitly set From: when generating an
5292 32. Exim crashed if a remote delivery caused a very long error message to be
5293 recorded - for instance if somebody sent an entire SpamAssassin report back
5294 as a large number of 550 error lines. This bug was coincidentally fixed by
5295 increasing the size of one of Exim's internal buffers (big_buffer) that
5296 happened as part of the Exiscan merge. However, to be on the safe side, I
5297 have made the code more robust (and fixed the comments that describe what
5300 33. Some experimental protocols are using DNS PTR records for new purposes. The
5301 keys for these records are domain names, not reversed IP addresses. The
5302 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5303 leaves it alone. Component reversal etc. now happens only for IP addresses.
5306 34. The host_aton() function is supposed to be passed a string that is known
5307 to be a valid IP address. However, in the case of IPv6 addresses, it was
5308 not checking this. This is a hostage to fortune. Exim now panics and dies
5309 if the condition is not met. A case was found where this could be provoked
5310 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5311 components; fortuitously, this particular loophole had already been fixed
5312 by change 4.50/55 or 4.44/33 above.
5314 If there are any other similar loopholes, the new check in host_aton()
5315 itself should stop them being exploited. The report I received stated that
5316 data on the command line could provoke the exploit when Exim was running as
5317 exim, but did not say which command line option was involved. All I could
5318 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5319 running as the user.
5322 35. There was a buffer overflow vulnerability in the SPA authentication code
5323 (which came originally from the Samba project). I have added a test to the
5324 spa_base64_to_bits() function which I hope fixes it.
5327 36. The daemon start-up calls getloadavg() while still root for those OS that
5328 need the first call to be done as root, but it missed one case: when
5329 deliver_queue_load_max is set with deliver_drop_privilege. This is
5330 necessary for the benefit of the queue runner, because there is no re-exec
5331 when deliver_drop_privilege is set.
5333 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5334 was in use, was not putting the data itself into the right store pool;
5335 consequently, it could be overwritten for a subsequent message in the same
5336 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5339 38. Sometimes the final signoff response after QUIT could fail to get
5340 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5341 < 0 before doing a fflush(). This bug looks as though it goes back to the
5342 introduction of TLS in release 3.20, but "sometimes" must have been rare
5343 because the tests only now provoked it.
5349 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5350 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5351 Now it returns an integer. A number of calls were still expecting a T/F
5352 return. Fortuitously, in all cases, the tests worked in OK situations,
5353 which is the norm. However, things would have gone wrong on any write
5354 failures on the smtp file descriptor. This function is used when sending
5355 messages over SMTP and also when doing verify callouts.
5357 2. When Exim is called to do synchronous delivery of a locally submitted
5358 message (the -odf or -odi options), it no longer closes stderr before doing
5361 3. Implemented the mua_wrapper option.
5363 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5365 5. Implemented the functions header_remove(), header_testname(),
5366 header_add_at_position(), and receive_remove_recipient(), and exported them
5369 6. If an ACL "warn" statement specified the addition of headers, Exim already
5370 inserted X-ACL-Warn: at the start if there was no header name. However, it
5371 was not making this test for the second and subsequent header lines if
5372 there were newlines in the string. This meant that an invalid header could
5373 be inserted if Exim was badly configured.
5375 7. Allow an ACL "warn" statement to add header lines at the start or after all
5376 the Received: headers, as well as at the end.
5378 8. Added the rcpt_4xx retry error code.
5380 9. Added postmaster_mailfrom=xxx to callout verification option.
5382 10. Added mailfrom=xxxx to the callout verification option, for verify=
5385 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5386 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5388 12. Inserted some casts to stop certain compilers warning when using pointer
5389 differences as field lengths or precisions in printf-type calls (mostly
5390 affecting debugging statements).
5392 13. Added optional readline() support for -be (dynamically loaded).
5394 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5395 same clock tick as a message's arrival, so that its received time was the
5396 same as the "first fail" time on the retry record, and that message
5397 remained on the queue past the ultimate address timeout, every queue runner
5398 would try a delivery (because it was past the ultimate address timeout) but
5399 after another failure, the ultimate address timeout, which should have then
5400 bounced the address, did not kick in. This was a "< instead of <=" error;
5401 in most cases the first failure would have been in the next clock tick
5402 after the received time, and all would be well.
5404 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5405 being recognized when the domain list was tested by the match_domain
5406 condition in an expansion string.
5408 16. Added the ${str2b64: operator.
5410 17. Exim was always calling setrlimit() to set a large limit for the number of
5411 processes, without checking whether the existing limit was already
5412 adequate. (It did check for the limit on file descriptors.) Furthermore,
5413 errors from getrlimit() and setrlimit() were being ignored. Now they are
5414 logged to the main and panic logs, but Exim does carry on, to try to do its
5415 job under whatever limits there are.
5417 18. Imported PCRE 5.0.
5419 19. Trivial typo in log message " temporarily refused connection" (the leading
5422 20. If the log selector return_path_on_delivery was set and an address was
5423 redirected to /dev/null, the delivery process crashed because it assumed
5424 that a return path would always be set for a "successful" delivery. In this
5425 case, the whole delivery is bypassed as an optimization, and therefore no
5428 21. Internal re-arrangement: the function for sending a challenge and reading
5429 a response while authentication was assuming a zero-terminated challenge
5430 string. It's now changed to take a pointer and a length, to allow for
5431 binary data in such strings.
5433 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5435 23. Exim was not respecting finduser_retries when seeking the login of the
5436 uid under which it was called; it was always trying 10 times. (The default
5437 setting of finduser_retries is zero.) Also, it was sleeping after the final
5438 failure, which is pointless.
5440 24. Implemented tls_on_connect_ports.
5442 25. Implemented acl_smtp_predata.
5444 26. If the domain in control=submission is set empty, Exim assumes that the
5445 authenticated id is a complete email address when it generates From: or
5446 Sender: header lines.
5448 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5449 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5450 chown and chgrp in /bin and hostname in /usr/bin.
5452 28. Exim was keeping the "process log" file open after each use, just as it
5453 does for the main log. This opens the possibility of it remaining open for
5454 long periods when the USR1 signal hits a daemon. Occasional processlog
5455 errors were reported, that could have been caused by this. Anyway, it seems
5456 much more sensible not to leave this file open at all, so that is what now
5459 29. The long-running daemon process does not normally write to the log once it
5460 has entered its main loop, and it closes the log before doing so. This is
5461 so that log files can straightforwardly be renamed and moved. However,
5462 there are a couple of unusual error situations where the daemon does write
5463 log entries, and I had neglected to close the log afterwards.
5465 30. The text of an SMTP error response that was received during a remote
5466 delivery was being truncated at 512 bytes. This is too short for some of
5467 the long messages that one sometimes sees. I've increased the limit to
5470 31. It is now possible to make retry rules that apply only when a message has a
5471 specific sender, in particular, an empty sender.
5473 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5474 it possible to be selective about when SMTP synchronization is enforced.
5476 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5478 32. Implemented hosts_connection_nolog.
5480 33. Added an ACL for QUIT.
5482 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5485 35. Added mailbox_size and mailbox_filecount to appendfile.
5487 36. Added control = no_multiline_responses to ACLs.
5489 37. There was a bug in the logic of the code that waits for the clock to tick
5490 in the case where the clock went backwards by a substantial amount such
5491 that the microsecond fraction of "now" was more than the microsecond
5492 fraction of "then" (but the whole seconds number was less).
5494 38. Added support for the libradius Radius client library this is found on
5495 FreeBSD (previously only the radiusclient library was supported).
5501 1. When certain lookups returned multiple values in the form name=value, the
5502 quoting of the values was not always being done properly. Specifically:
5503 (a) If the value started with a double quote, but contained no whitespace,
5505 (b) If the value contained whitespace other than a space character (i.e.
5506 tabs or newlines or carriage returns) it was not quoted.
5507 This fix has been applied to the mysql and pgsql lookups by writing a
5508 separate quoting function and calling it from the lookup code. The fix
5509 should probably also be applied to nisplus, ibase and oracle lookups, but
5510 since I cannot test any of those, I have not disturbed their existing code.
5512 2. A hit in the callout cache for a specific address caused a log line with no
5513 reason for rejecting RCPT. Now it says "Previous (cached) callout
5514 verification failure".
5516 3. There was an off-by-one bug in the queryprogram router. An over-long
5517 return line was truncated at 256 instead of 255 characters, thereby
5518 overflowing its buffer with the terminating zero. As well as fixing this, I
5519 have increased the buffer size to 1024 (and made a note to document this).
5521 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5522 when Exim is waiting for an SMTP response from a remote server, Exim
5523 restarts its select() call on the socket, thereby resetting its timeout.
5524 This is not a problem when such interrupts are rare. Somebody set up a cron
5525 job to run exiwhat every 2 minutes, which is less than the normal select()
5526 timeout (5 or 10 minutes). This meant that the select() timeout never
5527 kicked in because it was always reset. I have fixed this by comparing the
5528 time when an interrupt arrives with the time at the start of the first call
5529 to select(). If more time than the timeout has elapsed, the interrupt is
5530 treated as a timeout.
