1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
12 buffer overrun for (non-chunking) other transports.
14 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
15 TLS1.3, means that a server rejecting a client certificate is not visible
16 to the client until the first read of encrypted data (typically the
17 response to EHLO). Add detection for that case and treat it as a failed
18 TLS connection attempt, so that the normal retry-in-clear can work (if
21 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
22 and/or domain. Found and fixed by Jason Betts.
24 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
25 configuration). If a CNAME target was not a wellformed name pattern, a
32 JH/01 Remove code calling the customisable local_scan function, unless a new
33 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
35 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
36 non-signal-safe functions being used.
38 JH/03 Bug 2269: When presented with a received message having a stupidly large
39 number of DKIM-Signature headers, disable DKIM verification to avoid
40 a resource-consumption attack. The limit is set at twenty.
42 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
43 report of oldest_pass in ${authres } in consequence, and separate out
44 some descriptions of reasons for verification fail.
46 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
47 files in the spool were present and unlocked. A queue-runner could spot
48 them, resulting in a duplicate delivery. Fix that by doing the unlock
49 after the unlink. Investigation by Tim Stewart. Take the opportunity to
50 add more error-checking on spoolfile handling while that code is being
53 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
54 No known attacks, no CVE, this is defensive hardening.
56 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
57 a queue-runner could start a delivery while other operations were ongoing.
58 Cutthrough delivery was a common victim, resulting in duplicate delivery.
59 Found and investigated by Tim Stewart. Fix by using the open message data
60 file handle rather than opening another, and not locally closing it (which
61 releases a lock) for that case, while creating the temporary .eml format
62 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
64 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
65 $sender_verify_failure/$recipient_verify_failure to "random".
67 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
70 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
71 Previously this would segfault.
73 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
76 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
77 like zero, since the resolver should be doing this for us, But we need one
78 as a CNAME but no MX presence gets the CNAME returned; we need to check
79 that doesn't point to an MX to declare it "no MX returned" rather than
80 "error, loop". A new main option is added so the older capability of
81 following some limited number of chain links is maintained.
83 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
85 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
86 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
87 now-deprecated earlier definitions used only specified the range up to TLS
88 1.2 (in the older-version library docs).
90 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
92 JH/15 Rework TLS client-side context management. Stop using a global, and
93 explicitly pass a context around. This enables future use of TLS for
94 connections to service-daemons (eg. malware scanning) while a client smtp
95 connection is using TLS; with cutthrough connections this is quite likely.
97 JH/16 Fix ARC verification to do AS checks in reverse order.
99 JH/17 Support a "tls" option on the ${readsocket } expansion item.
101 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
102 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
103 Previously the "utf8" would be re-prepended for every additional message.
105 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
106 Previously thery were accepted, resulting in issues when attempting to
107 forward messages to a non-supporting MTA.
109 PP/02 Let -n work with printing macros too, not just options.
111 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
112 one parent address was copied, and bogus data was used at delivery-logging
113 time. Either a crash (after delivery) or bogus log data could result.
114 Discovery and analysis by Tim Stewart.
116 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
117 Previously if the string ended mid-character, we did not insert the
118 promised '?' replacement.
120 PP/04 Documentation: current string operators work on bytes, not codepoints.
122 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
123 should pack well giving a smaller memory footprint so better caching and
124 therefore performance. Group the declarations where this can't be done so
125 that the byte-sized flag variables are not interspersed among pointer
126 variables, giving a better chance of good packing by the compiler.
128 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
129 non-null, to avoid issues with sites running BATV. Previously reports were
130 sent with an empty envelope sender so looked like bounces.
132 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
133 The ignore_error flag wasn't being returned from the filter subprocess so
134 was not set for later routers. Investigation and fix by Matthias Kurz.
136 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
137 and a msg:complete for the whole, when a message is manually removed using
138 -Mrm. Developement by Matthias Kurz, hacked on by JH.
140 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
141 a "Gnu special" function, asprintf() in the DB utility binary builds; I
142 hope that is portable enough.
144 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
145 requiring a known-CA anchor certificate; make it now rely entirely on the
146 TLSA as an anchor. Checking the name on the leaf cert against the name
147 on the A-record for the host is still done for TA (but not for EE mode).
149 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
150 included in delivery lines for non-proxied connections, when compiled with
151 SUPPORT_SOCKS and running with proxy logging enabled.
153 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
154 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
155 move the existing event to fire before the normal logging of message
156 failure so that custom logging is bracketed by normal logging.
158 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
159 msg:fail:internal event. Developement by Matthias Kurz.
161 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
162 far too small for todays use of crypto signatures stored there. Go all
163 the way to the max DNS message size of 64kB, even though this might be
164 overmuch for IOT constrained device use.
166 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
167 copy a string over itself. The library routine is documented as not
168 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
170 JH/32 For main options check_spool_space and check_inode_space, where the
171 platform supports 64b integers, support more than the previous 2^31 kB
172 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
173 the previous G, M, k.
175 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
176 $authenticated_fail_id variable on authentication failure. Previously
179 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
180 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
181 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
182 GNUTLS_SEC_PARAM_MEDIUM.
184 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
185 side. Previously we would continue as if no SNI had been received.
187 JH/36 Harden the handling of string-lists. When a list consisted of a sole
188 "<" character, which should be a list-separator specification, we walked
189 off past the nul-terimation.
191 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
192 causes) even when the retry time is not yet met. Previously they were
193 not, meaning that when (say) an account was over-quota and temp-rejecting,
194 and multiple senders' messages were queued, only one sender would get
195 notified on each configured delay_warning cycle.
197 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
199 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
200 cipher-suites, an error can be left on the stack even for a succeeding
201 accept; this results in impossible error messages when a later operation
204 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
205 return error codes indicating retry. Under TLS1.3 this becomes required.
207 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
208 it only wrote the new authenticators, resulting in a lack of tracking of
209 peer changes of ESMTP extensions until the next cache flush.
211 JH/41 Fix the loop reading a message header line to check for integer overflow,
212 and more-often against header_maxsize. Previously a crafted message could
213 induce a crash of the recive process; now the message is cleanly rejected.
215 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
216 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
222 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
223 When redis_servers is set to a list of > 1 element, and the Redis servers
224 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
225 case of MOVED into a DEFER case instead, thus moving the query onto the
226 next server in the list. For a cluster of N elements, all N servers must
227 be defined in redis_servers.
229 GF/02 Catch and remove uninitialized value warning in exiqsumm
230 Check for existence of @ARGV before looking at $ARGV[0]
232 JH/01 Replace the store_release() internal interface with store_newblock(),
233 which internalises the check required to safely use the old one, plus
234 the allocate and data copy operations duplicated in both (!) of the
235 extant use locations.
237 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
238 modifier. This matches the restriction on the commandline.
240 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
241 Previously only the last row was returned.
243 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
244 we assumed that tags in the header were well-formed, and parsed the
245 element content after inspecting only the first char of the tag.
246 Assumptions at that stage could crash the receive process on malformed
249 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
250 While running the DKIM ACL we operate on the Permanent memory pool so that
251 variables created with "set" persist to the DATA ACL. Also (at any time)
252 DNS lookups that fail create cache records using the Permanent pool. But
253 expansions release any allocations made on the current pool - so a dnsdb
254 lookup expansion done in the DKIM ACL releases the memory used for the
255 DNS negative-cache, and bad things result. Solution is to switch to the
256 Main pool for expansions.
257 While we're in that code, add checks on the DNS cache during store_reset,
258 active in the testsuite.
259 Problem spotted, and debugging aided, by Wolfgang Breyha.
261 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
262 When none of the hosts presented to a transport match an already-open
263 connection, close it and proceed with the list. Previously we would
264 queue the message. Spotted by Lena with Yahoo, probably involving
267 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
268 Previously a spurious "250 OK id=" response was appended to the proper
271 JH/08 The "support for" informational output now, which built with Content
272 Scanning support, has a line for the malware scanner interfaces compiled
273 in. Interface can be individually included or not at build time.
275 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
276 by the template makefile "src/EDITME". The "STREAM" support for an older
277 ClamAV interface method is removed.
279 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
280 rows affected is given instead).
282 JH/11 The runtime Berkeley DB library version is now additionally output by
283 "exim -d -bV". Previously only the compile-time version was shown.
285 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
286 SMTP connection. Previously, when one had more receipients than the
287 first, an abortive onward connection was made. Move to full support for
288 multiple onward connections in sequence, handling cutthrough connection
289 for all multi-message initiating connections.
291 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
292 routers. Previously, a multi-recipient message would fail to match the
293 onward-connection opened for the first recipient, and cause its closure.
295 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
296 a timeout on read on a GnuTLS initiating connection, resulting in the
297 initiating connection being dropped. This mattered most when the callout
298 was marked defer_ok. Fix to keep the two timeout-detection methods
301 JH/15 Relax results from ACL control request to enable cutthrough, in
302 unsupported situations, from error to silently (except under debug)
303 ignoring. This covers use with PRDR, frozen messages, queue-only and
306 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
308 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
309 metadata, resulting in a crash in free().
311 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
312 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
313 Broken also in d185889f4, with init system revamp.
315 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
316 Previously we abruptly closed the connection after reading a malware-
317 found indication; now we go on to read the "scan ok" response line,
320 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
321 ACL. Previously, a crash would result.
323 JH/19 Speed up macro lookups during configuration file read, by skipping non-
324 macro text after a replacement (previously it was only once per line) and
325 by skipping builtin macros when searching for an uppercase lead character.
327 JH/20 DANE support moved from Experimental to mainline. The Makefile control
328 for the build is renamed.
330 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
331 was allocated for every new TLS startup, meaning one per message. Fix
332 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
334 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
335 reported the original. Fix to report (as far as possible) the ACL
336 result replacing the original.
338 JH/23 Fix memory leak during multi-message connections using STARTTLS under
339 OpenSSL. Certificate information is loaded for every new TLS startup,
340 and the resources needed to be freed.
342 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
344 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
347 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
348 DATA response info to the (existing) per-recipient response info for
349 the "C=" log element. It can have useful tracking info from the
350 destination system. Patch from Simon Arlott.
352 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
353 length value. Previously this would segfault.
355 HS/02 Support Avast multiline protoocol, this allows passing flags to
356 newer versions of the scanner.
358 JH/28 Ensure that variables possibly set during message acceptance are marked
359 dead before release of memory in the daemon loop. This stops complaints
360 about them when the debug_store option is enabled. Discovered specifically
361 for sender_rate_period, but applies to a whole set of variables.
362 Do the same for the queue-runner and queue-list loops, for variables set
363 from spool message files. Do the same for the SMTP per-message loop, for
364 certain variables indirectly set in ACL operations.
366 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
367 as a multi-recipient message from a mailinglist manager). The coding had
368 an arbitrary cutoff number of characters while checking for more input;
369 enforced by writing a NUL into the buffer. This corrupted long / fast
370 input. The problem was exposed more widely when more pipelineing of SMTP
371 responses was introduced, and one Exim system was feeding another.
372 The symptom is log complaints of SMTP syntax error (NUL chars) on the
373 receiving system, and refused recipients seen by the sending system
374 (propating to people being dropped from mailing lists).
375 Discovered and pinpointed by David Carter.
377 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
378 replaced by the ${authresults } expansion.
380 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
382 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
383 allows proper process termination in container environments.
385 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
386 Previously the "final dot" had a newline after it; ensure it is CR,LF.
388 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
389 and "err_perm", deprecated since 4.83 when the RFC-defined words
390 "temperror" and "permerror" were introduced.
392 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
393 transport-filters or DKIM-signing. The restriction was lost in the
394 consolidation of verify-callout and delivery SMTP handling.
395 Extend the restriction to also cover ARC-signing.
397 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
398 in defer=pass mode supply a 450 to the initiator. Previously the message
401 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
402 tls_require_ciphers is used as before.
404 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
405 "pass_unscanned". Only tmpfails from the scanner are written to
406 the paniclog, as they may require admin intervention (permission
407 denied, license issues). Other scanner errors (like decompression
408 bombs) do not cause a paniclog entry.
410 JH/36 Fix reinitialisation of DKIM logging variable between messages.
411 Previously it was possible to log spurious information in receive log
414 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
415 triggered odd behaviour from Outlook Express clients.
417 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
420 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
421 since the IETF WG has not yet settled on that versus the original
422 "bare" representation.
424 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
425 Previously the millisecond value corrupted the output.
426 Fix also for syslog_pid=no and log_selector +pid, for which the pid
427 corrupted the output.
433 JH/01 Rework error string handling in TLS interface so that the caller in
434 more cases is responsible for logging. This permits library-sourced
435 string to be attached to addresses during delivery, and collapses
436 pairs of long lines into single ones.
438 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
439 during configuration. Wildcards are allowed and expanded.