5532 5. Some internal re-factoring in preparation for the addition of Sieve
5533 extensions (by MH). In particular, the "personal" test is moved to a
5534 separate function, and given an option for scanning Cc: and Bcc: (which is
5535 not set for Exim filters).
5537 6. When Exim created an email address using the login of the caller as the
5538 local part (e.g. when creating a From: or Sender: header line), it was not
5539 quoting the local part when it contained special characters such as @.
5541 7. Installed new OpenBSD configuration files.
5543 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5544 try to make them clearer.
5546 9. Callout options, other than the timeout value, were being ignored when
5547 verifying sender addresses in header lines. For example, when using
5549 verify = header_sender/callout=no_cache
5551 the cache was (incorrectly) being used.
5553 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5554 only the Cygwin environment.
5556 11. When return_path_on_delivery was set as a log selector, if different remote
5557 addresses in the same message used different return paths and parallel
5558 remote delivery occurred, the wrong values would sometimes be logged.
5559 (Whenever a remote delivery process finished, the return path value from
5560 the most recently started remote delivery process was logged.)
5562 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5563 lines when AUTH and/or TLS are in use. This is the "received protocol"
5564 field. Exim used to use "asmtp" for authenticated SMTP, without any
5565 indication (in the protocol name) for TLS use. Now it follows the RFC and
5566 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5567 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5568 names appear in log lines as well as in Received: header lines.
5570 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5571 extensions, and comparison tests, and to fix some bugs.
5573 14. Changes to the "personal" filter test:
5575 (1) The test was buggy in that it was just doing the equivalent of
5576 "contains" tests on header lines. For example, if a user's address was
5577 anne@some.where, the "personal" test would incorrectly be true for
5579 To: susanne@some.where
5581 This test is now done by extracting each address from the header in turn,
5582 and checking the entire address. Other tests that are part of "personal"
5583 are now done using regular expressions (for example, to check local parts
5584 of addresses in From: header lines).
5586 (2) The list of non-personal local parts in From: addresses has been
5587 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5588 taken from the Sieve specification recommendations.
5590 (3) If the message contains any header line starting with "List-" it is
5591 treated as non-personal.
5593 (4) The test for "circular" in the Subject: header line has been removed
5594 because it now seems ill-conceived.
5596 15. Minor typos in src/EDITME comments corrected.
5598 16. Installed latest exipick from John Jetmore.
5600 17. If headers_add on a router specified a text string that was too long for
5601 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5602 of string_sprintf() is now avoided.
5604 18. $message_body_size was not set (it was always zero) when running the DATA
5605 ACL and the local_scan() function.
5607 19. For the "mail" command in an Exim filter, no default was being set for
5608 the once_repeat time, causing a random time value to be used if "once" was
5609 specified. (If the value happened to be <= 0, no repeat happened.) The
5610 default is now 0s, meaning "never repeat". The "vacation" command was OK
5611 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5612 (I found it when inspecting the code).
5614 20. There is now an overall timeout for performing a callout verification. It
5615 defaults to 4 times the callout timeout, which applies to individual SMTP
5616 commands during the callout. The overall timeout applies when there is more
5617 than one host that can be tried. The timeout is checked before trying the
5618 next host. This prevents very long delays if there are a large number of
5619 hosts and all are timing out (e.g. when the network connections are timing
5620 out). The value of the overall timeout can be changed by specifying an
5621 additional sub-option for "callout", called "maxwait". For example:
5623 verify = sender/callout=5s,maxwait=20s
5625 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5626 to the end before writing, but this should make it even safer).
5628 22. Exim was forgetting that it had advertised PIPELINING for the second and
5629 subsequent messages on an SMTP connection. It was also not resetting its
5630 memory on STARTTLS and an internal HELO.
5632 23. When Exim logs an SMTP synchronization error within a session, it now
5633 records whether PIPELINING has been advertised or not.
5635 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5636 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5637 rather than long int.
5639 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5641 26. Added the never_mail option to autoreply.
5647 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5648 crash if the getsockname() call failed; this can happen if a connection is
5649 closed very soon after it is established. The problem was simply in the
5650 order in which certain operations were done, causing Exim to try to write
5651 to the SMTP stream before it had set up the file descriptor. The bug has
5652 been fixed by making things happen in the correct order.
5658 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5659 before the connection was closed, thus losing the rejection response.
5661 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5662 some early Solaris releases, but causes trouble in current releases where
5663 socklen_t is defined.
5665 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5668 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5671 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5672 that releasing the top of it at the end releases what was used for sub-
5673 expansions (unless the block got too big). However, discard this block if
5674 the first thing is a variable or header, so that we can use its block when
5675 it is dynamic (useful for very large $message_headers, for example).
5677 6. Lookups now cache *every* query, not just the most recent. A new, separate
5678 store pool is used for this. It can be recovered when all lookup caches are
5679 flushed. Lookups now release memory at the end of their result strings.
5680 This has involved some general refactoring of the lookup sources.
5682 7. Some code has been added to the store_xxx() functions to reduce the amount
5683 of flapping under certain conditions.
5685 8. log_incoming_interface used to affect only the <= reception log lines. Now
5686 it causes the local interface and port to be added to several more SMTP log
5687 lines, for example "SMTP connection from", and rejection lines.
5689 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5691 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5693 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5694 could be overwritten at the end of the current message (or the start of a
5695 new message if it was set in a HELO ACL). The value is now preserved for
5696 the duration of the SMTP connection.
5698 12. If a transport had a headers_rewrite setting, and a matching header line
5699 contained an unqualified address, that address was qualified, even if it
5700 did not match any rewriting rules. The underlying bug was that the values
5701 of the flags that permit the existence of unqualified sender and recipient
5702 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5703 non-local messages, and by -bnq for local messages) were not being
5704 preserved with the message after it was received.
5706 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5707 "next input=" as part of the text comprising the host identity instead of
5708 the correct text. The code was using the same buffer for two different
5709 strings. However, depending on which order the printing function evaluated
5710 its arguments, the bug did not always show up. Under Linux, for example, my
5711 test suite worked just fine.
5713 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5714 doesn't work with some older versions of Perl. It has been changed to "my",
5715 which in any case is probably the better facility to use.
5717 15. A really picky compiler found some instances of statements for creating
5718 error messages that either had too many or two few arguments for the format
5721 16. The size of the buffer for calls to the DNS resolver has been increased
5722 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5723 for addresses that have a lot of PTR records. This alleviates a problem; it
5724 does not fully solve it.
5726 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5727 buffer now truncates the list and logs the incident, which is the same
5728 action as happens when Exim is looking up a host name and its aliases.
5729 Previously in this situation something unpredictable would happen;
5730 sometimes it was "internal error: store_reset failed".
5732 18. If a server dropped the connection unexpectedly when an Exim client was
5733 using GnuTLS and trying to read a response, the client delivery process
5734 crashed while trying to generate an error log message.
5736 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5737 string, for example:
5739 warn message = H1: something\nH2: something
5741 the text was added as a single header line from Exim's point of view
5742 though it ended up OK in the delivered message. However, searching for the
5743 second and subsequent header lines using $h_h2: did not work. This has been
5744 fixed. Similarly, if a system filter added multiple headers in this way,
5745 the routers could not see them.
5747 20. Expanded the error message when iplsearch is called with an invalid key to
5748 suggest using net-iplsearch in a host list.
5750 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5751 ACLs are no longer actually imposed (and a message to that effect is
5754 22. If a "gecos" field in a passwd entry contained escaped characters, in
5755 particular, if it contained a \" sequence, Exim got it wrong when building
5756 a From: or a Sender: header from that name. A second bug also caused
5757 incorrect handling when an unquoted " was present following a character
5758 that needed quoting.
5760 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5761 was not being matched caselessly.
5763 24. Arranged for all hyphens in the exim.8 source to be escaped with
5766 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5767 properly. Recipient callout cache records were still being keyed to include
5768 the sender, even when use_sender was set false. This led to far more
5769 callouts that were necessary. The sender is no longer included in the key
5770 when use_sender is false.
5772 26. Added "control = submission" modifier to ACLs.
5774 27. Added the ${base62d: operator to decode base 62 numbers.
5776 28. dnsdb lookups can now access SRV records.
5778 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5779 the configuration file.
5781 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5782 (-v) mode. This makes the output for a verbose queue run more intelligible.
5784 31. Added a use_postmaster feature to recipient callouts.
5786 32. Added the $body_zerocount variable, containing the number of binary zero
5787 bytes in the message body.
5789 33. The time of last modification of the "new" subdirectory is now used as the
5790 "mailbox time last read" when there is a quota error for a maildir
5793 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5795 35. Added +ignore_unknown as a special item in host lists.
5797 36. Code for decoding IPv6 addresses in host lists is now included, even if
5798 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5799 address was recognized as an IP address, but was then not correctly decoded
5800 into binary, causing unexpected and incorrect effects when compared with
5807 1. Very minor rewording of debugging text in manualroute to say "list of
5808 hosts" instead of "hostlist".
5810 2. If verify=header_syntax was set, and a header line with an unqualified
5811 address (no domain) and a large number of spaces between the end of the
5812 name and the colon was received, the reception process suffered a buffer
5813 overflow, and (when I tested it) crashed. This was caused by some obsolete
5814 code that should have been removed. The fix is to remove it!
5816 3. When running in the test harness, delay a bit after writing a bounce
5817 message to get a bit more predictability in the log output.
5819 4. Added a call to search_tidyup() just before forking a reception process. In
5820 theory, someone could use a lookup in the expansion of smtp_accept_max_
5821 per_host which, without the tidyup, could leave open a database connection.
5823 5. Added the variables $recipient_data and $sender_data which get set from a
5824 lookup success in an ACL "recipients" or "senders" condition, or a router
5825 "senders" option, similar to $domain_data and $local_part_data.
5827 6. Moved the writing of debug_print from before to after the "senders" test
5830 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5831 problems for message scanning, either using a data ACL, or using
5832 local_scan() because the Received: header was not generated till after they
5833 were called (in order to set the time as the time of reception completion).
5834 I have revised the way this works. The header is now generated after the
5835 body is received, but before the ACL or local_scan() are called. After they
5836 are run, the timestamp in the header is updated.
5842 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5843 before starting a queue runner without re-exec. This happened only when
5844 deliver_drop_privilege was set or when the Exim user was set to root. The
5845 effect of the bug was that timeouts during subsequent deliveries caused
5846 crashes instead of being properly handled. The handler is now left at its
5847 default (and expected) setting.
5849 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5850 message, again when deliver_drop_privilege is set or Exim is run as root.
5851 The bug described in (1) was not present in this case, but the tidying up
5852 of the other signals was missing. I have made the two cases consistent.
5854 3. The ignore_target_hosts setting on a manualroute router was being ignored
5855 for hosts that were looked up using the /MX notation.
5857 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5860 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5861 operated on the sender address. After changing the $sender_address to <>
5862 for the sender address verify, Exim was re-instated it as the original
5863 (before rewriting) address, but remembering that it had rewritten it, so it
5864 wasn't rewriting it again. This bug also had the effect of breaking the
5865 sender address verification caching when the sender address was rewritten.
5867 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5868 This has been changed so that if the ip literal address matches
5869 ignore_target_hosts, the router declines.
5871 7. Added expansion conditions match_domain, match_address, and match_local_
5872 part (NOT match_host).
5874 8. The placeholder for the Received: header didn't have a length field set.
5876 9. Added code to Exim itself and to exim_lock to test for a specific race
5877 condition that could lead to file corruption when using MBX delivery. The
5878 issue is with the lockfile that is created in /tmp. If this file is removed
5879 after a process has opened it but before that process has acquired a lock,
5880 there is the potential for a second process to recreate the file and also
5881 acquire a lock. This could lead to two Exim processes writing to the file
5882 at the same time. The added code performs the same test as UW imapd; it
5883 checks after acquiring the lock that its file descriptor still refers to
5884 the same named file.
5886 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5887 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5890 11. Added the smtp_active_hostname option. If used, this will typically be made
5891 to depend on the incoming interface address. Because $interface_address is
5892 not set up until the daemon has forked a reception process, error responses
5893 that can happen earlier (such as "too many connections") no longer contain
5896 12. If an expansion in a condition on a "warn" statement fails because a lookup
5897 defers, the "warn" statement is abandoned, and the next ACL statement is
5898 processed. Previously this caused the whole ACL to be aborted.
5900 13. Added the iplsearch lookup type.
5902 14. Added ident_timeout as a log selector.
5904 15. Added tls_certificate_verified as a log selector.
5906 16. Added a global option tls_require_ciphers (compare the smtp transport
5907 option of the same name). This controls incoming TLS connections.
5909 17. I finally figured out how to make tls_require_ciphers do a similar thing
5910 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5911 before starting the TLS session.
5913 18. Tabs are now shown as \t in -bP output.
5915 19. If the log selector return_path_on_delivery was set, Exim crashed when
5916 bouncing a message because it had too many Received: header lines.
5918 20. If two routers both had headers_remove settings, and the first one included
5919 a superfluous trailing colon, the final name in the first list and the
5920 first name in the second list were incorrectly joined into one item (with a
5921 colon in the middle).
5927 1. Added -C and -D options to the exinext utility, mainly to make it easier
5928 to include in the automated testing, but these could be helpful when
5929 multiple configurations are in use.
5931 2. The exinext utility was not formatting the output nicely when there was
5932 an alternate port involved in the retry record key, nor when there was a
5933 message id as well (for retries that were specific to a specific message
5934 and a specific host). It was also confused by IPv6 addresses, because of
5935 the additional colons they contain. I have fixed the IPv4 problem, and
5936 patched it up to do a reasonable job for IPv6.
5938 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5939 delivery, the log line now contains "pipelined" if PIPELINING was used.
5941 4. An SMTP transport process used to panic and die if the bind() call to set
5942 an explicit outgoing interface failed. This has been changed; it is now
5943 treated in the same way as a connect() failure.
5945 5. A reference to $sender_host_name in the part of a conditional expansion
5946 that was being skipped was still causing a DNS lookup. This no longer
5949 6. The def: expansion condition was not recognizing references to header lines
5950 that used bh_ and bheader_.
5952 7. Added the _cache feature to named lists.
5954 8. The code for checking quota_filecount in the appendfile transport was
5955 allowing one more file than it should have been.
5957 9. For compatibility with Sendmail, the command line option
5965 and sets the incoming protocol and host name (for trusted callers). The
5966 host name and its colon can be omitted when only the protocol is to be set.
5967 Note the Exim already has two private options, -pd and -ps, that refer to
5968 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5969 "s", but I don't think that's a major issue.
5971 10. A number of refactoring changes to the code, none of which should affect
5974 (a) The number of logging options was getting close to filling up the
5975 32-bit word that was used as a bit map. I have split them into two classes:
5976 those that are passed in the argument to log_write(), and those that are
5977 only ever tested independently outside of that function. These are now in
5978 separate 32-bit words, so there is plenty of room for expansion again.
5979 There is no change in the user interface or the logging behaviour.
5981 (b) When building, for example, log lines, the code previously used a
5982 macro that called string_cat() twice, in order to add two strings. This is
5983 not really sufficiently general. Furthermore, there was one instance where
5984 it was actually wrong because one of the argument was used twice, and in
5985 one call a function was used. (As it happened, calling the function twice
5986 did not affect the overall behaviour.) The macro has been replaced by a
5987 function that can join an arbitrary number of extra strings onto a growing
5990 (c) The code for expansion conditions now uses a table and a binary chop
5991 instead of a serial search (which was left over from when there were very
5992 few conditions). Also, it now recognizes conditions like "pam" even when
5993 the relevant support is not compiled in: a suitably worded error message is
5994 given if an attempt is made to use such a condition.
5996 11. Added ${time_interval:xxxxx}.
5998 12. A bug was causing one of the ddress fields not to be passed back correctly
5999 from remote delivery subprocesses. The field in question was not being
6000 subsequently used, so this caused to problems in practice.
6002 13. Added new log selectors queue_time and deliver_time.
6004 14. Might have fixed a bug in maildirsizefile handling that threw up
6005 "unexpected character" debug warnings, and recalculated the data
6006 unnecessarily. In any case, I expanded the warning message to give more
6009 15. Added the message "Restricted characters in address" to the statements in
6010 the default ACL that block characters like @ and % in local parts.
6012 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6013 Three changes have been made:
6015 (a) There was a serious bug; a negative response to MAIL caused the whole
6016 recipient domain to be cached as invalid, thereby blocking all messages
6017 to all local parts at the same domain, from all senders. This bug has
6018 been fixed. The domain is no longer cached after a negative response to
6019 MAIL if the sender used is not empty.
6021 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6024 (c) A new callout option, "use_sender" has been added for people who want
6025 the modified behaviour.
6031 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6034 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6035 indeed breaks things for older releases.
6037 3. Added additional logging to the case where there is a problem reading data
6038 from a filter that is running in a subprocess using a pipe, in order to
6039 try to track down a specific problem.
6041 4. Testing facility fudge: when running in the test harness and attempting
6042 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6043 getting "No route to host". Convert this to a timeout.
6045 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6048 6. Some OS don't have socklen_t but use size_t instead. This affects the
6049 fifth argument of getsockopt() amongst other things. This is now
6050 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6051 can be set for individual OS. I have set it for SunOS5, OSF1, and
6052 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6053 some earlier ones do not.
6055 7. Change 4.30/15 was not doing the test caselessly.
6057 8. The standard form for an IPv6 address literal was being rejected by address
6058 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6059 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6060 this, as well as the form without the "IPv6" on the front (but only when
6061 address literals are enabled, of course).
6063 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6065 10. Exim crashed if a message with an empty sender address specified by -f
6066 encountered a router with an errors_to setting. This could be provoked only
6067 by a command such as
6071 where an empty string was supplied; "<>" did not hit this bug.
6073 11. Installed PCRE release 4.5.
6075 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6076 remained set. It is now erased.
6078 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6079 times from message ids (which are base 36 rather than the normal 62).
6081 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6082 were being counted as actual protocol errors, and logged if the log
6083 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6084 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6085 and DATA following a set of rejected RCPTs do not count as protocol errors.
6086 In other words, Exim assumes they were pipelined, though this may not
6087 actually be the case. Of course, in all cases the client gets an
6088 appropriate error code.