441 JH/02 Rework error string handling in DKIM to pass more info back to callers.
442 This permits better logging.
444 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
445 do not close it down and have the child transport start it up again on
446 the passed-on TCP connection. Instead, proxy the child (and any
447 subsequent ones) for TLS via a unix-domain socket channel. Logging is
448 affected: the continued delivery log lines do not have any DNSSEC, TLS
449 Certificate or OCSP information. TLS cipher information is still logged.
451 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
452 identical IP addresses on different listening ports. Will also affect
455 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
456 add noisy ifdef guards to special-case this sillyness.
457 Patch from Bernd Kuhls.
459 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
460 than 255 are no longer allowed.
462 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
463 Disable the session-cache too, which might reduce our load. Since we
464 currrectly use a new context for every connection, both as server and
465 client, there is no benefit for these.
466 GnuTLS appears to not support tickets server-side by default (we don't
467 call gnutls_session_ticket_enable_server()) but client side is enabled
468 by default on recent versions (3.1.3 +) unless the PFS priority string
471 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
472 <https://reproducible-builds.org/specs/source-date-epoch/>.
474 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
475 the check for any unsuccessful recipients did not notice the limit, and
476 erroneously found still-pending ones.
478 JH/08 Pipeline CHUNKING command and data together, on kernels that support
479 MSG_MORE. Only in-clear (not on TLS connections).
481 JH/09 Avoid using a temporary file during transport using dkim. Unless a
482 transport-filter is involved we can buffer the headers in memory for
483 creating the signature, and read the spool data file once for the
484 signature and again for transmission.
486 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
487 4.77 as the kernel support then wasn't solid, having issues in 64bit
488 mode. Now, it's been long enough. Add support for FreeBSD also.
490 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
491 case where the routing stage had gathered several addresses to send to
492 a host before calling the transport for the first, we previously failed
493 to close down TLS in the old transport process before passing the TCP
494 connection to the new process. The new one sent a STARTTLS command
495 which naturally failed, giving a failed delivery and bloating the retry
496 database. Investigation and fix prototype from Wolfgang Breyha.
498 JH/12 Fix check on SMTP command input synchronisation. Previously there were
499 false-negatives in the check that the sender had not preempted a response
500 or prompt from Exim (running as a server), due to that code's lack of
501 awareness of the SMTP input buffering.
503 PP/04 Add commandline_checks_require_admin option.
504 Exim drops privileges sanely, various checks such as -be aren't a
505 security problem, as long as you trust local users with access to their
506 own account. When invoked by services which pass untrusted data to
507 Exim, this might be an issue. Set this option in main configuration
508 AND make fixes to the calling application, such as using `--` to stop
511 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
512 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
513 responses to those, into a single TLS record each way (this usually means
514 a single packet). As a side issue, smtp_enforce_sync now works on TLS
517 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
518 affects you only if you're dancing at the edge of the param size limits.
519 If you are, and this message makes sense to you, then: raise the
520 configured limit or use OpenSSL 1.1. Nothing we can do for older
523 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
524 cmdline element to get the documented default one. Previously it was
527 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
528 are unsuable for non-message-specific reasons - eg. connection timeout,
529 banner-time rejection.
531 JH/16 Fix logging of delivery remote port, when specified by router, under
534 PP/06 Repair manualroute's ability to take options in any order, even if one
535 is the name of a transport.
538 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
540 JH/17 Change the list-building routines interface to use the expanding-string
541 triplet model, for better allocation and copying behaviour.
543 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
544 Previously it was constructed the first time a possibly-matching string
545 was met in the configuration file input during startup; now it is done
548 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
549 compatible one, to avoid the (poorly documented) possibility of a config
550 file in the working directory redirecting the DB files, possibly correpting
551 some existing file. CVE-2017-10140 assigned for BDB.
553 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
554 cache-hot. Previously, although the result was properly cached, the
555 initial verify call returned a defer.
557 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
558 the main verify for receipient in uncached-mode.
560 JH/22 Retire historical build files to an "unsupported" subdir. These are
561 defined as "ones for which we have no current evidence of testing".
563 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
564 if present. Previously it was ignored.
566 JH/24 Start using specified-initialisers in C structure init coding. This is
567 a C99 feature (it's 2017, so now considered safe).
569 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
570 if was a fixed-sized field and bitmask ops via macros; it is now more
573 PP/07 GitHub PR 56: Apply MariaDB build fix.
574 Patch provided by Jaroslav Škarvada.
576 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
577 during Coverity cleanups [4.87 JH/47]
578 Diagnosis and fix provided by Michael Fischer v. Mollard.
580 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
581 the right size to place the terminating semicolon on its own folded
582 line, the header hash was calculated to an incorrect value thanks to
583 the (relaxed) space the fold became.
585 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
586 and confused the parent.
588 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
589 which could crash as a result. This could lead to undeliverable messages.
591 JH/28 Logging: "next input sent too soon" now shows where input was truncated
594 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
595 matters on fast-turnover and PID-randomising systems, which were getting
596 out-of-order delivery.
598 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
599 a possibly-overlapping copy. The symptom was that "Remote host closed
600 connection in response to HELO" was logged instead of the actual 4xx
603 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
604 Previously only that bufferd was discarded, resulting in SYMTP command
607 JH/32 DKIM: when a message has multiple signatures matching an identity given
608 in dkim_verify_signers, run the dkim acl once for each. Previously only
609 one run was done. Bug 2189.
611 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
612 "panic the current process" to "deliberately defer". The panic log is
613 still written with the problem list name; the mail and reject logs now
614 get a temp-reject line for the message that was being handled, saying
615 something like "domains check lookup or other defer". The SMTP 451
616 message is still "Temporary local problem".
618 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
619 A crafted sequence of BDAT commands could result in in-use memory beeing
620 freed. CVE-2017-16943.
622 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
623 from SMTP input. Previously it was always done; now only done for DATA
624 and not BDAT commands. CVE-2017-16944.
626 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
627 to the message (such as an overlong header line). Previously this was
628 not done and we did not exit BDAT mode. Followon from the previous item
629 though a different problem.
635 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
636 than -2003 did; needs libidn2 in addition to libidn.
638 JH/02 The path option on a pipe transport is now expanded before use.
640 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
641 Patch provided by "Björn", documentation fix added too.
643 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
644 missing a wire-to-host endian conversion.
646 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
647 close after a BDAT command line could be taken as a following command,
648 giving a synch failure. Fix by only checking for synch immediately
649 before acknowledging the chunk.
651 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
652 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
653 macro. Patches provided by Josh Soref.
655 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
656 Previously we did not; the RFC seems ambiguous and VRFY is not listed
657 by IANA as a service extension. However, John Klensin suggests that we
660 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
661 the dkim code may be unix-mode line endings rather than smtp wire-format
662 CRLF, so prepend a CR to any bare LF.
664 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
665 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
667 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
668 the first were themselves being wrongly included in the feed into dkim
669 processing; with most chunk sizes in use this resulted in an incorrect
670 body hash calculated value.
672 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
673 DKIM signature block, for verification. Although advised against by
674 standards it is specifically not ruled illegal.
676 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
678 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
679 missing a body hash (the bh= tag).
681 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
682 It seems that HAProxy sends the Proxy Protocol information in clear and
683 only then does a TLS startup, so do the same.
685 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
686 TCP connections (such as for Spamd) unless the daemon successfully set
687 Fast Open mode on its listening sockets. This fixes breakage seen on
688 too-old kernels or those not configured for Fast Open, at the cost of
689 requiring both directions being enabled for TFO, and TFO never being used
690 by non-daemon-related Exim processes.
692 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
693 endings, at least on the first header line. Try to canonify any that get
694 past that check, despite the cost.
696 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
697 now limited to an arbitrary five deep, while parsing addresses with the
698 strip_excess_angle_brackets option enabled.
700 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
701 instead leave the unprompted TLS handshake in socket buffer for the
702 TLS library to consume.
704 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
706 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
708 JH/16 Drop variables when they go out of scope. Memory management drops a whole
709 region in one operation, for speed, and this leaves assigned pointers
710 dangling. Add checks run only under the testsuite which checks all
711 variables at a store-reset and panics on a dangling pointer; add code
712 explicitly nulling out all the variables discovered. Fixes one known
713 bug: a transport crash, where a dangling pointer for $sending_ip_address
714 originally assigned in a verify callout, is re-used.
716 PP/06 Drop '.' from @INC in various Perl scripts.
718 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
720 PP/08 Reduce a number of compilation warnings under clang; building with
721 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
722 should be warning-free.
724 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
726 HS/01 Fix portability problems introduced by PP/08 for platforms where
727 realloc(NULL) is not equivalent to malloc() [SunOS et al].
729 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
730 chunk. This allows us to accept broken chunked messages. We need a more
731 general solution here.
733 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
734 already-broken messages in the queue.
736 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
738 JH/19 Fix reference counting bug in routing-generated-address tracking.
744 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
745 supports it and a size is available (ie. the sending peer gave us one).
747 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
748 years of being deprecated). The replacements are the ACLs
749 acl_smtp_mime and acl_not_smtp_mime.
751 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
752 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
753 or even an in-clear connection were permitted. Now, if the host lookup
754 was dnssec and dane was requested then the host is only used if the
755 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
756 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
757 if one fails this test.
758 This means that a poorly-configured remote DNS will make it incommunicado;
759 but it protects against a DNS-interception attack on it.
761 JH/04 Bug 1810: make continued-use of an open smtp transport connection
762 non-noisy when a race steals the message being considered.
764 JH/05 If main configuration option tls_certificate is unset, generate a
765 self-signed certificate for inbound TLS connections.
767 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
768 in rewrites and routers.
770 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
771 and logged a warning sing 4.83; now they are a configuration file error.
773 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
774 (lacking @domain). Apply the same qualification processing as RCPT.
776 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
778 JH/10 Support ${sha256:} applied to a string (as well as the previous
781 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
782 a cutthrough deliver is pending, as we always want to make a connection.
783 This also avoids re-routing the message when later placing the cutthrough
784 connection after a verify cache hit.
785 Do not update it with the verify result either.
787 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
788 when routing results in more than one destination address.
790 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
791 signing (which inhibits the cutthrough capability). Previously only
792 the presence of an option was tested; now an expansion evaluating as
793 empty is permissible (obviously it should depend only on data available
794 when the cutthrough connection is made).
796 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
797 the relevant preceding SMTP command did not note the pipelining mode.
799 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
800 Previously they were not counted.
802 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
803 as one having no matching records. Previously we deferred the message
804 that needed the lookup.
806 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
807 distinguished as "(=".
809 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
810 for missing MX records. Previously it only worked for missing A records.
812 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
814 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
815 after the data-go-ahead and data-ack. Patch from Jason Betts.
817 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
818 even for a "none" policy. Patch from Tony Meyer.
820 JH/22 Fix continued use of a connection for further deliveries. If a port was
821 specified by a router, it must also match for the delivery to be
824 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
825 When one of the recipients of a message was unsuitable for the connection
826 (has no matching addresses), we lost track of needing to mark it
827 deferred. As a result mail would be lost.
829 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
831 JH/25 Decoding ACL controls is now done using a binary search; the source code
832 takes up less space and should be simpler to maintain. Merge the ACL
833 condition decode tables also, with similar effect.
835 JH/26 Fix problem with one_time used on a redirect router which returned the
836 parent address unchanged. A retry would see the parent address marked as
837 delivered, so not attempt the (identical) child. As a result mail would
840 JH/27 Fix a possible security hole, wherein a process operating with the Exim
841 UID can gain a root shell. Credit to http://www.halfdog.net/ for
842 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
845 JH/28 Enable {spool,log} filesystem space and inode checks as default.
846 Main config options check_{log,spool}_{inodes,space} are now
847 100 inodes, 10MB unless set otherwise in the configuration.
849 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
850 Previously it only applied to the main-section connection policy
853 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
855 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
856 by me. Added RFC7919 DH primes as an alternative.
858 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
859 are not in the system include path.
861 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
862 GnuTLS, when a session startup failed (eg because the client disconnected)
863 Exim did stdio operations after fclose. This was exposed by a recent
864 change which nulled out the file handle after the fclose.
866 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
867 signed directly by the cert-signing cert, rather than an intermediate
868 OCSP-signing cert. This is the model used by LetsEncrypt.
870 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
872 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
873 an incoming connection.
875 HS/02 Bug 1802: Do not half-close the connection after sending a request
878 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
879 fallback to "prime256v1".
881 JH/34 SECURITY: Use proper copy of DATA command in error message.
882 Could leak key material. Remotely exploitable. CVE-2016-9963.
888 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
889 and 3.4.4 - once the server is enabled to respond to an OCSP request
890 it does even when not requested, resulting in a stapling non-aware
891 client dropping the TLS connection.
893 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
894 support variable-length bit vectors. No functional change.