6090 15. If a lookup fails in an ACL condition, a message about the failure may
6091 be available; it is used if testing the ACL cannot continue, because most
6092 such messages specify what the cause of the deferral is. However, some
6093 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6094 that caused an old message to be retained and used if a later statement
6095 caused a defer, replacing the real cause of the deferral.
6097 16. If an IP address had so many PTR records that the DNS lookup buffer
6098 was not large enough to hold them, Exim could crash while trying to process
6099 the truncated data. It now detects and logs this case.
6101 17. Further to 4.21/58, another change has been made: if (and only if) the
6102 first line of a message (the first header line) ends with CRLF, a bare LF
6103 in a subsequent header line has a space inserted after it, so as not to
6104 terminate the header.
6106 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6107 unnecessarily, used atoi() instead of strtol(), and didn't check the
6108 termination when getting file sizes from file names by regex.
6110 19. Completely re-implemented the support for maildirsize files, in the light
6111 of a number of problems with the previous contributed implementation
6112 (4.30/29). In particular:
6114 . If the quota is zero, the maildirsize file is maintained, but no quota is
6117 . If the maildir directory does not exist, it is created before any attempt
6118 to write a maildirsize file.
6120 . The quota value in the file is just a cache; if the quota is changed in
6121 the transport, the new value overrides.
6123 . A regular expression is available for excluding directories from the
6126 20. The autoreply transport checks the characters in options that define the
6127 message's headers; it allows continued headers, but it was checking with
6128 isspace() after an embedded newline instead of explicitly looking for a
6131 21. If all the "regular" hosts to which an address was routed had passed their
6132 expiry times, and had not reached their retry times, the address was
6133 bounced, even if fallback hosts were defined. Now Exim should go on to try
6136 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6137 equivalent code in the SMTP transport. Some hosts send humungous responses
6138 to HELO/EHLO, more than 1024 it seems.
6140 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6141 gives compiler warnings in some environments. I've now done it "properly",
6144 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6145 (because of problems with the built-in one) was declared to return uschar *
6146 instead of char *, causing compiler failure.
6148 25. Fixed a file descriptor leak when processing alias/forward files.
6150 26. Fixed a minor format string issue in dbfn.c.
6152 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6154 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6155 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6158 29. When a sender address is verified, it is cached, to save repeating the test
6159 when there is more than one recipient in a message. However, when the
6160 verification involves a callout, it is possible for different callout
6161 options to be set for different recipients. It is too complicated to keep
6162 track of this in the cache, so now Exim always runs a verification when a
6163 callout is required, relying on the callout cache for the optimization.
6164 The overhead is duplication of the address routing, but this should not be
6167 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6168 to be verified with callout=postmaster, and the main callout worked but the
6169 postmaster check failed, the verification correctly failed. However, if a
6170 subsequent RCPT command asked for sender verification *without* the
6171 postmaster check, incorrect caching caused this verification also to fail,
6174 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6175 it was not caching the DNS options (qualify_single, search_parents) that
6176 were used when the lookup failed. A subsequent lookup with different
6177 options therefore always gave the same answer, though there were cases
6178 where it should not have. (Example: a "domains = !$mx_any" option on a
6179 dnslookup router: the "domains" option is always processed without any
6180 widening, but the router might have qualify_single set.) Now Exim uses the
6181 cached value only when the same options are set.
6183 32. Added John Jetmore's "exipick" utility to the distribution.
6185 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6186 than a timeout (e.g. a certificate is required, and is not provided), an
6187 Exim server now closes the connection immediately. Previously it waited for
6188 the client to close - but if the client is SSL, it seems that they each
6189 wait for each other, leading to a delay before one of them times out.
6191 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6192 maintained 0.8.x compatibility because I don't think many are using it, and
6193 it is clearly obsolete.
6195 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6198 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6199 client certificate was expired. A simple patch fixes this, though I don't
6200 understand the full logic of why the verify callback is called multiple
6203 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6204 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6205 which causes problems with some clients (such as the Certicom SSL Plus
6206 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6207 disables the coutermeasure allowing Eudora to connect."
6209 38. Exim was not checking that a write() to a log file succeeded. This could
6210 lead to Bad Things if a log got too big, in particular if it hit a file
6211 size limit. Exim now panics and dies if it cannot write to a log file, just
6212 as it does if it cannot open a log file.
6214 39. Modified OS/Makefile-Linux so that it now contains
6216 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6218 The two -D definitions ensure that Exim is compiled with large file
6219 support, which makes it possible to handle log files that are bigger than
6222 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6223 instance) a domain was checked against a named list that involved a lookup,
6224 causing $domain_data to be set, then another domain was checked against the
6225 same list, then the first domain was re-checked, the value of $domain_data
6226 after the final check could be wrong. In particular, if the second check
6227 failed, it could be set empty. This bug probably also applied to
6230 41. The strip_trailing_dot option was not being applied to the address given
6231 with the -f command-line option.
6233 42. The code for reading a message's header from the spool was incrementing
6234 $received_count, but never initializing it. This meant that the value was
6235 incorrect (doubled) while delivering a message in the same process in which
6236 it was received. In the most common configuration of Exim, this never
6237 happens - a fresh exec is done - but it can happen when
6238 deliver_drop_privilege is set.
6240 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6241 it now includes up to 150 characters of the unexpected data in the log
6244 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6245 and building data strings. The size of both of these buffers was 10 000
6246 bytes - far larger than anybody would *ever* want, thought I. Needless to
6247 say, somebody hit the limit. I have increased the maximum line length to
6248 20 000 and the maximum data length of concatenated lines to 100 000. I have
6249 also fixed two bugs, because there was no checking on these buffers. Tsk,
6250 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6251 buffer is too small.
6253 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6254 lsearch lookups. Now it does.
6256 46. When parsing a route_list item in a manualroute router, a fixed-length
6257 buffer was used for the list of hosts. I made this 1024 bytes long,
6258 thinking that nobody would ever have a list of hosts that long. Wrong.
6259 Somebody had a whole pile of complicated expansion conditions, and the
6260 string was silently truncated, leading to an expansion error. It turns out
6261 that it is easier to change to an unlimited length (owing to other changes
6262 that have happened since this code was originally written) than to build
6263 structure for giving a limitation error. The length of the item that
6264 expands into the list of hosts is now unlimited.
6266 47. The lsearch lookup could not handle data where the length of text line was
6267 more than 4095 characters. Such lines were truncated, leading to shortened
6268 data being returned. It should now handle lines of any length.
6270 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6271 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6274 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6275 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6276 no longer happens in comment lines. A list of replacements is now placed
6277 at the head of all of the source files, except those whose only change is
6278 to replace PERL_COMMAND in the very first #! line.
6280 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6281 messages on the queue, with a bottom-up merge sort, using code contributed
6282 by Michael Haardt. This should make operations like -bp somewhat faster on
6283 large queues. It won't affect queue runners, except when queue_run_in_order
6286 51. Installed eximstats 1.31 in the distribution.
6288 52. Added support for SRV lookups to the dnslookup router.
6290 53. If an ACL referred to $message_body or $message_body_end, the value was not
6291 reset for any messages that followed in the same SMTP session.
6293 54. The store-handling optimization for building very long strings was not
6294 differentiating between the different store pools. I don't think this
6295 actually made any difference in practice, but I've tidied it.
6297 55. While running the routers to verify a sender address, $sender_address
6298 was still set to the sender address. This is wrong, because when routing to
6299 send a bounce to the sender, it would be empty. Therefore, I have changed
6300 it so that, while verifying a sender address, $sender_address is set to <>.
6301 (There is no change to what happens when verifying a recipient address.)
6303 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6304 target A or AAAA records (if not already returned) without resetting the
6305 qualify_single or search_parents options of the DNS resolver. These are
6306 inappropriate in this case because the targets of MX and SRV records must
6307 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6308 target that, when qualified, matched something in the local domain. These
6309 two options are now turned off when doing these lookups.
6311 57. It seems that at least some releases of Reiserfs (which does not have the
6312 concept of a fixed number of inodes) returns zero and not -1 for the
6313 number of available inodes. This interacted badly with check_spool_inodes,
6314 which assumed that -1 was the "no such thing" setting. What I have done is
6315 to check that the total number of inodes is greater than zero before doing
6316 the test of how many are available.
6318 58. When a "warn" ACL statement has a log_message modifier, the message is
6319 remembered, and not repeated. This is to avoid a lot of repetition when a
6320 message has many recipients that cause the same warning to be written.
6321 However, Exim was preserving the list of already written lines for an
6322 entire SMTP session, which doesn't seem right. The memory is now reset if a
6323 new message is started.
6325 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6326 debugging output unless log_rewrite was also set.
6328 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6329 of (int)(handle) when we know that handle contains (void *)(-1).
6331 61. The Exim daemon panic-logs an error return when it closes the incoming
6332 connection. However "connection reset by peer" seems to be common, and
6333 isn't really an error worthy of noting specially, so that particular error
6336 62. When Exim is trying to find all the local interfaces, it used to panic and
6337 die if the ioctl to get the interface flags failed. However, it seems that
6338 on at least one OS (Solaris 9) it is possible to have an interface that is
6339 included in the list of interfaces, but for which you get a failure error
6340 for this call. This happens when the interface is not "plumbed" into a
6341 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6342 failure of the "get flags" call assumes that the interface is down.
6344 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6345 makes life easier for people who are doing arithmetic on fields extracted
6346 from dates, where you often get leading zeros that should not be
6347 interpreted as octal.
6349 64. Added qualify_domain to the redirect router, to override the global
6352 65. If a pathologically long header line contained very many addresses (the
6353 report of this problem mentioned 10 000) and each of them was rewritten,
6354 Exim could use up a very large amount of memory. (It kept on making new
6355 copies of the header line as it rewrote, and never released the old ones.)
6356 At the expense of a bit more processing, the header rewriting function has
6357 been changed so that it no longer eats memory in this way.
6359 66. The generation of the Received: header has been moved from the time that a
6360 message starts to be received, to the time that it finishes. The timestamp
6361 in the Received: header should now be very close to that of the <= log
6362 line. There are two side-effects of this change:
6364 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6365 logged header lines no longer include the local Received: line, because
6366 it has not yet been created. The same applies to a copy of the message
6367 that is returned to a non-SMTP sender when a message is rejected.
6369 (b) When a filter file is tested using -bf, no additional Received: header
6370 is added to the test message. After some thought, I decided that this
6373 This change does not affect the value of $received_for. It is still set
6374 after address rewriting, but before local_scan() is called.
6376 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6378 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6379 gave an unhelpful panic error message, and a defer error. I have managed to
6380 change this behaviour so that it now rejects any supplied certificate,
6381 which seems right, as the list of acceptable certificates is empty.
6383 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6384 gave an unhelpful defer error. I have not managed to make this reject any
6385 supplied certificates, but the error message it gives is "no certificate
6386 supplied", which is not helpful.
6388 70. exigrep's output now also includes lines that are not associated with any
6389 message, but which match the given pattern. Implemented by a patch from
6390 Martin Sluka, which also tidied up the Perl a bit.
6392 71. Recipient callout verification, like sender verification, was using <> in
6393 the MAIL FROM command. This isn't really the right thing, since the actual
6394 sender may affect whether the remote host accepts the recipient or not. I
6395 have changed it to use the actual sender in the callout; this means that
6396 the cache record is now keyed on a recipient/sender pair, not just the
6397 recipient address. There doesn't seem to be a real danger of callout loops,
6398 since a callout by the remote host to check the sender would use <>.
6399 [SEE ABOVE: changed after hitting problems.]
6401 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6402 temporary errors. However, in the case of such a code being given after
6403 the end of a data transmission (i.e. after ".") Exim was failing to write
6404 a retry record for the message. (Yes, there was some broken host that was
6405 actually sending 8xx at this point.)
6407 73. An unknown lookup type in a host list could cause Exim to panic-die when
6408 the list was checked. (An example that provoked this was putting <; in the
6409 middle of a list instead of at the start.) If this happened during a DATA
6410 ACL check, a -D file could be left lying around. This kind of configuration
6411 error no longer causes Exim to die; instead it causes a defer error. The
6412 incident is still logged to the main and panic logs.
6414 74. Buglet left over from Exim 3 conversion. The message "too many messages
6415 in one connection" was written to the rejectlog but not the mainlog, except
6416 when address rewriting (yes!) was being logged.
6418 75. Added write_rejectlog option.
6420 76. When a system filter was run not as root (that is, when system_filter_user
6421 was set), the values of the $n variables were not being returned to the
6422 main process; thus, they were not subsequently available in the $sn
6425 77. Added +return_path_on_delivery log selector.
6427 78. A connection timeout was being treated differently from recipients deferred
6428 when testing hosts_max_try with a message that was older than the host's
6429 retry timeout. (The host should not be counted, thus allowing all hosts to
6430 be tried at least once before bouncing.) This may have been the cause of an
6431 occasionally reported bug whereby a message would remain on the queue
6432 longer than the retry timeout, but would be bounced if a delivery was
6433 forced. I say "may" because I never totally pinned down the problem;
6434 setting up timeout/retry tests is difficult. See also the next item.
6436 79. The ultimate address timeout was not being applied to errors that involved
6437 a combination of host plus message (for example, a timeout on a MAIL
6438 command). When an address resolved to a number of possible hosts, and they
6439 were not all tried for each delivery (e.g. because of hosts_max_try), a
6440 message could remain on the queue longer than the retry timeout.
6442 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6445 81. Fixed an obscure SMTP outgoing bug which required at least the following
6446 conditions: (a) there was another message waiting for the same server;
6447 (b) the server returned 5xx to all RCPT commands in the first message so
6448 that the message was not completed; (c) the server dropped the connection
6449 or gave a negative response to the RSET that Exim sends to abort the
6450 transaction. The observed case was a dropped connection after DATA that had
6451 been sent in pipelining mode. That is, the server had advertised PIPELINING
6452 but was not implementing it correctly. The effect of the bug was incorrect
6453 behaviour, such as trying another host, and this could lead to a crash.
6459 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6460 and daemon.c were passed as pointers to ints; they should have been
6461 pointers to socklen_t variables (which are typically unsigned ints).
6463 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6466 3. Fixed a really odd bug that affected only the testing scheme; patching a
6467 certain fixed string in the binary changed the value of another string that
6468 happened to be identical to the end of the original first string.
6470 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6471 name", it returns that address as the IP address. On some operating
6472 systems (e.g. Solaris), it also passes back the IP address string as the
6473 "host name". However, on others (e.g. Linux), it passes back an empty
6474 string. Exim wasn't checking for this, and was changing the host name to an
6475 empty string, assuming it had been canonicalized.
6477 5. Although rare, it is permitted to have more than one PTR record for a given
6478 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6479 all the names associated with an address, because they do in Solaris.
6480 However, it seems that they do not in Linux for data that comes from the
6481 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6482 I found this out when I moved to a new Linux workstation and tried to run
6483 the Exim test suite.
6485 To get round this problem I have changed the code so that it now does its
6486 own call to the DNS to look up PTR records when searching for a host name.
6487 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6488 addresses that are only in /etc/hosts are still found.
6490 This behaviour is, however, controlled by an option called host_lookup_
6491 order, which defaults to "bydns:byaddr". If people want to use the other
6492 order, or indeed, just use one or the other means of lookup, they can
6493 specify it in this variable.
6495 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6496 some operating systems, this comes back from gethostbyaddr() as an empty
6497 string, and this is what Exim used to test for. However, it seems that in
6498 other systems, "." is yielded. Exim now tests for this case too.
6500 7. The values of check_spool_space and check_log_space are now held internally
6501 as a number of kilobytes instead of an absolute number of bytes. If a
6502 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6503 kilobyte. This means that much larger values can be stored.
6505 8. Exim monitor: an attempt to get the action menu when not actually pointing
6506 at a message produces an empty menu entitled "No message selected". This
6507 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6508 no entries in it ("Shell widget menu has zero width and/or height"). So I
6509 have added a single, blank menu entry in this case.
6511 9. Added ${quote_local_part.
6513 10. MIME decoding is now applied to the contents of Subject: header lines when
6516 11. Now that a reference to $sender_host_address automatically causes a reverse
6517 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6518 host lookup before query-style lookups in lists that might use this
6519 variable. This has therefore been abolished, and the "net-" prefix is no
6520 longer necessary for query-style lookups.
6522 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6523 have been a typo for LFLAGS, so it has been changed.
6525 13. The install script calls Exim with "-C /dev/null" in order to find the
6526 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6527 to be output. However, since Exim outputs its version number before the
6528 error, it didn't break the script. It just looked ugly. I fixed this by
6529 always allowing "-C /dev/null" if the caller is root.
6531 14. Ignore overlarge ACL variable number when reading spool file - insurance
6532 against a later release with more variables having written the file.
6534 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6535 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6536 this, as well as the form without the "IPv6" on the front.
6538 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6539 OS/Makefile-Darwin file.
6541 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6542 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6544 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6545 variable. However, some OS use ints for inodes. I've added cast to long int
6546 to get rid of the compiler warning.
6548 19. I had forgotten to lock out "/../" in configuration file names when
6549 ALT_CONFIG_PREFIX was set.
6551 20. Routers used for verification do not need to specify transports. However,
6552 if such a router generated a host list, and callout was configured, Exim
6553 crashed, because it could not find a port number from the (non-existent)
6554 transport. It now assumes port 25 in this circumstance.
6556 21. Added the -t option to exigrep.
6558 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6559 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6560 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6561 them is set, LOOKUP_LSEARCH is forced.
6563 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6565 24. Added sender and host information to the "rejected by local_scan()" log
6566 line; previously there was no indication of these.
6568 25. Added .include_if_exists.
6570 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6571 Linux. It turns out that not all file systems support this. Apparently some
6572 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6573 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6574 not supported on the file descriptor, is now ignored when Exim is trying to
6575 sync a directory. This applies only to Linux.
6577 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6579 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6580 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6583 29. Added support for maildirsize files from supplied patch (modified a bit).
6585 30. The use of :fail: followed by an empty string could lead Exim to respond to
6586 sender verification failures with (e.g.):
6588 550 Verification failed for <xxx>
6589 550 Sender verify failed
6591 where the first response line was missing the '-' that indicates it is not
6592 the final line of the response.