896 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
897 The I= interface field on outgoing lines is now after the H= remote
898 host field, same as incoming lines. There is a separate
899 outgoing_interface log selector which allows you to disable the
902 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
903 If not running log_selector +smtp_connection the mainlog would be held
904 open indefinitely after a "too many connections" event, including to a
905 deleted file after a log rotate. Leave the per net connection logging
906 leaving it open for efficiency as that will be quickly detected by the
907 check on the next write.
909 HS/01 Bug 1671: Fix post transport crash.
910 Processing the wait-<transport> messages could crash the delivery
911 process if the message IDs didn't exist for some reason. When
912 using 'split_spool_directory=yes' the construction of the spool
913 file name failed already, exposing the same netto behaviour.
915 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
916 mime_regex ACL conditions.
918 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
919 to DSN fail messages (bounces): remote IP, remote greeting, remote response
920 to HELO, local diagnostic string.
922 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
923 log line to debug. Even when configured with a tls authenticator many
924 client connections are expected to not authenticate in this way, so
925 an authenticate fail is not an error.
927 HS/02 Add the Exim version string to the process info. This way exiwhat
928 gives some more detail about the running daemon.
930 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
931 matter for fast-change records such as DNSBLs.
933 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
934 retry record, even if constant. There may be multiple transports with
935 different interface settings and the retry behaviour needs to be kept
938 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
940 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
942 JH/10 Bug 840: fix log_defer_output option of pipe transport
944 JH/11 Bug 830: use same host for all RCPTS of a message, even under
945 hosts_randomize. This matters a lot when combined with mua_wrapper.
947 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
948 ${quote_pgsql:<string>} operator.
950 JH/13 Bug 1708: avoid misaligned access in cached lookup.
952 JH/14 Change header file name for freeradius-client. Relevant if compiling
953 with Radius support; from the Gentoo tree and checked under Fedora.
955 JH/15 Bug 1712: Introduce $prdr_requested flag variable
957 JH/16 Bug 1714: Permit an empty string as expansion result for transport
958 option transport_filter, meaning no filtering.
960 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
962 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
963 defaults to "*" (all hosts). The variable is now available when not built
964 with TLS, default unset, mainly to enable keeping the testsuite sane.
965 If a server certificate is not supplied (via tls_certificate) an error is
966 logged, and clients will find TLS connections fail on startup. Presumably
967 they will retry in-clear.
968 Packagers of Exim are strongly encouraged to create a server certificate
969 at installation time.
971 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
972 with the $config_file variable.
974 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
975 in transport context, after the attempt, and per-recipient. The latter type
976 is per host attempted. The event data is the error message, and the errno
977 information encodes the lookup type (A vs. MX) used for the (first) host,
978 and the trailing two digits of the smtp 4xx response.
980 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
981 to write to mainlog (or rejectlog, paniclog) in the window between file
982 creation and permissions/ownership being changed. Particularly affects
983 installations where exicyclog is run as root, rather than exim user;
984 result is that the running daemon panics and dies.
986 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
988 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
989 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
990 "pri" and "weight". Note that the previous implicit priority given by the
991 list order is no longer honoured.
993 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
996 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
997 by defining SUPPORT_SOCKS.
999 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1000 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1001 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1002 variables are renamed to proxy_{local,external}_{address,port}.
1004 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1005 for a build by defining SUPPORT_I18N
1007 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1008 of the query string, and make ${quote_redis:} do that quoting.
1010 JH/29 Move Events support from Experimental to mainline, enabled by default
1011 and removable for a build by defining DISABLE_EVENT.
1013 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1015 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1016 cached by the daemon.
1018 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1019 by defining LOOKUP_REDIS. The libhiredis library is required.
1021 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1022 keys are given for lookup.
1024 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1025 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1026 only supported when built with TLS support. The PolarSSL SHA routines
1027 are still used when the TLS library is too old for convenient support.
1029 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1030 openssl_options), for security. OpenSSL forces this from version 1.1.0
1031 server-side so match that on older versions.
1033 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1034 allocation for $value could be released as the expansion processing
1035 concluded, but leaving the global pointer active for it.
1037 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1038 and to use the domains and local_parts ACL conditions.
1040 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1041 incorrectly not doubled on cutthrough transmission, hence seen as a
1042 body-termination at the receiving system - resulting in truncated mails.
1043 Commonly the sender saw a TCP-level error, and retransmitted the message
1044 via the normal store-and-forward channel. This could result in duplicates
1045 received - but deduplicating mailstores were liable to retain only the
1046 initial truncated version.
1048 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1050 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1052 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1053 we're in there, support oversigning also; bug 1309.
1055 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1057 HS/04 Add support for keep_environment and add_environment options.
1059 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1060 either intentional arithmetic overflow during PRNG, or testing config-
1063 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1064 delivery resulted in actual delivery. Cancel cutthrough before DATA
1067 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1068 rejects a recipient: pass the reject to the originator.
1070 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1071 Many were false-positives and ignorable, but it's worth fixing the
1074 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1075 for the new environment-manipulation done at startup. Move the routines
1076 from being local to tls.c to being global via the os.c file.
1078 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1079 an extract embedded as result-arg for a map, the first arg for extract
1080 is unavailable so we cannot tell if this is a numbered or keyed
1081 extraction. Accept either.
1087 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1090 JH/02 The smtp transport option "multi_domain" is now expanded.
1092 JH/03 The smtp transport now requests PRDR by default, if the server offers
1095 JH/04 Certificate name checking on server certificates, when exim is a client,
1096 is now done by default. The transport option tls_verify_cert_hostnames
1097 can be used to disable this per-host. The build option
1098 EXPERIMENTAL_CERTNAMES is withdrawn.
1100 JH/05 The value of the tls_verify_certificates smtp transport and main options
1101 default to the word "system" to access the system default CA bundle.
1102 For GnuTLS, only version 3.0.20 or later.
1104 JH/06 Verification of the server certificate for a TLS connection is now tried
1105 (but not required) by default. The verification status is now logged by
1106 default, for both outbound TLS and client-certificate supplying inbound
1109 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1112 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1113 Status Notification (bounce) messages are now MIME format per RFC 3464.
1114 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1115 under the control of the dsn_advertise_hosts option, and routers may
1116 have a dsn_lasthop option.
1118 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1119 default, modifiable by a malware= option. The list separator for
1120 the options can now be changed in the usual way. Bug 68.
1122 JH/10 The smtp_receive_timeout main option is now expanded before use.
1124 JH/11 The incoming_interface log option now also enables logging of the
1125 local interface on delivery outgoing connections.
1127 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1128 if the interface and destination host and port all match.
1130 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1133 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1134 Patch from Andrew Lewis.
1136 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1137 now supports optional time-restrictions, weighting, and priority
1138 modifiers per server. Patch originally by <rommer@active.by>.
1140 JH/16 The spamd_address main option now supports a mixed list of local
1141 and remote servers. Remote servers can be IPv6 addresses, and
1142 specify a port-range.
1144 JH/17 Bug 68: The spamd_address main option now supports an optional
1145 timeout value per server.
1147 JH/18 Bug 1581: Router and transport options headers_add/remove can
1148 now have the list separator specified.
1150 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1153 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1156 JH/21 Support for the A6 type of dns record is withdrawn.
1158 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1159 rather than the verbs used.
1161 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1162 from 255 to 1024 chars.
1164 JH/24 Verification callouts now attempt to use TLS by default.
1166 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1167 are generic router options now. The defaults didn't change.
1169 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1170 Original patch from Alexander Shikoff, worked over by JH.
1172 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1173 files if ZCAT_COMMAND is not executable.
1175 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1177 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1179 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1180 Normally benign, it bites when the pair was led to by a CNAME;
1181 modern usage is to not canonicalize the domain to a CNAME target
1182 (and we were inconsistent anyway for A-only vs AAAA+A).
1184 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1186 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1187 when evaluating $sender_host_dnssec.
1189 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1190 $sender_helo_dnssec variable.
1192 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1194 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1196 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1198 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1199 documented as working, but never had. Support all but $spam_report.
1201 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1202 added for tls authenticator.
1204 HS/03 Add perl_taintmode main config option
1210 TL/01 When running the test suite, the README says that variables such as
1211 no_msglog_check are global and can be placed anywhere in a specific
1212 test's script, however it was observed that placement needed to be near
1213 the beginning for it to behave that way. Changed the runtest perl
1214 script to read through the entire script once to detect and set these
1215 variables, reset to the beginning of the script, and then run through
1216 the script parsing/test process like normal.
1218 TL/02 The BSD's have an arc4random API. One of the functions to induce
1219 adding randomness was arc4random_stir(), but it has been removed in
1220 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1221 function when detected.
1223 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1224 cause callback expansion.
1226 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1227 syntax errors in an expansion can be treated as a string instead of
1228 logging or causing an error, due to the internal use of bool_lax
1229 instead of bool when processing it.
1231 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1232 server certificates when making smtp deliveries.
1234 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1236 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1238 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1240 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1241 Merged patch from Sebastian Wiedenroth.
1243 JH/05 Fix results-pipe from transport process. Several recipients, combined
1244 with certificate use, exposed issues where response data items split
1245 over buffer boundaries were not parsed properly. This eventually
1246 resulted in duplicates being sent. This issue only became common enough
1247 to notice due to the introduction of connection certificate information,
1248 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1250 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1251 size buffer was used, resulting in syntax errors when an expansion
1254 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1255 version 3.3.6 or later.
1257 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1258 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1259 both become "event_action", the variables become $event_name, $event_data
1260 and $event_defer_errno. There is a new variable $verify_mode, usable in
1261 routers, transports and related events. The tls:cert event is now also
1262 raised for inbound connections, if the main configuration event_action
1265 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1266 early OCSP support, but no stapling (appears to be less than 1.0.0).
1268 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1269 server certificate names available under the smtp transport option
1270 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1273 JH/10 Time-related extraction expansions from certificates now use the main
1274 option "timezone" setting for output formatting, and are consistent
1275 between OpenSSL and GnuTLS compilations. Bug 1541.
1277 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1278 encoded parameter in the incoming message. Bug 1558.
1280 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1281 include certificate info, eximon was claiming there were spoolfile
1284 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1286 JH/14 Log delivery-related information more consistently, using the sequence
1287 "H=<name> [<ip>]" wherever possible.
1289 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1290 are problematic for Debian distribution, omit them from the release
1293 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1295 JH/16 Fix string representation of time values on 64bit time_t architectures.
1298 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1299 output list separator was used.
1304 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1305 checkers that were complaining about end of non-void function with no
1308 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1309 This was a regression introduced in 4.83 by another bugfix.
1311 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1313 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1314 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1320 TF/01 Correctly close the server side of TLS when forking for delivery.
1322 When a message was received over SMTP with TLS, Exim failed to clear up
1323 the incoming connection properly after forking off the child process to
1324 deliver the message. In some situations the subsequent outgoing
1325 delivery connection happened to have the same fd number as the incoming
1326 connection previously had. Exim would try to use TLS and fail, logging
1327 a "Bad file descriptor" error.
1329 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1330 utilities have not been installed.
1332 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1333 temporary space as the ACL may create new global variables.
1335 TL/01 LDAP support uses per connection or global context settings, depending
1336 upon the detected version of the libraries at build time.
1338 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1339 to extract and use the src ip:port in logging and expansions as if it
1340 were a direct connection from the outside internet. PPv2 support was
1341 updated based on HAProxy spec change in May 2014.
1343 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1345 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1346 Properly escape header and check for NULL return.
1348 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1351 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1353 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1354 characters in header names, implemented as a verify condition.
1355 Contributed by Michael Fischer v. Mollard.
1357 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1358 results permerror and temperror. Previous values are deprecated but
1359 still accepted. In a future release, err_perm and err_temp will be
1360 completely removed, which will be a backward incompatibility if the
1361 ACL tests for either of these two old results. Patch contributed by
1362 user bes-internal on the mailing list.
1364 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1366 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1367 selectors, in both main and reject logs.
1369 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1372 JH/07 Add malware type "sock" for talking to simple daemon.
1374 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1376 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1377 routers/transports under cutthrough routing.
1379 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1380 numbers. Touch up "bool" conditional to keep the same definition.
1382 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1384 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1386 JH/12 Expand items in router/transport headers_add or headers_remove lists
1387 individually rather than the list as a whole. Bug 1452.
1389 Required for reasonable handling of multiple headers_ options when
1390 they may be empty; requires that headers_remove items with embedded
1391 colons must have them doubled (or the list-separator changed).
1393 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1394 view the policy declared in the DMARC record. Currently, $dmarc_status
1395 is a combined value of both the record presence and the result of the
1398 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1400 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1401 dnslookup router and the smtp transport (applying to the forward
1404 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1405 of ldap servers used for a specific lookup. Patch provided by Heiko
1408 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1409 New variable $lookup_dnssec_authenticated for observability.
1411 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1412 Patch submitted by Lars Timman.
1414 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1416 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1417 Requires trusted mode and valid format message id, aborts otherwise.
1418 Patch contributed by Heiko Schlichting.
1420 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1421 certextract with support for various fields. Bug 1358.
1423 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1424 is requested by default, modifiable by smtp transport option
1427 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1428 operate on certificate variables to give certificate fingerprints
1429 Also new ${sha256:cert_variable}.
1431 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1433 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1434 Christian Aistleitner.
1436 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1438 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1439 file. Patch from Wolfgang Breyha.
1441 JH/25 Expand the coverage of the delivery $host and $host_address to
1442 client authenticators run in verify callout. Bug 1476.
1444 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1445 align with daemon_smtp_ports. Bug 72.
1447 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1448 support and error reporting did not work properly.
1450 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1451 and is readable. Patch from Andrew Colin Kissa.
1453 TL/14 Enhance documentation of ${run expansion and how it parses the
1454 commandline after expansion, particularly in the case when an
1455 unquoted variable expansion results in an empty value.
1457 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1459 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1460 by renaming away from it. Addresses GH issue 15, reported by
1463 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1465 TL/15 SECURITY: prevent double expansion in math comparison functions
1466 (can expand unsanitized data). Not remotely exploitable.
1473 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1475 PP/02 Make -n do something, by making it not do something.
1476 When combined with -bP, the name of an option is not output.
1478 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1481 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1482 $sender_host_name and config options to manage this, and basic check
1485 PP/05 DSCP support for outbound connections and control modifier for inbound.
1487 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1488 (Only plugin which currently uses this is kerberos4, which nobody should
1489 be using, but we should make it available and other future plugins might
1490 conceivably use it, even though it would break NAT; stuff *should* be
1491 using channel bindings instead).
1493 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1494 name; added for Sendmail compatibility; requires admin caller.
1495 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1496 just ignore it); requires trusted caller.
1497 Also parse but ignore: -Ac -Am -X<logfile>
1500 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1502 TL/02 Add +smtp_confirmation as a default logging option.
1504 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1505 Patch by Magnus Holmgren from 2007-02-20.
1507 TL/04 Bugzilla 1281 - Spec typo.
1508 Bugzilla 1283 - Spec typo.
1509 Bugzilla 1290 - Spec grammar fixes.
1511 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1513 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1515 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1516 mailing list by by Dmitry Isaikin.
1518 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1520 JH/02 Support "G" suffix to numbers in ${if comparisons.
1522 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1524 NM/01 Bugzilla 1197 - Spec typo
1525 Bugzilla 1196 - Spec examples corrections
1527 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1529 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1530 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1533 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1534 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1536 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1537 "acl {{name}{arg}...}", and optional args on acl condition
1540 JH/05 Permit multiple router/transport headers_add/remove lines.
1542 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1544 JH/07 Avoid using a waiting database for a single-message-only transport.
1545 Performance patch from Paul Fisher. Bugzilla 1262.
1547 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1550 JH/09 Add $headers_added variable, with content from use of ACL modifier
1551 add_header (but not yet added to the message). Bugzilla 199.
1553 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1554 Pulled from Bugzilla 817 by Wolfgang Breyha.
1556 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1558 (nb: this is the same fix as in Exim 4.80.1)
1560 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1563 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1566 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1568 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1569 advertise SMTP AUTH mechanism to us, instead of a generic
1570 protocol violation error. Also, make Exim more robust to bad
1571 data from the Dovecot auth socket.
1573 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1575 When a queue runner is handling a message, Exim first routes the
1576 recipient addresses, during which it prunes them based on the retry
1577 hints database. After that it attempts to deliver the message to
1578 any remaining recipients. It then updates the hints database using
1581 So if a recipient address works intermittently, it can get repeatedly
1582 deferred at routing time. The retry hints record remains fresh so the
1583 address never reaches the final cutoff time.
1585 This is a fairly common occurrence when a user is bumping up against
1586 their storage quota. Exim had some logic in its local delivery code
1587 to deal with this. However it did not apply to per-recipient defers
1588 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1590 This change adds a proper retry rule check during routing so that the
1591 final cutoff time is checked against the message's age. We only do
1592 this check if there is an address retry record and there is not a
1593 domain retry record; this implies that previous attempts to handle
1594 the address had the retry_use_local_parts option turned on. We use
1595 this as an approximation for the destination being like a local
1596 delivery, as in LMTP.
1598 I suspect this new check makes the old local delivery cutoff check
1599 redundant, but I have not verified this so I left the code in place.
1601 TF/02 Correct gecos expansion when From: is a prefix of the username.
1603 Test 0254 submits a message to Exim with the header
1607 When I ran the test suite under the user fanf2, Exim expanded
1608 the header to contain my full name, whereas it should have added
1609 a Resent-Sender: header. It erroneously treats any prefix of the
1610 username as equal to the username.
1612 This change corrects that bug.
1614 GF/01 DCC debug and logging tidyup
1615 Error conditions log to paniclog rather than rejectlog.
1616 Debug lines prefixed by "DCC: " to remove any ambiguity.
1618 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1620 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1621 Bug spotted by Jeremy Harris; was flawed since initial commit.
1622 Would have resulted in OCSP responses post-SNI triggering an Exim
1623 NULL dereference and crash.
1625 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1627 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1628 Bug detection, analysis and fix by Samuel Thibault.
1629 Bugzilla 1331, Debian bug #698092.
1631 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1633 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1634 Server implementation by Todd Lyons, client by JH.
1635 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1636 config variable "prdr_enable" controls whether the server
1637 advertises the facility. If the client requests PRDR a new
1638 acl_data_smtp_prdr ACL is called once for each recipient, after
1639 the body content is received and before the acl_smtp_data ACL.
1640 The client is controlled by both of: a hosts_try_prdr option
1641 on the smtp transport, and the server advertisement.
1642 Default client logging of deliveries and rejections involving
1643 PRDR are flagged with the string "PRDR".
1645 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1646 fclose(). Diagnosis by Todd Lyons.
1648 PP/17 Update configure.default to handle IPv6 localhost better.
1649 Patch by Alain Williams (plus minor tweaks).
1652 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1653 This is now consistent with GnuTLS, and is now documented: the
1654 previous undocumented portable approach to treating the option as
1655 unset was to force an expansion failure. That still works, and
1656 an empty string is now equivalent.
1658 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1659 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1660 not performing validation itself.
1662 PP/20 Added force_command boolean option to pipe transport.
1663 Patch from Nick Koston, of cPanel Inc.
1665 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1668 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1670 PP/21 Fix eximon continuous updating with timestamped log-files.
1671 Broken in a format-string cleanup in 4.80, missed when I repaired the
1672 other false fix of the same issue.
1673 Report and fix from Heiko Schlichting.
1676 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1677 Report from Prashanth Katuri.
1679 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1680 It's SecureTransport, so affects any MacOS clients which use the
1681 system-integrated TLS libraries, including email clients.
1683 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1684 using a MIME ACL for non-SMTP local injection.
1685 Report and assistance in diagnosis by Warren Baker.
1687 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1689 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1691 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1692 last id that failed so it may be referenced in subsequent ACL's.
1694 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1697 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1698 ldap library initialization, allowing self-signed CA's to be
1699 used. Also properly sets require_cert option later in code by
1700 using NULL (global ldap config) instead of ldap handle (per
1701 session). Bug diagnosis and testing by alxgomz.
1703 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1704 the src/util/ subdirectory.
1706 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1707 renamed to Transport Post Delivery Action by Jeremy Harris, as
1710 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1711 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1712 redis_servers = needs to be configured which will be used by the redis
1713 lookup. Patch from Warren Baker, of The Packet Hub.
1715 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1717 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1718 hostname or reverse DNS when processing a host list. Used suggestions
1719 from multiple comments on this bug.
1721 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1723 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1724 Missed a few lines, added it to make the runtest require no keyboard
1727 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1728 contains upper case chars. Make router use caseful_local_part.
1730 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1731 support when GnuTLS has been built with p11-kit.
1737 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1739 This, or similar/improved, will also be change PP/11 of 4.82.
1745 PP/01 Handle short writes when writing local log-files.
1746 In practice, only affects FreeBSD (8 onwards).
1747 Bugzilla 1053, with thanks to Dmitry Isaikin.
1749 NM/01 Bugzilla 949 - Documentation tweak
1751 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1754 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1756 PP/02 Implemented gsasl authenticator.
1758 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1760 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1761 `pkg-config foo` for cflags/libs.
1763 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1764 with rest of GSASL and with heimdal_gssapi.
1766 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1767 `pkg-config foo` for cflags/libs for the TLS implementation.
1769 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1770 properties get this fed in as external SSF. A number of robustness
1771 and debugging improvements to the cyrus_sasl authenticator.
1773 PP/08 cyrus_sasl server now expands the server_realm option.
1775 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1776 Patch by Jeremy Harris.
1778 PP/10 Added dbmjz lookup type.
1780 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1782 PP/12 MAIL args handles TAB as well as SP, for better interop with
1783 non-compliant senders.
1784 Analysis and variant patch by Todd Lyons.
1786 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1787 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1788 Patch from Dirk Mueller <dmueller@suse.com>
1790 PP/13 tls_peerdn now print-escaped for spool files.
1791 Observed some $tls_peerdn in wild which contained \n, which resulted
1792 in spool file corruption.
1794 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1795 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1796 or write after TLS renegotiation, which otherwise led to messages
1799 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1800 as a tracking header (ie: a signed header comes before the signature).
1801 Patch from Wolfgang Breyha.
1803 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1804 comma-sep list; embedded commas doubled.
1806 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1808 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1810 Report and patch from Dmitry Banschikov.
1812 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1813 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1814 failures. We appear to now support TLS1.1+ with Exim.
1816 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1817 lets Exim select keys and certificates based upon TLS SNI from client.
1818 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1819 before an outbound SMTP session. New log_selector, +tls_sni.
1821 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1822 NULL dereference. Report and patch from Alun Jones.
1824 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1825 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1826 Not seeing resolver debug output on NetBSD, but suspect this is a
1827 resolver implementation change.
1829 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1830 Left warnings. Added "eximon gdb" invocation mode.
1832 PP/21 Defaulting "accept_8bitmime" to true, not false.
1834 PP/22 Added -bw for inetd wait mode support.
1836 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1837 locate the relevant includes and libraries. Made this the default.
1839 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1840 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1842 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1843 This may cause build issues on older platforms.
1845 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1846 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1847 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1848 Added SNI support via GnuTLS too.
1849 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1851 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1853 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1854 Applied second patch from Janne, implementing suggestion to default
1855 multiple-strings-in-record handling to match SPF spec.
1857 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1859 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1860 read-only, out of scope).
1861 Patch from Wolfgang Breyha, report from Stuart Northfield.
1863 PP/29 Fix three issues highlighted by clang analyser static analysis.
1864 Only crash-plausible issue would require the Cambridge-specific
1865 iplookup router and a misconfiguration.
1866 Report from Marcin Mirosław.
1868 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1870 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1871 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1872 As part of this, removing so much warning spew let me fix some minor
1873 real issues in debug logging.
1875 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1876 assignment on my part. Fixed.
1878 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1879 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1880 Janne Snabb (who went above and beyond: thank you).
1882 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1883 string otherwise requires a connection and a bunch more work and it's
1884 relatively easy to get wrong. Should also expose TLS library linkage
1887 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1888 64-bit ${eval} (JH/03).
1890 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1891 GNU libc to support some of the 64-bit stuff, should not lead to
1892 conflicts. Defined before os.h is pulled in, so if a given platform
1893 needs to override this, it can.
1895 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1896 protection layer was required, which is not implemented.
1897 Bugzilla 1254, patch from Wolfgang Breyha.
1899 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1900 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1901 tls_dhparam take prime identifiers. Also unbreak combination of
1902 OpenSSL+DH_params+TLSSNI.
1904 PP/39 Disable SSLv2 by default in OpenSSL support.
1910 PP/01 Solaris build fix for Oracle's LDAP libraries.
1911 Bugzilla 1109, patch from Stephen Usher.
1913 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1915 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1918 TF/02 Fix a couple more cases where we did not log the error message
1919 when unlink() failed. See also change 4.74-TF/03.
1921 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1922 lock up or crash if it happened to be inside a call to libc when it
1923 got a SIGUSR1 from exiwhat.
1925 The SIGUSR1 handler appends the current process status to the process
1926 log which is later printed by exiwhat. It used to use the general
1927 purpose logging code to do this, but several functions it calls are
1928 not safe for signals.
1930 The new output code in the SIGUSR1 handler is specific to the process
1931 log, and simple enough that it's easy to inspect for signal safety.
1932 Removing some special cases also simplifies the general logging code.
1933 Removing the spurious timestamps from the process log simplifies
1936 TF/04 Improved ratelimit ACL condition.
1938 The /noupdate option has been deprecated in favour of /readonly which
1939 has clearer semantics. The /leaky, /strict, and /readonly update modes
1940 are mutually exclusive. The update mode is no longer included in the
1941 database key; it just determines when the database is updated. (This
1942 means that when you upgrade Exim will forget old rate measurements.)