6594 31. The loop for finding the name of the user that called Exim had a hardwired
6595 limit of 10; it now uses the value of finduser_retries, which is used for
6596 all other user lookups.
6598 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6601 33. Exim was neglecting to zero errno before one call of strtol() when
6602 expanding a string and expecting an integer value. On some systems this
6603 resulted in spurious "integer overflow" errors. Also, it was casting the
6604 result into an int without checking.
6606 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6607 did not work. The code looks as if it has *never* worked, though it appears
6608 to have been documented since at least release 1.62. I have made it work.
6610 35. The "timeout_DNS" error in retry rules, also documented since at least
6611 1.62, also never worked. As it isn't clear exactly what this means, and
6612 clearly it isn't a major issue, I have abolished the feature by treating it
6613 as "timeout", and writing a warning to the main and panic logs.
6615 36. The display of retry rules for -brt wasn't always showing the error code
6618 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6619 timeout_connect_A, timeout_connect_MX.
6621 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6622 to the empty sender.
6624 39. The daemon was not analysing the content of -oX till after it had closed
6625 stderr and disconnected from the controlling terminal. This meant that any
6626 syntax errors were only noted on the panic log, and the return code from
6627 the command was 0. By re-arranging the code a little, I've made the
6628 decoding happen first, so such errors now appear on stderr, and the return
6629 code is 1. However, the actual setting up of the sockets still happens in
6630 the disconnected process, so errors there are still only recorded on the
6633 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6634 connections (as happens on some IP stacks) was logged at start up time as
6635 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6636 it from "IPv6 and IPv4", which means that two separate sockets are being
6639 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6640 says whether AF_INET or AF_INET6 was passed as an argument.
6642 42. Exiwhat output was messed up when time zones were included in log
6645 43. Exiwhat now gives more information about the daemon's listening ports,
6646 and whether -tls-on-connect was used.
6648 44. The "port" option of the smtp transport is now expanded.
6650 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6651 silently ignored. Now an error message is written to the main and panic
6654 46. There's a new ACL modifier called "logwrite" which writes to a log file
6655 as soon as it is encountered.
6657 47. Added $local_user_uid and $local_user_gid at routing time.
6659 48. Exim crashed when trying to verify a sender address that was being
6662 49. Exim was recognizing only a space character after ".include". It now also
6663 recognizes a tab character.
6665 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6666 extracting the relevant information from the specification. The man page no
6667 longer contains scrambled data for the -d option, and I've added a section
6668 at the front about calling Exim under different names.
6670 51. Added "extra_headers" argument to the "mail" command in filter files.
6672 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6675 53. Installed eximstats 1.29.
6677 54. Added transport_filter_timeout as a generic transport option.
6679 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6680 Cc: header lines. This was required by RFC 822, but it not required by RFC
6683 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6684 incoming messages that did not have them. Now it does so only if the
6685 message originates locally, that is, if there is no associated remote host
6686 address. When Resent- header lines are present, this applies to the Resent-
6687 lines rather than the non-Resent- lines.
6689 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6690 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6692 58. Messages for configuration errors now include the name of the main
6693 configuration file - useful now that there may be more than one file in a
6694 list (.included file names were always shown).
6696 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6697 for those rare installations that do not start the daemon as root or run it
6698 setuid root. I've cut out the call to initgroups() if the daemon is not
6701 60. The Exim user and group can now be bound into the binary as text strings
6702 that are looked up at the start of Exim's processing.
6704 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6706 62. Added $mailstore_basename variable.
6708 63. Installed patch to sieve.c from Michael Haardt.
6710 64. When Exim failed to open the panic log after failing to open the main log,
6711 the original message it was trying to log was written to stderr and debug
6712 output, but if they were not available (the usual case in production), it
6713 was lost. Now it is written to syslog before the two lines that record the
6714 failures to open the logs.
6716 65. Users' Exim filters run in subprocesses under the user's uid. It is
6717 possible for a "deliver" command or an alias in a "personal" command to
6718 provoke an address rewrite. If logging of address rewriting is configured,
6719 this fails because the process is not running as root or exim. There may be
6720 a better way of dealing with this, but for the moment (because 4.30 needs
6721 to be released), I have disabled address rewrite logging when running a
6722 filter in a non-root, non-exim process.
6728 1. The buildconfig auxiliary program wasn't quoting the value set for
6729 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6730 not defined. This bug was masked in 4.22 by the effect that was fixed in
6733 2. Some messages that were rejected after a message id was allocated were
6734 shown as "incomplete" by exigrep. It no longer does this for messages that
6735 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6737 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6738 have allow_domain_literals set, the ID did not get logged in the <= line.
6739 Domain literals are now always recognized in Message-ID: header lines.
6741 4. The first argument for a ${extract expansion item is the key name or field
6742 number. Leading and trailing spaces in this item were not being ignored,
6743 causing some misleading effects.
6745 5. When deliver_drop_privilege was set, single queue runner processes started
6746 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6747 same command in the process it spins off) were not dropping privilege.
6749 6. When the daemon running as "exim" started a queue runner, it always
6750 re-executed Exim in the spun-off process. This is a waste of effort when
6751 deliver_drop_privilege is set. The new process now just calls the
6752 queue-runner function directly.
6758 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6761 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6762 ignored. Though the use of -oP was forcing the writing of a pid file, it
6763 was always written to the default place.
6765 3. If the message "no IP address found for host xxxx" is generated during
6766 incoming verification, it is now followed by identification of the incoming
6767 connection (so you can more easily find what provoked it).
6769 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6771 5. Added some features to "harden" Exim a bit more against certain attacks:
6773 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6774 be put in Local/Makefile. This is like the never_users runtime option,
6775 but it cannot be overridden. The default setting is "root".
6777 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6778 prefix string with which any file named in a -C command line option
6781 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6782 is retained for -C and -D only if the caller of Exim is root. Without
6783 it, the exim user may also use -C and -D and retain privilege.
6785 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6786 command line option is disabled.
6788 6. Macro names set by the -D option must start with an upper case letter, just
6789 like macro names defined in the configuration file.
6791 7. Added "dereference=" facility to LDAP.
6793 8. Two instances of the typo "uknown" in the source files are fixed.
6795 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6796 the Configure-Makefile script screwed up while processing it.
6798 10. Incorporated PCRE 4.4.
6800 11. The SMTP synchronization check was not operating right at the start of an
6801 SMTP session. For example, it could not catch a HELO sent before the client
6802 waited for the greeting. There is now a check for outstanding input at the
6803 point when the greeting is written. Because of the duplex, asynchronous
6804 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6805 way, but not yet received, when the check is performed.
6807 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6808 on TCP/IP sockets, because this apparently causes some broken clients to
6811 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6812 unchanged) from the Cygwin maintainer.
6814 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6815 was supported instead of testing for mbx. Effectively a typo.
6817 15. The spa authenticator server code was not checking that the input it
6818 received was valid base64.
6820 16. The debug output line for the "set" modifier in ACLs was not showing the
6821 name of the variable that was being set.
6823 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6825 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6826 lookup on its own. Something else was needed to trigger the lookup. For
6827 example, a match in host_lookup or the need for a host name in a host list.
6828 Now, if $sender_host_name is referenced and the host name has not yet been
6829 looked up, a lookup is performed. If the lookup fails, the variable remains
6830 empty, and $host_lookup_failed is set to "1".
6832 19. Added "eqi" as a case-independent comparison operator.
6834 20. The saslauthd authentication condition could segfault if neither service
6835 nor realm was specified.
6837 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6838 error message that was logged was misleading, and incoming SMTP
6839 connections were dropped. The message is now more accurate, and temporary
6840 errors are given to SMTP connections.
6842 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6843 (except RSET) with a 421 error, until QUIT is received. However, it was
6844 failing to send a response to QUIT.
6846 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6847 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6848 ACL is now run after the helo_try_verify_hosts code.
6850 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6851 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6852 case-independent, so other case variants are also recognized). Apparently
6853 some systems use these upper case variants.
6855 25. If more than two messages were waiting for the same host, and a transport
6856 filter was specified for the transport, Exim sent two messages over the
6857 same TCP/IP connection, and then failed with "socket operation on non-
6858 socket" when it tried to send the third.
6860 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6862 27. The extern definition of crypt16() in expand.c was not being excluded when
6863 the OS had its own crypt16() function.
6865 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6866 as a preferred synonym for return_size_limit, both as an option and as an
6869 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6871 30. Changed the default configuration ACL to relax the local part checking rule
6872 for addresses that are not in any local domains. For these addresses,
6873 slashes and pipe symbols are allowed within local parts, but the sequence
6874 /../ is explicitly forbidden.
6876 31. SPA server authentication was not clearing the challenge buffer before
6879 32. log_message in a "warn" ACL statement was writing to the reject log as
6880 well as to the main log, which contradicts the documentation and doesn't
6881 seem right (because no rejection is happening). So I have stopped it.
6883 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6884 However, I am unable to do any testing of this.
6886 34. Fixed an infelicity in the appendfile transport. When checking directories
6887 for a mailbox, to see if any needed to be created, it was accidentally
6888 using path names with one or more superfluous leading slashes; tracing
6889 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6891 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6892 contents are added to the log line that is written for every discarded
6893 recipient. (Previously a log_message setting was ignored.)
6895 36. The ${quote: operator now quotes the string if it is empty.