1944 Exim now checks that the per_* options are used with an update mode that
1945 makes sense for the current ACL. For example, when Exim is processing a
1946 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1947 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1948 must specify per_mail/readonly. If you omit the update mode it defaults to
1949 /leaky where that makes sense (as before) or /readonly where required.
1951 The /noupdate option is now undocumented but still supported for
1952 backwards compatibility. It is equivalent to /readonly except that in
1953 ACLs where /readonly is required you may specify /leaky/noupdate or
1954 /strict/noupdate which are treated the same as /readonly.
1956 A useful new feature is the /count= option. This is a generalization
1957 of the per_byte option, so that you can measure the throughput of other
1958 aggregate values. For example, the per_byte option is now equivalent
1959 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1961 The per_rcpt option has been generalized using the /count= mechanism
1962 (though it's more complicated than the per_byte equivalence). When it is
1963 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1964 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1965 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1966 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1967 Note that using per_rcpt with a non-readonly update mode in more than
1968 one ACL will cause the recipients to be double-counted. (The per_mail
1969 and per_byte options don't have this problem.)
1971 The handling of very low rates has changed slightly. If the computed rate
1972 is less than the event's count (usually one) then this event is the first
1973 after a long gap. In this case the rate is set to the same as this event's
1974 count, so that the first message of a spam run is counted properly.
1976 The major new feature is a mechanism for counting the rate of unique
1977 events. The new per_addr option counts the number of different
1978 recipients that someone has sent messages to in the last time period. It
1979 behaves like per_rcpt if all the recipient addresses are different, but
1980 duplicate recipient addresses do not increase the measured rate. Like
1981 the /count= option this is a general mechanism, so the per_addr option
1982 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1983 example, measure the rate that a client uses different sender addresses
1984 with the options per_mail/unique=$sender_address. There are further
1985 details in the main documentation.
1987 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1989 TF/06 Removed a few PCRE remnants.
1991 TF/07 Automatically extract Exim's version number from tags in the git
1992 repository when doing development or release builds.
1994 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1995 Bugzilla 879. Patch from Paul Fisher.
1997 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1998 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2001 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2003 PP/05 Variable $av_failed, true if the AV scanner deferred.
2004 Bugzilla 1078. Patch from John Horne.
2006 PP/06 Stop make process more reliably on build failure.
2007 Bugzilla 1087. Patch from Heiko Schlittermann.
2009 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2010 Bugzilla 1089. Patch from Heiko Schlittermann.
2012 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2013 Bugzilla 1131. Patch from Holger Weiß.
2015 PP/09 Handle IPv6 addresses with SPF.
2016 Bugzilla 860. Patch from Wolfgang Breyha.
2018 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2020 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2023 PP/11 match_* no longer expand right-hand-side by default.
2024 New compile-time build option, EXPAND_LISTMATCH_RHS.
2025 New expansion conditions, "inlist", "inlisti".
2027 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2029 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2031 PP/14 fix log_write() format string regression from TF/03.
2032 Bugzilla 1152. Patch from Dmitry Isaikin.
2038 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2040 PP/02 Harmonised TLS library version reporting; only show if debugging.
2041 Layout now matches that introduced for other libraries in 4.74 PP/03.
2043 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2045 PP/04 New "dns_use_edns0" global option.
2047 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2050 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2051 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2053 TK/01 Updated PolarSSL code to 0.14.2.
2054 Bugzilla 1097. Patch from Andreas Metzler.
2056 PP/07 Catch divide-by-zero in ${eval:...}.
2057 Fixes bugzilla 1102.
2059 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2062 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2063 format-string attack -- SECURITY: remote arbitrary code execution.
2065 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2066 time unintentionally subject to list matching rules, letting the header
2067 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2068 arbitrary string expansion). This allowed for information disclosure.
2070 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2071 INT_MIN/-1 -- value coerced to INT_MAX.
2077 NM/01 Workaround for PCRE version dependency in version reporting
2080 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2081 This fixes portability to compilers other than gcc, notably
2082 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2084 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2085 makefiles for portability to HP-UX and POSIX correctness.
2087 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2088 Also via indented variable definition in the Makefile.
2089 (Debugging by Oliver Heesakkers).
2091 PP/02 Restore caching of spamd results with expanded spamd_address.
2092 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2094 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2095 Improves build reliability. Fix from: Frank Elsner
2097 NM/02 Fix wide character breakage in the rfc2047 coding
2098 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2100 NM/03 Allow underscore in dnslist lookups
2101 Fixes bug 1026. Patch from Graeme Fowler
2103 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2104 Code patches from Adam Ciarcinski of NetBSD.
2106 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2109 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2110 is logged, to avoid truncation. Patch from John Horne.
2112 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2113 Patch from Jakob Hirsch.
2115 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2116 SQL string expansion failure details.
2117 Patch from Andrey Oktyabrski.
2119 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2120 Patch from Simon Arlott.
2122 PP/09 New lookups functionality failed to compile on old gcc which rejects
2123 extern declarations in function scope.
2124 Patch from Oliver Fleischmann
2126 PP/10 Use sig_atomic_t for flags set from signal handlers.
2127 Check getgroups() return and improve debugging.
2128 Fixed developed for diagnosis in bug 927 (which turned out to be
2131 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2132 Patch from Mark Zealey.
2134 PP/12 Bugzilla 1056: Improved spamd server selection.
2135 Patch from Mark Zealey.
2137 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2138 Based on patch from Heiko Schlittermann.
2140 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2141 Patch from Uwe Doering, sign-off by Michael Haardt.
2143 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2144 variable declaration deep within a block. Bug and patch from
2147 PP/15 lookups-Makefile IRIX compatibility coercion.
2149 PP/16 Make DISABLE_DKIM build knob functional.
2151 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2152 Patch by Simon Arlott
2154 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2155 variable argument macros. Our copy now differs from upstream.
2161 TF/01 Failure to get a lock on a hints database can have serious
2162 consequences so log it to the panic log.
2164 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2165 controlled using the smtp_confirmation log selector.
2167 TF/03 Include the error message when we fail to unlink a spool file.
2169 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2170 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2171 for maintaining out-of-tree patches for some time.
2173 PP/01 Bugzilla 139: Documentation and portability issues.
2174 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2175 Handle per-OS dynamic-module compilation flags.
2177 PP/02 Let /dev/null have normal permissions.
2178 The 4.73 fixes were a little too stringent and complained about the
2179 permissions on /dev/null. Exempt it from some checks.
2180 Reported by Andreas M. Kirchwitz.
2182 PP/03 Report version information for many libraries, including
2183 Exim version information for dynamically loaded libraries. Created
2184 version.h, now support a version extension string for distributors
2185 who patch heavily. Dynamic module ABI change.
2187 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2188 privilege escalation vulnerability whereby the Exim run-time user
2189 can cause root to append content of the attacker's choosing to
2192 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2195 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2196 If dropping privileges for untrusted macros, we disabled normal logging
2197 on the basis that it would fail; for the Exim run-time user, this is not
2198 the case, and it resulted in successful deliveries going unlogged.
2199 Fixed. Reported by Andreas Metzler.
2205 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2206 only prepend for the Resent-* case. Fixes regression introduced in
2207 Exim 4.70 by NM/22 for Bugzilla 607.
2209 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2210 increasing numbers of administrators be bitten by this.
2212 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2214 PP/03 Bugzilla 994: added openssl_options main configuration option.
2216 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2218 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2220 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2222 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2223 without a peer certificate, leading to a segfault because of an
2224 assumption that peers always have certificates. Be a little more
2225 paranoid. Problem reported by Martin Tscholak.
2227 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2228 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2229 NB: ClamAV planning to remove STREAM in "middle of 2010".
2230 CL also introduces -bmalware, various -d+acl logging additions and
2231 more caution in buffer sizes.
2233 PP/09 Implemented reverse_ip expansion operator.
2235 PP/10 Bugzilla 937: provide a "debug" ACL control.
2237 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2239 PP/12 Bugzilla 973: Implement --version.
2241 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2243 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2245 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2247 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2248 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2249 ignore trailing whitespace.
2251 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2253 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2256 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2257 Notification from Dr Andrew Aitchison.
2259 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2260 ExtendedDetectionInfo response format.
2261 Notification from John Horne.
2263 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2266 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2267 XSL and documented dependency on system catalogs, with examples of how
2270 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2273 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2274 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2275 configuration file which is writeable by the Exim user or group.
2277 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2278 of configuration files to cover files specified with the -C option if
2279 they are going to be used with root privileges, not just the default
2282 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2283 option (effectively making it always true).
2285 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2286 files to be used while preserving root privileges.
2288 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2289 that rogue child processes cannot use them.
2291 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2292 run-time user, instead of root.
2294 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2295 Exim run-time user without dropping privileges.
2297 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2298 result string, instead of calling string_vformat() twice with the same
2301 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2302 for other users. Others should always drop root privileges if they use
2303 -C on the command line, even for a whitelisted configure file.
2305 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2307 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2313 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2314 $data_path, and $header_path variables; fixed documentation bugs and
2317 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2318 exipick to access non-standard spools, including the "frozen" queue
2321 NM/01 Bugzilla 965: Support mysql stored procedures.
2322 Patch from Alain Williams
2324 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2326 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2327 Patch from Andreas Metzler
2329 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2330 Patch from Kirill Miazine
2332 NM/05 Bugzilla 671: Added umask to procmail example.
2334 JJ/03 installed exipick 20100323.0, fixing doc bug
2336 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2337 directory. Notification and patch from Dan Rosenberg.
2339 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2341 TK/02 Improve log output when DKIM signing operation fails.
2343 MH/01 Treat the transport option dkim_domain as a colon separated
2344 list, not as a single string, and sign the message with each element,
2345 omitting multiple occurences of the same signer.
2347 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2348 Bugzilla 985, 986. Patch by Simon Arlott
2350 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2351 Patch by Simon Arlott
2353 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2354 MBX locking. Notification from Dan Rosenberg.
2360 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2362 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2364 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2366 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2368 NM/04 Bugzilla 915: Flag broken perl installation during build.
2374 TK/01 Added patch by Johannes Berg that expands the main option
2375 "spamd_address" if it starts with a dollar sign.
2377 TK/02 Write list of recipients to X-Envelope-Sender header when building
2378 the mbox-format spool file for content scanning (suggested by Jakob
2381 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2382 (http://www.dcc-servers.net/) support via dccifd. Activated by
2383 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2385 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2386 by Mark Daniel Reidel <mr@df.eu>.
2388 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2389 When building exim an external PCRE library is now needed -
2390 PCRE is a system library on the majority of modern systems.
2391 See entry on PCRE_LIBS in EDITME file.
2393 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2394 conversation. Added nologin parameter to request.
2395 Patch contributed by Kirill Miazine.
2397 TF/01 Do not log submission mode rewrites if they do not change the address.
2399 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2401 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2402 log files in place. Contributed by Roberto Lima.
2404 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2406 TF/03 Bugzilla 615: When checking the local_parts router precondition
2407 after a local_part_suffix or local_part_prefix option, Exim now
2408 does not use the address's named list lookup cache, since this
2409 contains cached lookups for the whole local part.
2411 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2412 Robert Millan. Documentation is in experimental-spec.txt.
2414 TF/04 Bugzilla 668: Fix parallel build (make -j).
2416 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2418 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2419 Patch provided by Jan Srzednicki.
2421 TF/05 Leading white space used to be stripped from $spam_report which
2422 wrecked the formatting. Now it is preserved.
2424 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2425 that they are available at delivery time.
2427 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2429 TF/08 TLS error reporting now respects the incoming_interface and
2430 incoming_port log selectors.
2432 TF/09 Produce a more useful error message if an SMTP transport's hosts
2433 setting expands to an empty string.
2435 NM/06 Bugzilla 744: EXPN did not work under TLS.
2436 Patch provided by Phil Pennock.
2438 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2439 Patch provided by Richard Godbee.
2441 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2442 acl_smtp_notquit, added index entry.
2444 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2445 Patch provided by Eugene Bujak.
2447 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2448 accept(). Patch provided by Maxim Dounin.
2450 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2451 Patch provided by Phil Pennock.
2453 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2455 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2456 Patch provided by Brad "anomie" Jorsch.
2458 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2459 Patch provided by Dean Brooks.
2461 TK/05 Add native DKIM support (does not depend on external libraries).
2463 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2464 Patch provided by Graeme Fowler.
2466 NM/16 Bugzilla 851: Documentation example syntax fix.
2468 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2470 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2473 NM/19 Bugzilla 745: TLS version reporting.
2474 Patch provided by Phil Pennock.
2476 NM/20 Bugzilla 167: bool: condition support.
2477 Patch provided by Phil Pennock.
2479 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2480 clients. Patch provided by Phil Pennock.
2482 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2483 Patch provided by Brad "anomie" Jorsch.