6897 37. The install script runs exim in order to find its version number. If for
6898 some reason other than non-existence or emptiness, which it checks, it
6899 could not run './exim', it was installing it with an empty version number,
6900 i.e. as "exim-". This error state is now caught, and the installation is
6903 38. An argument was missing from the function that creates an error message
6904 when Exim fails to connect to the socket for saslauthd authentication.
6905 This could cause Exim to crash, or give a corrupted message.
6907 39. Added isip, isip4, and isip6 to ${if conditions.
6909 40. The ACL variables $acl_xx are now saved with the message, and can be
6910 accessed later in routers, transports, and filters.
6912 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6913 strings in the file are not string-expanded.
6915 42. If a MAIL command specified a SIZE value that was too large to fit into an
6916 int variable, the check against message_size_limit failed. Such values are
6917 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6918 day this will have to be increased, but I don't think I want to be around
6919 when emails are that large.
6926 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6927 iconv() is not standard in FreeBSD.
6929 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6930 IPv6 enabled. The observed symptom was a segmentation fault on return from
6931 the function os_common_find_running_interfaces() in src/os.c.
6933 3. In the check_special_case() function in daemon.c I had used "errno" as an
6934 argument name, which causes warnings on some systems. This was basically a
6935 typo, since it was named "eno" in the comments!
6937 4. The code that waits for the clock to tick (at a resolution of some fraction
6938 of a second) so as to ensure message-id uniqueness was always waiting for
6939 at least one whole tick, when it could have waited for less. [This is
6940 almost certainly not relevant at current processor speeds, where it is
6941 unlikely to ever wait at all. But we try to future-proof.]
6943 5. The function that sleeps for a time interval that includes fractions of a
6944 second contained a race. It did not block SIGALRM between setting the
6945 timer, and suspending (a couple of lines later). If the interval was short
6946 and the sigsuspend() was delayed until after it had expired, the suspension
6947 never ended. On busy systems this could lead to processes getting stuck for
6950 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6951 process, before it forks any delivery processes. The open lookup caching
6952 mechanism meant that the open file or database connection was passed into
6953 the delivery process. The problem was that delivery processes always tidy
6954 up cached lookup data. This could cause a problem for the next delivery
6955 process started by the queue runner, because the external queue runner
6956 process does not know about the closure. So the next delivery process
6957 still has data in the lookup cache. In the case of a file lookup, there was
6958 no problem because closing a file descriptor in a subprocess doesn't affect
6959 the parent. However, if the lookup was caching a connection to a database,
6960 the connection was closed, and the second delivery process was likely to
6961 see errors such as "PGSQL: query failed: server closed the connection
6962 unexpectedly". The problem has been fixed by closing all cached lookups
6963 in a queue runner before running a delivery process.
6965 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6966 seem to have the "const" qualifier which it has on other OS. I've
6969 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6970 *of the same type* (client or server) with the same public name that an
6971 error should be diagnosed.
6973 9. When Exim looked up a host name for an IP address, but failed to find the
6974 original IP address when looking up the host name (a safety check), it
6975 output the message "<ip address> does not match any IP for NULL", which was
6976 confusing, to say the least. The bug was that the host name should have
6977 appeared instead of "NULL".
6979 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6980 user that is built into the binary, and the -C or -D options is used, root
6981 privilege is dropped before the configuration file is read. In addition,
6982 logging is switched to stderr instead of the normal log files. If the
6983 configuration then re-defines the Exim user, the unprivileged environment
6984 is probably not what is expected, so Exim logs a panic warning message (but
6987 However, if deliver_drop_privilege is set, the unprivileged state may well
6988 be exactly what is intended, so the warning has been cut out in that case,
6989 and Exim is allowed to try to write to its normal log files.
6995 1. smtp_return_error_details was not giving details for temporary sender
6996 or receiver verification errors.
6998 2. Diagnose a configuration error if two authenticators have the same public
7001 3. Exim used not to create the message log file for a message until the first
7002 delivery attempt. This could be confusing when incoming messages were held
7003 for policy or load reasons. The message log file is now created at the time
7004 the message is received, and an initial "Received" line is written to it.
7006 4. The automatically generated man page for command line options had a minor
7007 bug that caused no ill effects; however, a more serious problem was that
7008 the procedure for building the man page automatically didn't always
7009 operate. Consequently, release 4.20 contains an out-of-date version. This
7010 shouldn't happen again.
7012 5. When building Exim with embedded Perl support, the script that builds the
7013 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7014 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7016 6. The freeze_tell option was not being used for messages that were frozen on
7017 arrival, either by an ACL or by local_scan().
7019 7. Added the smtp_incomplete_transaction log selector.
7021 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7022 was accepting AUTH without a new EHLO.
7024 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7025 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7028 10. Logging of TCP/IP connections (when configured) now happens in the main
7029 daemon process instead of the child process, so that the TCP/IP connection
7030 count is more accurate (but it can never be perfect).
7032 11. The use of "drop" in a nested ACL was not being handled correctly in the
7033 outer ACL. Now, if condition failure induced by the nested "drop" causes
7034 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7035 or "require"), the connection is dropped.
7037 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7038 that yield "discard" can now be used with an "accept" or a "discard" verb,
7039 but an error is generated for any others (because I can't see a useful way
7040 to define what should happen).
7042 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7043 are now processed in the same way as lines in the Exim configuration file.
7044 In particular, continuation lines are supported.
7046 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7048 15. Added -ti meaning -t -i.
7050 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7051 domains, and warn by logging if others are found.
7053 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7054 returned from GIFCONF when Exim is trying to find the list of interfaces on
7055 a host. The code in os.c has been modified to copy each ifreq to an aligned
7056 structure in all cases.
7058 Also, in some cases, the returned ifreq's were being copied to a 'struct
7059 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7060 means the last couple of bytes of an IPv6 address could be chopped if the
7061 ifreq contained only a normal sockaddr (14 bytes storage).
7063 18. Named domain lists were not supported in the hosts_treat_as_local option.
7064 An entry such as +xxxx was not recognized, and was treated as a literal
7067 19. Ensure that header lines added by a DATA ACL are included in the reject log
7068 if the ACL subsequently rejects the message.
7070 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7071 MD5 (which is deprecated).
7073 21. When testing a filter file using -bf, Exim was writing a message when it
7074 took the sender from a "From " line in the message, but it was not doing so
7075 when it took $return_path from a Return-Path: header line. It now does.
7077 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7078 with a valid header line field name (a series of printing characters
7079 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7081 23. Changed "disc" in the source to "disk" to conform to the documentation and
7082 the book and for uniformity.
7084 24. Ignore Sendmail's -Ooption=value command line item.
7086 25. When execve() failed while trying to run a command in a pipe transport,
7087 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7088 could be confused with a return value of 69 from the command itself. This
7089 has been changed to 127, the value the shell returns if it is asked to run
7090 a non-existent command. The wording for the related log line suggests a
7091 non-existent command as the problem.
7093 26. If received_header_text expands to an empty string, do not add a Received:
7094 header line to the message. (Well, it adds a token one on the spool, but
7095 marks it "old" so that it doesn't get used or transmitted.)
7097 27. Installed eximstats 1.28 (addition of -nt option).
7099 28. There was no check for failure on the call to getsockname() in the daemon
7100 code. This can fail if there is a shortage of resources on the system, with
7101 ENOMEM, for example. A temporary error is now given on failure.
7103 29. Contrary to the C standard, it seems that in some environments, the
7104 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7105 program. Exim now does this explicitly; it affects the formatting of
7106 timestamps using strftime().
7108 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7109 complaints. I've now initialized all the variables, to avoid this.
7111 32. Header lines added by a system filter were not being "seen" during
7112 transport-time rewrites.
7114 33. The info_callback() function passed to OpenSSL is set up with type void
7115 (*)(SSL *, int, int), as described somewhere. However, when calling the
7116 function (actually a macro) that sets it up, the type void(*)() is
7117 expected. I've put in a cast to prevent warnings from picky compilers.
7119 34. If a DNS black list lookup found a CNAME record, but there were no A
7120 records associated with the domain it pointed at, Exim crashed.
7122 35. If a DNS black list lookup returned more than one A record, Exim ignored
7123 all but the first. It now scans all returned addresses if a particular IP
7124 value is being sought. In this situation, the contents of the
7125 $dnslist_value variable are a list of all the addresses, separated by a
7128 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7129 to accept a host name and all its aliases if the forward lookup for any of
7130 them yielded the IP address of the incoming connection. Now it accepts only
7131 those names whose forward lookup yields the correct IP address. Any other
7132 names are discarded. This closes a loophole whereby a rogue DNS
7133 administrator could create reverse DNS records to break through a
7134 wildcarded host restriction in an ACL.
7136 37. If a user filter or a system filter that ran in a subprocess used any of
7137 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7138 the wrong values were passed to the pipe command ($thisaddress had the
7139 value of $0, $0 had the value of $1, etc). This bug was introduced by
7140 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7142 38. Improved the line breaking for long SMTP error messages from ACLs.