2485 NM/23 Bugzilla 687: Fix misparses in eximstats.
2486 Patch provided by Heiko Schlittermann.
2488 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2489 Patch provided by Heiko Schlittermann.
2491 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2492 plus update to original patch.
2494 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2496 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2497 Patch provided by David Brownlee.
2499 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2501 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2503 NM/30 Bugzilla 888: TLS documentation bugfixes.
2505 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2507 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2508 Unlike the original bugzilla I have changed all shell scripts in src tree.
2510 NM/33 Bugzilla 898: Transport filter timeout fix.
2511 Patch by Todd Rinaldo.
2513 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2514 Patch by Serge Demonchaux.
2516 NM/35 Bugzilla 39: Base64 decode bug fixes.
2517 Patch by Jakob Hirsch.
2519 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2521 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2523 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2525 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2531 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2532 ALT-N's libdkim that I have put here:
2533 http://duncanthrax.net/exim-experimental/
2535 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2536 were called 'true' and it seems that C99 defines that as a reserved
2537 keyword to be used with 'bool' variable types. That means you could
2538 not include C99-style headers which use bools without triggering
2539 build errors in sieve.c.
2541 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2542 as mailq or other aliases. Changed the --help handling significantly
2543 to do whats expected. exim_usage() emits usage/help information.
2545 SC/01 Added the -bylocaldomain option to eximstats.
2547 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2549 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2551 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2557 PH/01 Another patch from the Sieve maintainer.
2559 PH/02 When an IPv6 address is converted to a string for single-key lookup
2560 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2561 dots are used instead of colons so that keys in lsearch files need not
2562 contain colons. This was done some time before quoting was made available
2563 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2564 (notated using the quote facility) so as to distinguish them from IPv4
2565 keys. This meant that lookups for IP addresses in host lists did not work
2566 for iplsearch lookups.
2568 This has been fixed by arranging for IPv6 addresses to be expressed with
2569 colons if the lookup type is iplsearch. This is not incompatible, because
2570 previously such lookups could never work.
2572 The situation is now rather anomalous, since one *can* have colons in
2573 ordinary lsearch keys. However, making the change in all cases is
2574 incompatible and would probably break a number of configurations.
2576 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2579 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2580 conversion specification without a maximum field width, thereby enabling
2581 a rogue spamd server to cause a buffer overflow. While nobody in their
2582 right mind would setup Exim to query an untrusted spamd server, an
2583 attacker that gains access to a server running spamd could potentially
2584 exploit this vulnerability to run arbitrary code as the Exim user.
2586 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2587 $primary_hostname instead of what libspf2 thinks the hosts name is.
2589 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2590 a directory entry by the name of the lookup key. Previously, if a
2591 symlink pointed to a non-existing file or a file in a directory that
2592 Exim lacked permissions to read, a lookup for a key matching that
2593 symlink would fail. Now it is enough that a matching directory entry
2594 exists, symlink or not. (Bugzilla 503.)
2596 PH/03 The body_linecount and body_zerocount variables are now exported in the
2599 PH/04 Added the $dnslist_matched variable.
2601 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2602 This means they are set thereafter only if the connection becomes
2605 PH/06 Added the client_condition to authenticators so that some can be skipped
2606 by clients under certain conditions.
2608 PH/07 The error message for a badly-placed control=no_multiline_responses left
2609 "_responses" off the end of the name.
2611 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2613 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2614 (without spaces) instead of just copying the configuration text.
2616 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2618 PH/11 Added $max_received_linelength.
2620 PH/12 Added +ignore_defer and +include_defer to host lists.
2622 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2623 way in which PCRE > 7.0 is built.
2625 PH/14 Implemented queue_only_load_latch.
2627 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2628 MAIL command. The effect was to mangle the value on 64-bit systems.
2630 PH/16 Another patch from the Sieve maintainer.
2632 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2634 PH/18 If a system quota error occurred while trying to create the file for
2635 a maildir delivery, the message "Mailbox is full" was not appended to the
2636 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2637 only to a quota excession during the actual writing of the file.
2639 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2640 characters?) which causes problems in log lines. The DN values are now
2641 passed through string_printing() before being added to log lines.
2643 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2644 and InterBase are left for another time.)
2646 PH/21 Added message_body_newlines option.
2648 PH/22 Guard against possible overflow in moan_check_errorcopy().
2650 PH/23 POSIX allows open() to be a macro; guard against that.
2652 PH/24 If the recipient of an error message contained an @ in the local part
2653 (suitably quoted, of course), incorrect values were put in $domain and
2654 $local_part during the evaluation of errors_copy.
2660 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2661 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2664 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2665 issue a MAIL command.
2667 PH/02 In an ACL statement such as
2669 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2671 if a client was not listed at all, or was listed with a value other than
2672 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2673 the condition was not true (as it should be), so access was not denied.
2674 The bug was that the ! inversion was incorrectly passed on to the second
2675 item. This has been fixed.
2677 PH/03 Added additional dnslists conditions == and =& which are different from
2678 = and & when the dns lookup returns more than one IP address.
2680 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2681 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2683 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2684 FSYNC, which compiles an option called disable_fsync that allows for
2685 bypassing fsync(). The documentation is heavily laced with warnings.
2687 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2689 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2690 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2691 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2692 including adding "make clean"; (3) Added -fPIC when compiling the test
2693 dynamically loaded module, to get rid of a warning.
2695 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2696 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2697 = 0s. The bug is otherwise harmless.
2699 PH/07 There was a bug in the dovecot authenticator such that the value of
2700 $auth1 could be overwritten, and so not correctly preserved, after a
2701 successful authentication. This usually meant that the value preserved by
2702 the server_setid option was incorrect.
2704 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2706 PH/09 Installed PCRE release 7.0.
2708 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2709 run for batched SMTP input. It is now run at the start of every message
2710 in the batch. While fixing this I discovered that the process information
2711 (output by running exiwhat) was not always getting set for -bs and -bS
2712 input. This is fixed, and it now also says "batched" for BSMTP.
2714 PH/11 Added control=no_pipelining.
2716 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2717 patch, slightly modified), and move the expansion of helo_data till after
2718 the connection is made in the smtp transport (so it can use these
2721 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2723 PH/14 Added log_selector = +pid.
2725 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2727 PH/16 Add ${if forany and ${if forall.
2729 PH/17 Added dsn_from option to vary the From: line in DSNs.
2731 PH/18 Flush SMTP output before performing a callout, unless control =
2732 no_callout_flush is set.
2734 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2735 was true (the default) a successful delivery failed to delete the retry
2736 item, thus causing premature timeout of the address. The bug is now
2739 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2741 PH/21 Long custom messages for fakedefer and fakereject are now split up
2742 into multiline responses in the same way that messages for "deny" and
2743 other ACL rejections are.
2745 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2746 with slight modification.
2748 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2749 draft, changing the syntax and factoring some duplicate code".
2751 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2752 for deliveries of the second and subsequent messages over the same SMTP
2755 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2756 ${reduce, with only minor "tidies".
2758 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2760 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2761 expansion side effects.
2763 PH/27 When a message times out after an over-quota error from an Exim-imposed
2764 quota, the bounce message says "mailbox is full". This message was not
2765 being given when it was a system quota that was exceeded. It now should
2768 MH/03 Made $recipients available in local_scan(). local_scan() already has
2769 better access to the recipient list through recipients_list[], but
2770 $recipients can be useful in postmaster-provided expansion strings.
2772 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2773 in the case of a MAIL command with additional options following the
2774 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2775 were accidentally chopped off.
2777 PH/29 SMTP synchronization checks are implemented when a command is read -
2778 there is a check that no more input is waiting when there shouldn't be
2779 any. However, for some commands, a delay in an ACL can mean that it is
2780 some time before the response is written. In this time, more input might
2781 arrive, invalidly. So now there are extra checks after an ACL has run for
2782 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2783 pipelining has not been advertised.
2785 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2787 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2788 returned in the SMTP response when smtp_return_error_details was false.
2789 This has been fixed.
2791 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2792 instead of the C I/O that was originally supplied, because problems were
2793 reported on Solaris.
2795 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2796 Exim which did not show up earlier: it was assuming that a call to
2797 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2798 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2799 was a macro that became an assignment, so it seemed to work. This has
2800 changed to a proper function call with a void return, hence the compile
2801 error. Exim's code has been fixed.
2803 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2806 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2808 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2810 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2811 because it turns out to be too big or there is a timeout) there is no
2812 "Completed" line in the log. When some messages of this type were
2813 selected by exigrep, they were listed as "not completed". Others were
2814 picked up by some special patterns. I have improved the selection
2815 criteria to be more general.
2817 PH/38 The host_find_failed option in the manualroute router can now be set
2818 to "ignore", to completely ignore a host whose IP address cannot be
2819 found. If all hosts are ignored, the behaviour is controlled by the new
2820 host_all_ignored option.
2822 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2823 homing or because of multiple MX records with /mx) generated more than
2824 one IP address, and the following item turned out to be the local host,
2825 all the secondary addresses of the first item were incorrectly removed
2826 from the list, along with the local host and any following hosts (which
2827 is what is supposed to happen).
2829 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2830 whoever called Exim into the -H file. In the case of the daemon it was
2831 behaving confusingly. When first started, it used values for whoever
2832 started the daemon, but after a SIGHUP it used the Exim user (because it
2833 calls itself on a restart). I have changed the code so that it now always
2836 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2837 message are rejected with the same error (e.g. no authentication or bad
2838 sender address), and a DATA command is nevertheless sent (as can happen
2839 with PIPELINING or a stupid MUA), the error message that was given to the
2840 RCPT commands is included in the rejection of the DATA command. This is
2841 intended to be helpful for MUAs that show only the final error to their
2844 PH/42 Another patch from the Sieve maintainer.
2846 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2847 Eximstats - Fixed some broken HTML links and added missing column headers
2849 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2850 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2852 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2854 PH/43 Yet another patch from the Sieve maintainer.
2856 PH/44 I found a way to check for a TCP/IP connection going away before sending
2857 the response to the final '.' that terminates a message, but only in the
2858 case where the client has not sent further data following the '.'
2859 (unfortunately, this is allowed). However, in many cases there won't be
2860 any further data because there won't be any more messages to send. A call
2861 to select() can be used: if it shows that the input is "ready", there is
2862 either input waiting, or the socket has been closed. An attempt to read
2863 the next input character can distinguish the two cases. Previously, Exim
2864 would have sent an OK response which the client would never have see.
2865 This could lead to message repetition. This fix should cure that, at
2866 least in a lot of common cases.
2868 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2869 advertised in response to EHLO.
2875 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2876 fixed by 4.65/MH/01 (is this a record?) are fixed:
2878 (i) An empty string was always treated as zero by the numeric comparison
2879 operators. This behaviour has been restored.
2881 (ii) It is documented that the numeric comparison operators always treat
2882 their arguments as decimal numbers. This was broken in that numbers
2883 starting with 0 were being interpreted as octal.
2885 While fixing these problems I realized that there was another issue that
2886 hadn't been noticed. Values of message_size_limit (both the global option
2887 and the transport option) were treated as octal if they started with 0.
2888 The documentation was vague. These values are now always treated as
2889 decimal, and I will make that clear in the documentation.
2895 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2896 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2899 MH/01 Don't check that the operands of numeric comparison operators are
2900 integers when their expansion is in "skipping" mode (fixes bug
2901 introduced by 4.64-PH/07).
2903 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2904 child addresses, Exim now panics and dies. Previously, because the count
2905 is held in a short int, deliveries were likely to be lost. As such a
2906 large number of recipients for a single message is ridiculous
2907 (performance will be very, very poor), I have chosen to impose a limit
2908 rather than extend the field.
2914 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2915 leftover -K file (the existence of which was triggered by #402).
2916 While we were at it, introduced process PID as part of the -K
2917 filename. This should rule out race conditions when creating
2920 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2921 processing considerably. Previous code took too long for large mails,
2922 triggering a timeout which in turn triggers #401.
2924 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2925 in the DK code in transports.c. sendfile() is not really portable,
2926 hence the _LINUX specificness.
2928 TF/01 In the add_headers option to the mail command in an Exim filter,
2929 there was a bug that Exim would claim a syntax error in any
2930 header after the first one which had an odd number of characters
2933 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2934 callout verification, Exim cached a "reject" for the entire domain. This
2935 is correct for most verifications, but it is not correct for a recipient
2936 verification with use_sender or use_postmaster set, because in that case
2937 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2938 case of MAIL FROM:<> rejection from other early rejections (e.g.
2939 rejection of HELO). When verifying a recipient using a non-null MAIL
2940 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2941 Whatever the result of the callout, the value of the domain cache is
2942 left unchanged (for any other kind of callout, getting as far as trying
2943 RCPT means that the domain itself is ok).
2945 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2948 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2949 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2950 session, the daemon ignores ECONNECT errors and logs others; it now
2951 ignores EPIPE as well.
2953 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2954 (quoted-printable decoding).
2956 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2957 later the small subsequent patch to fix an introduced bug.
2959 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2961 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2963 PH/08 An error is now given if message_size_limit is specified negative.
2965 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2966 to be given (somewhat) arbitrary names.
2968 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2971 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2972 miscellaneous code fixes
2974 PH/10 Added the log_reject_target ACL modifier to specify where to log
2977 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2978 hostname. This is wrong, because it relates to the incoming message (and
2979 probably the interface on which it is arriving) and not to the outgoing
2980 callout (which could be using a different interface). This has been
2981 changed to use the value of the helo_data option from the smtp transport
2982 instead - this is what is used when a message is actually being sent. If
2983 there is no remote transport (possible with a router that sets up host
2984 addresses), $smtp_active_hostname is used.
2986 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2987 tweaks were necessary in order to get it to work (see also 21 below):
2988 (a) The code assumed that strncpy() returns a negative number on buffer
2989 overflow, which isn't the case. Replaced with Exim's string_format()
2991 (b) There were several signed/unsigned issues. I just did the minimum
2992 hacking in of casts. There is scope for a larger refactoring.
2993 (c) The code used strcasecmp() which is not a standard C function.
2994 Replaced with Exim's strcmpic() function.
2995 (d) The code set only $1; it now sets $auth1 as well.
2996 (e) A simple test gave the error "authentication client didn't specify
2997 service in request". It would seem that Dovecot has changed its
2998 interface. Fortunately there's a specification; I followed it and
2999 changed what the client sends and it appears to be working now.
3001 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3004 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3005 address A is aliased to B and C, where B exists and C does not. Without
3006 -v the output is "A verified" because verification stops after a
3007 successful redirection if more than one address is generated. However,
3008 with -v the child addresses are also verified. Exim was outputting "A
3009 failed to verify" and then showing the successful verification for C,
3010 with its parentage. It now outputs "B failed to verify", showing B's
3011 parentage before showing the successful verification of C.
3013 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3014 look up a TXT record in a specific list after matching in a combined
3017 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3018 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3019 they consult the DNS. I had assumed they would set it the way they
3020 wanted; and indeed my experiments on Linux seem to show that in some
3021 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3022 To be on the safe side, however, I have now made the interface to
3023 host_find_byname() similar to host_find_bydns(), with an argument
3024 containing the DNS resolver options. The host_find_byname() function now
3025 sets these options at its start, just as host_find_bydns() does. The smtp
3026 transport options dns_qualify_single and dns_search_parents are passed to
3027 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3028 of host_find_byname() use the default settings of RES_DEFNAMES
3029 (qualify_single) but not RES_DNSRCH (search_parents).
3031 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3032 spool_read_header() do less string testing, by means of a preliminary
3033 switch on the second character of optional "-foo" lines. (This is
3034 overdue, caused by the large number of possibilities that now exist.
3035 Originally there were few.) While I was there, I also converted the
3036 str(n)cmp tests so they don't re-test the leading "-" and the first
3037 character, in the hope this might squeeze out yet more improvement.
3039 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3040 flag allowing group syntax was set by the header_syntax check but not
3041 turned off, possible causing trouble later; (2) The flag was not being
3042 set at all for the header_verify test, causing "group"-style headers to
3043 be rejected. I have now set it in this case, and also caused header_
3044 verify to ignore an empty address taken from a group. While doing this, I
3045 came across some other cases where the code for allowing group syntax
3046 while scanning a header line wasn't quite right (mostly, not resetting
3047 the flag correctly in the right place). These bugs could have caused
3048 trouble for malformed header lines. I hope it is now all correct.
3050 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3051 with the "reply" argument non-NULL. The code, however (which originally
3052 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3053 but it didn't always do it. This confused somebody who was copying the
3054 code for some other use. I have removed all the tests.
3056 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3057 feature that was used to support insecure browsers during the U.S. crypto
3058 embargo. It requires special client support, and Exim is probably the
3059 only MTA that supported it -- and would never use it because real RSA is
3060 always available. This code has been removed, because it had the bad
3061 effect of slowing Exim down by computing (never used) parameters for the
3062 RSA_EXPORT functionality.
3064 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3065 authenticator to fail if there's a tab character in the incoming data
3066 (there should never be unless someone is messing about, as it's supposed
3067 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3068 if the connection is using TLS or if the remote IP is the same as the
3069 local IP, and the "valid-client-cert option" if a client certificate has
3072 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3073 authenticators. This can be used for authorization after authentication
3074 succeeds. (In the case of plaintext, it servers for both authentication
3077 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3078 if any retry times were supplied.
3080 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3081 connection, where there is no client IP address to check. In this
3082 situation, the verify now always succeeds.
3084 PH/25 Applied John Jetmore's -Mset patch.
3086 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3088 PH/27 In a string expansion for a processed (not raw) header when multiple
3089 headers of the same name were present, leading whitespace was being
3090 removed from all of them, but trailing whitespace was being removed only
3091 from the last one. Now trailing whitespace is removed from each header
3092 before concatenation. Completely empty headers in a concatenation (as
3093 before) are ignored.
3095 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3096 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3098 PH/29 [Removed. This was a change that I later backed out, and forgot to
3099 correct the ChangeLog entry (that I had efficiently created) before
3100 committing the later change.]
3102 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3103 address errors (4xx response to RCPT) over the same connection as other
3104 messages routed to the same hosts. Such deliveries are always "forced",
3105 so retry times are not inspected. This resulted in far too many retries
3106 for the affected addresses. The effect occurred only when there were more
3107 hosts than the hosts_max_try setting in the smtp transport when it had
3108 the 4xx errors. Those hosts that it had tried were not added to the list
3109 of hosts for which the message was waiting, so if all were tried, there
3110 was no problem. Two fixes have been applied:
3112 (i) If there are any address or message errors in an SMTP delivery, none
3113 of the hosts (tried or untried) are now added to the list of hosts
3114 for which the message is waiting, so the message should not be a
3115 candidate for sending over the same connection that was used for a
3116 successful delivery of some other message. This seems entirely
3117 reasonable: after all the message is NOT "waiting for some host".
3118 This is so "obvious" that I'm not sure why it wasn't done
3119 previously. Hope I haven't missed anything, but it can't do any
3120 harm, as the worst effect is to miss an optimization.
3122 (ii) If, despite (i), such a delivery is accidentally attempted, the
3123 routing retry time is respected, so at least it doesn't keep
3124 hammering the server.
3126 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3127 in ${readsocket because some servers need this prod.
3129 PH/32 Added some extra debug output when updating a wait-xxx database.
3131 PH/33 The hint "could be header name not terminated by colon", which has been
3132 given for certain expansion errors for a long time, was not being given
3133 for the ${if def:h_colon_omitted{... case.
3135 PH/34 The spec says: "With one important exception, whenever a domain list is
3136 being scanned, $domain contains the subject domain." There was at least
3137 one case where this was not true.
3139 PH/35 The error "getsockname() failed: connection reset by peer" was being
3140 written to the panic log as well as the main log, but it isn't really
3141 panic-worthy as it just means the connection died rather early on. I have
3142 removed the panic log writing for the ECONNRESET error when getsockname()
3145 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3146 runs only) independently of the message's sender address. This meant
3147 that, if the 4xx error was in fact related to the sender, a different
3148 message to the same recipient with a different sender could confuse
3149 things. In particular, this can happen when sending to a greylisting
3150 server, but other circumstances could also provoke similar problems.
3151 I have changed the default so that the retry time for these errors is now
3152 based a combination of the sender and recipient addresses. This change
3153 can be overridden by setting address_retry_include_sender=false in the
3156 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3157 remote server are returned as part of bounce messages. This was not
3158 happening for LMTP over a pipe (the lmtp transport), but now it is the
3159 same for both kinds of LMTP.
3161 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3162 in header lines that were in fact CNAMEs. This is no longer the case.
3164 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3165 and queue runs started by the daemon processed all messages. This has
3166 been fixed so that -R and -S can now usefully be given with -q<time>.
3168 PH/40 Import PCRE release 6.7 (fixes some bugs).
3170 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3172 PH/42 Give an error if -q is specified more than once.
3174 PH/43 Renamed the variables $interface_address and $interface_port as
3175 $received_ip_address and $received_port, to make it clear that these
3176 values apply to message reception, and not to the outgoing interface when
3177 a message is delivered. (The old names remain recognized, of course.)
3179 PH/44 There was no timeout on the connect() call when using a Unix domain
3180 socket in the ${readsocket expansion. There now is.
3182 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3183 be meaningful with "accept".
3185 SC/01 Eximstats V1.43
3186 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3188 SC/02 Eximstats V1.44
3189 Use a glob alias rather than an array ref in the generated
3190 parser. This improves both readability and performance.
3192 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3193 Collect SpamAssassin and rejection statistics.
3194 Don't display local sender or destination tables unless
3195 there is data to show.
3196 Added average volumes into the top table text output.
3198 SC/04 Eximstats V1.46
3199 Collect data on the number of addresses (recipients)
3200 as well as the number of messages.
3202 SC/05 Eximstats V1.47
3203 Added 'Message too big' to the list of mail rejection
3204 reasons (thanks to Marco Gaiarin).
3206 SC/06 Eximstats V1.48
3207 Mainlog lines which have GMT offsets and are too short to
3208 have a flag are now skipped.
3210 SC/07 Eximstats V1.49 (Alain Williams)
3211 Added the -emptyok flag.
3213 SC/08 Eximstats V1.50
3214 Fixes for obtaining the IP address from reject messages.
3216 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3217 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3218 whitespace changes from 4.64-PH/27
3220 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3223 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3224 are found, allow negative numbers in numeric criteria)
3226 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3228 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3229 to match changes made in 4.64-PH/43
3231 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3233 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3234 "domain" is actually an IP address, and force a failure. This locks out
3235 those revolvers/nameservers that support "A-for-A" lookups, in
3236 contravention of the specifications.
3238 PH/48 When a host name was looked up from an IP address, and the subsequent
3239 forward lookup of the name timed out, the host name was left in
3240 $sender_host_name, contrary to the specification.
3242 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3243 restricted to single-key lookups, Exim was not diagnosing an error if
3244 * or *@ was used with a query-style lookup.
3246 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3248 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3249 long ago, but noone interested enough thought of it. Let's just say that
3250 the "1.1" means that there are some new functions that weren't there at
3251 some point in the past.
3253 PH/51 Error processing for expansion failure of helo_data from an smtp
3254 transport during callout processing was broken.
3256 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3257 tested/used via the -bh/-bhc/-bs options.
3259 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3260 bug, fixed in subsequent PCRE releases).
3262 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3263 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3265 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3271 SC/01 Use a glob alias rather than an array ref in eximstats generated
3272 parser. This improves both readability and performance.
3274 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3275 Don't display local sender or destination tables in eximstats unless
3276 there is data to show.
3277 Added average volumes into the eximstats top table text output.
3279 SC/03 Collect data on the number of addresses (recipients) as well
3280 as the number of messages in eximstats.
3282 TF/01 Correct an error in the documentation for the redirect router. Exim
3283 does (usually) call initgroups() when daemonizing.
3285 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3286 with consistent privilege compared to when running as a daemon.
3288 TF/03 Note in the spec that $authenticated_id is not set for local
3289 submissions from trusted users.
3291 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3292 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3294 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3295 by adding some example configuration directives to the default
3296 configuration file. A little bit of work is required to uncomment the
3297 directives and define how usernames and passwords are checked, but
3298 there is now a framework to start from.
3300 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3301 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3302 without this. I don't know how relevant this is to other LDAP libraries.
3304 PH/02 Add the verb name to the "unknown ACL verb" error.
3306 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3308 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3310 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3311 directory not expanded when it should be if an expanded home directory
3312 was set for the address (which is overridden by the transport).
3314 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3317 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3318 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3319 because it is too late at that time, and has no effect.
3321 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3322 security issue with \' (bugzilla #107). I could not use the
3323 PQescapeStringConn() function, because it needs a PGconn value as one of
3326 PH/08 When testing addresses using -bt, indicate those final addresses that
3327 are duplicates that would not cause an additional delivery. At least one
3328 person was confused, thinking that -bt output corresponded to deliveries.
3329 (Suppressing duplicates isn't a good idea as you lose the information
3330 about possibly different redirections that led to the duplicates.)
3332 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3333 systems where poll() doesn't work, in particular OS X.
3335 PH/10 Added more information to debugging output for retry time not reached.
3337 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3338 operations in malware.c.
3340 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3343 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3344 syslog_duplication set false, log lines that would normally be written
3345 both the the main log and to the reject log were not written to syslog at
3348 PH/14 In the default configuration, change the use of "message" in ACL warn
3349 statements to "add_header".
3351 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3352 not followed by a command (e.g. "seen endif").