7143 Previously, if there was no break point between 40 and 75 characters, Exim
7144 left the rest of the message alone. Two changes have been made: (a) I've
7145 reduced the minimum length to 35 characters; (b) if it can't find a break
7146 point between 35 and 75 characters, it looks ahead and uses the first one
7147 that it finds. This may give the occasional overlong line, but at least the
7148 remaining text gets split now.
7150 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7151 file descriptors might be low, and that setting 1000 would always raise it.
7152 It turns out that in some environments, the limit is already over 1000 and
7153 that lowering it causes trouble. So now Exim takes care not to decrease it.
7155 40. When delivering a message, the value of $return_path is set to $sender_
7156 address at the start of routing (routers may change the value). By an
7157 oversight, this default was not being set up when an address was tested by
7158 -bt or -bv, which affected the outcome if any router or filter referred to
7161 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7162 log only when "message" or "log_message" are set. However, if one of the
7163 conditions was an address verification, or a call to a nested ACL, the
7164 messages generated by the underlying test were being passed through. This
7165 no longer happens. The underlying message is available in $acl_verify_
7166 message for both "message" and "log_message" expansions, so it can be
7167 passed through if needed.
7169 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7170 new expansion $bh_ to give the encoded byte string without charset
7171 translation. Translation happens only if iconv() is available; HAVE_ICONV
7172 indicates this at build time. HEADERS_CHARSET gives the charset to
7173 translate to; headers_charset can change it in the configuration, and
7174 "headers charset" can change it in an individual filter file.
7176 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7177 that creates RFC 2047 encoded "words" labels them as that charset instead
7178 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7179 expansion operator; (ii) when Exim creates a From: line for a local
7180 message; (iii) when a header line is rewritten to include a "phrase" part.
7182 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7183 buggy, causing it to skip the first lines of messages whose message ID
7184 ended in 'D'. This would not have bitten before Exim release 4.14, because
7185 message IDs were unlikely to end in 'D' before then. The effect was to have
7186 incorrect size information for certain domains.
7188 45. #include "config.h" was missing at the start of the crypt16.c module. This
7189 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7192 46. If there was a timeout during a "random" callout check, Exim treated it as
7193 a failure of the random address, and carried on sending RSET and the real
7194 address. If the delay was just some slowness somewhere, the response to the
7195 original RCPT would be taken as a response to RSET and so on, causing
7196 mayhem of various kinds.
7198 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7199 when I implemented it. It didn't allow for the fact that some option values
7200 may legitimately be negative (e.g. size_addition), and it didn't even do
7201 the right test for positive values.
7203 48. Domain names in DNS records are case-independent. Exim always looks them up
7204 in lower case. Some resolvers return domain names in exactly the case they
7205 appear in the zone file, that is, they may contain uppercase letters. Not
7206 all resolvers do this - some return always lower case. Exim was treating a
7207 change of case by a resolver as a change of domain, similar to a widening
7208 of a domain abbreviation. This triggered its re-routing code and so it was
7209 trying to route what was effectively the same domain again. This normally
7210 caused routing to fail (because the router wouldn't handle the domain
7211 twice). Now Exim checks for this case specially, and just changes the
7212 casing of the domain that it ultimately uses when it transmits the message
7215 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7218 50. If a filter generated a file delivery with a non-absolute name (possible if
7219 no home directory exists for the router), the forbid_file option was not
7222 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7223 the existing equality matching.
7225 52. Exim was using ints instead of ino_t variables in some places where it was
7226 dealing with inode numbers.
7228 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7229 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7230 called TMPDIR, and if it finds it is different, it changes its value.
7232 54. The smtp_printf() function is now made available to local_scan() so
7233 additional output lines can be written before returning. There is also an
7234 smtp_fflush() function to enable the detection of a dropped connection.
7235 The variables smtp_input and smtp_batched_input are exported to
7238 55. Changed the default runtime configuration: the message "Unknown user"
7239 has been removed from the ACL, and instead placed on the localuser router,
7240 using the cannot_route_message feature. This means that any verification
7241 failures that generate their own messages won't get overridden. Similarly,
7242 the "Unrouteable address" message that was in the ACL for unverifiable
7243 relay addresses has also been removed.
7245 56. Added hosts_avoid_esmtp to the smtp transport.
7247 57. The exicyclog script was not checking for the esoteric option
7248 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7249 will work only if exicyclog is run under the appropriate euid.
7251 58. Following a discussion on the list, the rules by which Exim recognises line
7252 endings on incoming messages have been changed. The -dropcr and drop_cr
7253 options are now no-ops, retained only for backwards compatibility. The
7254 following line terminators are recognized: LF CRLF CR. However, special
7255 processing applies to CR:
7257 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7258 nor a local message in the state where . is a terminator.
7260 (ii) If a bare CR is encountered in a header line, an extra space is added
7261 after the line terminator so as not to end the header. The reasoning
7262 behind this is that bare CRs in header lines are most likely either
7263 to be mistakes, or people trying to play silly games.
7265 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7266 was being incorrectly given as 18 bytes larger than it should have been.
7267 This is a VOB (very old bug).
7269 60. This may never have affected anything current, but just in case it has:
7270 When the local host is found other than at the start of a list of hosts,
7271 the local host, those with the same MX, and any that follow, are discarded.
7272 When the list in question was part of a longer list of hosts, the following
7273 hosts (not currently being processed) were also being discarded. This no
7274 longer happens. I'm not sure if this situation could ever has previously
7277 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7280 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7281 header. This is something that is recommended in a new Internet Draft, and
7282 is something that is documented as being done by Sendmail. There are two
7283 possible values. For messages generated by the autoreply transport, Exim
7286 Auto-Submitted: auto-replied
7288 whereas for all other generated messages (e.g. bounces) it adds
7290 Auto-Submitted: auto-generated
7292 63. The "personal" condition in filters now includes a test for the
7293 Auto-Submitted: header. If it contains the string "auto-" the message it
7294 not considered personal.
7296 64. Added rcpt_include_affixes as a generic transport option.
7298 65. Added queue_only_override (default true).
7300 66. Added the syslog_duplication option.
7302 67. If what should have been the first header line of a message consisted of
7303 a space followed by a colon, Exim was mis-interpreting it as a header line.
7304 It isn't of course - it is syntactically invalid and should therefore be
7305 treated as the start of the message body. The misbehaviour could have
7306 caused a number of strange effects, including loss of data in subsequent
7307 header lines, and spool format errors.
7309 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7310 client host had authenticated. This control can now be exercised by an ACL
7311 for more flexibility.
7313 69. By default, callouts do not happen when testing with -bh. There is now a
7314 variant, -bhc, which does actually run the callout code, including
7315 consulting and updating the callout cache.
7317 70. Added support for saslauthd authentication, courtesy of Alexander
7320 71. If statvfs() failed on the spool or log directories while checking their
7321 size for availability, Exim confusingly gave the error "space shortage".
7322 Furthermore, in debugging mode it crashed with a floating point exception.
7323 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7324 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7325 serious problem, Exim now writes to the main and panic logs when this
7326 happens, with details of the failure. It then refuses to accept the
7327 incoming message, giving the message "spool directory problem" or "log
7328 directory problem" with a 421 code for SMTP messages.
7330 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7331 0, 1, and 2 exist, because some OS complain for execs without them (see
7332 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7333 descriptors. However, the code omitted to check that the open succeeded,
7334 causing mysterious errors if for some reason the permissions on /dev/null
7335 got screwed. Now Exim writes a message to the main and panic logs, and
7336 bombs out if it can't open /dev/null.
7338 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7339 interact so that it is all more flexible. It is supposed to remain
7340 backwards compatible. Also added extra_local_interfaces.
7342 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7343 to bomb out with an assertion failure - to the client this appears as a
7344 connection drop. This problem occurs in the part of the code that was taken
7345 from the Samba project. Fortunately, the assertion is in a very simple
7346 function, so I have fixed this by reproducing the function inline in the
7347 one place where it is called, and arranging for authentication to fail
7348 instead of killing the process with assert().
7350 75. The SPA client code was not working when the server requested OEM rather
7351 than Unicode encoding.
7353 76. Added code to make require_files with a specific uid setting more usable in
7354 the case where statting the file as root fails - usually a non-root-mounted
7355 NFS file system. When this happens and the failure is EACCES, Exim now
7356 forks a subprocess and does the per-uid checking as the relevant uid.
7358 77. Added process_log_path.
7360 78. If log_file_path was not explicitly set, a setting of check_log_space or
7361 check_log_inodes was ignored.
7363 79. If a space check for the spool or log partitions fails, the incident is now
7364 logged. Of course, in the latter case the data may get lost...
7366 80. Added the %p formatting code to string_format() so that it can be used to
7367 print addresses in debug_print(). Adjusted all the address printing in the
7368 debugging in store.c to use %p rather than %d.
7370 81. There was a concern that a line of code in smtp_in.c could overflow a
7371 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7372 initially expressed, the concern was not well-founded, because trailing
7373 spaces are removed early. However, if the trailing spaces were followed by
7374 a NULL, they did not get removed, so the overflow was possible. Two fixes
7377 (a) I re-wrote the offending code in a cleaner fashion.
7378 (b) If an incoming SMTP command contains a NULL character, it is rejected
7381 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7382 runs initgroups(), so that if the Exim user is in any additional groups,
7383 they will be used during message reception.
7389 The change log for 4.20 and earlier releases has been archived.