3354 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3355 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3358 PH/17 Added extra conditions to the default value of delay_warning_condition
3362 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3363 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3364 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3367 The Auto-Submitted: and various List- headers are standardised, whereas I
3368 don't think Precedence: ever was.
3370 PH/18 Refactored debugging code in route_finduser() to show more information,
3371 in particular, the error code if getpwnam() issues one.
3373 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3374 This is apparently needed in addition to the PH/07 change above to avoid
3375 any possible encoding problems.
3377 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3378 but not after initializing Perl.
3380 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3381 output them only if debugging. By default they are written stderr,
3382 apparently, which is not desirable.
3384 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3387 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3390 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3392 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3393 authenticated or an ident call has been made. Suppress the default
3394 values for $authenticated_id and $authenticated_sender (but permit -oMai
3395 and -oMas) when testing with -bh.
3397 PH/24 Re-jigged the order of the tests in the default configuration so that the
3398 tests for valid domains and recipients precede the DNS black list and CSA
3399 tests, on the grounds that those ones are more expensive.
3401 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3402 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3403 This bug exists in every version of Exim that I still have, right back to
3406 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3407 However, an attempt to turn on case-sensitivity in a regex key by
3408 including (?-i) didn't work because the subject string was already
3409 lowercased, and the effects were non-intuitive. It turns out that a
3410 one-line patch can be used to allow (?-i) to work as expected.
3416 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3417 other effects) broke the use of negated acl sub-conditions.
3419 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3422 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3423 "Deny" causes Exim to reject the incoming connection with a 554 error.
3424 Unfortunately, if there is a major crisis, such as a disk failure,
3425 tcp-wrappers gives "deny", whereas what one would like would be some
3426 kind of temporary error. A kludge has been added to help with this.
3427 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3428 554 error is used if errno is still zero or contains ENOENT (which occurs
3429 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3432 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3434 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3435 errors. Otherwise a message that provokes a temporary error (when other
3436 messages do not) can cause a whole host to time out.
3438 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3439 addresses were routed directly to files or pipes from a redirect router.
3440 File deliveries just didn't batch; pipe deliveries might have suffered
3443 PH/06 A failure to get a lock for a hints database would erroneously always say
3444 "Failed to get write lock", even when it was really a read lock.
3446 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3447 of 0600. This has been changed to use the value of the lockfile_mode
3448 option (which defaults to 0600).
3450 PH/08 Applied small patch from the Sieve maintainer.
3452 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3453 folder from quota calculations, a direct delivery into this folder messed
3454 up the contents of the maildirsize file. This was because the regex was
3455 used only to exclude .Trash (or whatever) when the size of the mailbox
3456 was calculated. There was no check that a delivery was happening into an
3457 excluded directory. This bug has been fixed by ignoring all quota
3458 processing for deliveries into excluded directories.
3460 PH/10 Added the maildirfolder_create_regex option to appendfile.
3466 PH/01 The code for finding all the local interface addresses on a FreeBSD
3467 system running IPv6 was broken. This may well have applied to all BSD
3468 systems, as well as to others that have similar system calls. The broken
3469 code found IPv4 interfaces correctly, but gave incorrect values for the
3470 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3471 that it would not match correctly against @[] and not recognize the IPv6
3474 PH/02 The ipliteral router was not recognizing addresses of the form user@
3475 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3477 PH/03 Added disable_ipv6.
3479 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3480 decoded form, because it is most often used to construct To: headers
3481 lines in autoreplies, and the decoded form may well be syntactically
3482 invalid. However, $reply_address has leading white space removed, and all
3483 newlines turned into spaces so that the autoreply transport does not
3486 PH/05 If group was specified without a user on a router, and no group or user
3487 was specified on a transport, the group from the router was ignored.
3489 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3490 for visible compile-time settings that can be used to change these
3491 numbers, for those that want even more. Backwards compatibility with old
3492 spool files has been maintained. However, going back to a previous Exim
3493 release will lost any variables that are in spool files.
3495 PH/07 Two small changes when running in the test harness: increase delay when
3496 passing a TCP/IP connection to a new process, in case the original
3497 process has to generate a bounce, and remove special handling of
3498 127.0.0.2 (sic), which is no longer necessary.
3500 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3501 be the same on different OS.
3503 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3506 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3507 whether --show-vars was specified or not
3509 JJ/02 exipick: Added support for new ACL variable spool format introduced
3512 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3513 syntactically invalid From: or Reply-to: line, and a filter used this to
3514 generate an autoreply, and therefore failed to obtain an address for the
3515 autoreply, Exim could try to deliver to a non-existent relative file
3516 name, causing unrelated and misleading errors. What now happens is that
3517 it logs this as a hard delivery error, but does not attempt to create a
3520 PH/11 The exinext utility has a -C option for testing purposes, but although
3521 the given file was scanned by exinext itself; it wasn't being passed on
3522 when Exim was called.
3524 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3525 an end-of-file indication when reading a command response.
3527 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3528 compiled. In many other places in Exim, IPv6 addresses are always
3529 recognized, so I have changed this. It also means that IPv4 domain
3530 literals of the form [IPV4:n.n.n.n] are now always recognized.
3532 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3533 used if the router is not running as root, for example, when verifying at
3534 ACL time, or when using -bh. The debugging output from this situation was
3535 non-existent - all you got was a failure to exec. I have made two
3538 (a) Failures to set uid/gid, the current directory, or a process leader
3539 in a subprocess such as that created by queryprogram now generate
3540 suitable debugging output when -d is set.
3542 (b) The queryprogram router detects when it is not running as root,
3543 outputs suitable debugging information if -d is set, and then runs
3544 the subprocess without attempting to change uid/gid.
3546 PH/15 Minor change to Makefile for building test_host (undocumented testing
3549 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3550 additional section of a DNS packet that returns MX or SRV records.
3551 Instead, it always explicitly searches for A/AAAA records. This avoids
3552 major problems that occur when a DNS server includes only records of one
3553 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3554 fixed another bug: if SRV records were looked up and the corresponding
3555 address records were *not* found in the additional section, the port
3556 values from the SRV records were lost.
3558 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3559 using the correct key (the original address) when searching the retry
3560 rules in order to find which one to use for generating the retry hint.
3562 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3563 adding the default one. Similarly, if it contains a Reply-To: header, the
3564 errors_reply_to option, if set, is not used.
3566 PH/19 When calculating a retry time, Exim used to measure the "time since
3567 failure" by looking at the "first failed" field in the retry record. Now
3568 it does not use this if it is later than than the arrival time of the
3569 message. Instead it uses the arrival time. This makes for better
3570 behaviour in cases where some deliveries succeed, thus re-setting the
3571 "first failed" field. An example is a quota failure for a huge message
3572 when small messages continue to be delivered. Without this change, the
3573 "time since failure" will always be short, possible causing more frequent
3574 delivery attempts for the huge message than are intended.
3575 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3577 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3578 $1, $2, $3) because the numerical variables can be reset during some
3579 expansion items (e.g. "match"), thereby losing the authentication data.
3581 PH/21 Make -bV show the size of off_t variables so that the test suite can
3582 decide whether to run tests for quotas > 2G.
3584 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3585 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3586 filecount value is greater than 2G or if a quota value is greater than 2G
3587 on a system where the size of off_t is not greater than 4, a panic error
3590 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3591 never match. The debug and -bh output now contains an explicit error
3592 message indicating a malformed IPv4 address or mask.
3594 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3595 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3596 PH/23 above applies.
3598 PH/25 Do not write to syslog when running in the test harness. The only
3599 occasion when this arises is a failure to open the main or panic logs
3600 (for which there is an explicit test).
3602 PH/26 Added the /no_tell option to "control=freeze".
3604 PH/27 If a host name lookup failed very early in a connection, for example, if
3605 the IP address matched host_lookup and the reverse lookup yielded a name
3606 that did not have a forward lookup, an error message of the form "no IP
3607 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3608 could be logged. Now it outputs the IP address instead of "NULL".
3610 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3611 allows the sender and the authenticated sender to be set when
3612 submitting a message from within Exim. Since child_open_exim() is
3613 documented for local_scan(), the new function should be too.
3615 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3616 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3617 results in an empty string is now treated as unset.
3619 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3621 PH/31 Added sender_verify_fail logging option.
3623 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3624 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3625 tidied the source and removed it altogether.
3627 PH/33 When a queue run was abandoned because the load average was too high, a
3628 log line was always written; now it is written only if the queue_run log
3629 selector is set. In addition, the log line for abandonment now contains
3630 information about the queue run such as the pid. This is always present
3631 in "start" and "stop" lines but was omitted from the "abandon" line.
3633 PH/34 Omit spaces between a header name and the colon in the error message that
3634 is given when verify = headers_syntax fails (if there are lots of them,
3635 the message gets confusing).
3637 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3638 names, as there are now some PTR records that contain slashes. This check
3639 is only to protect against broken name servers that fall over on strange
3640 characters, so the fact that it applies to all lookups doesn't matter.
3642 PH/36 Now that the new test suite is complete, we can remove some of the
3643 special code in Exim that was needed for the old test suite. For example,
3644 sorting DNS records because real resolvers return them in an arbitrary
3645 order. The new test suite's fake resolver always returns records in the
3648 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3649 bounces) except when queue_only is set, to avoid logging races between
3650 the different processes.
3652 PH/38 Panic-die if .include specifies a non-absolute path.
3654 PH/39 A tweak to the "H" retry rule from its user.
3656 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3657 a label. They prevented compilation on older perls.
3659 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3660 a warning to be raised on newish perls.
3662 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3663 on queue. Changes to match documented behaviour of showing count of
3664 messages matching specified criteria.
3666 PH/40 Changed the default ident timeout from 30s to 5s.
3668 PH/41 Added support for the use of login_cap features, on those BSD systems
3669 that have them, for controlling the resources used by pipe deliveries.
3671 PH/42 The content-scanning code uses fopen() to create files in which to put
3672 message data. Previously it was not paying any attention to the mode of
3673 the files. Exim runs with umask(0) because the rest of the code creates
3674 files with open(), and sets the required mode explicitly. Thus, these
3675 files were ending up world-writeable. This was not a big issue, because,
3676 being within the spool directory, they were not world-accessible. I have
3677 created a function called modefopen, which takes an additional mode
3678 argument. It sets umask(777), creates the file, chmods it to the required
3679 mode, then resets the umask. All the relevant calls to fopen() in the
3680 content scanning code have been changed to use this function.
3682 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3683 to 24 hours. This avoids potential overflow problems when processing G
3684 and H retry rules. I suspect nobody ever tinkers with this value.
3686 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3688 PH/45 When the plaintext authenticator is running as a client, the server's
3689 challenges are checked to ensure they are valid base64 strings. By
3690 default, the authentication attempt is cancelled if an invalid string is
3691 received. Setting client_ignore_invalid_base64 true ignores these errors.
3692 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3693 they are received. Thus, the responses can be made to depend on the
3694 challenges. If an invalid string is ignored, an empty string is placed in
3697 PH/46 Messages that are created by the autoreply transport now contains a
3698 References: header, in accordance with RFCs 2822 and 3834.
3700 PH/47 Added authenticated_sender_force to the smtp transport.
3702 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3704 PH/49 Installed latest patch from the Sieve maintainer.
3706 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3707 was also set, the appendfile transport was unnecessarily scanning a
3708 directory of message files (e.g. for maildir delivery) to find the count
3709 of files (along with the size), even though it did not need this
3710 information. It now does the scan only if it needs to find either the
3711 size of the count of files.
3713 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3715 PH/52 Two bugs concerned with error handling when the smtp transport is
3718 (i) Exim was not creating retry information for temporary errors given
3719 for individual recipients after the DATA command when the smtp transport
3720 was used in LMTP mode. This meant that they could be retried too
3721 frequently, and not timed out correctly.
3723 (ii) Exim was setting the flag that allows error details to be returned
3724 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3725 recipients that were returned after the DATA command.
3727 PH/53 This is related to PH/52, but is more general: for any failing address,
3728 when detailed error information was permitted to be returned to the
3729 sender, but the error was temporary, then after the final timeout, only
3730 "retry timeout exceeded" was returned. Now it returns the full error as
3731 well as "retry timeout exceeded".
3733 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3734 do this, and (what is worse) MTAs that accept it.
3736 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3737 will now be deprecated.
3739 PH/56 New os.c-cygwin from the Cygwin maintainer.
3741 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3742 formats (previously only available in exim formats via -bpr, -bpru,
3743 and -bpra. Now also available in native and exiqgrep formats)
3745 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3746 with very large, slow to parse queues
3748 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3750 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3752 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3753 responses to authentication challenges, though it was showing the
3754 challenges; (ii) I've removed the CR characters from the debug output for
3757 PH/58 Allow for the insertion of a newline as well as a space when a string
3758 is turned into more than one encoded-word during RFC 2047 encoding. The
3759 Sieve code now uses this.
3761 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3762 data_4xx, lost_connection, tls_required.
3764 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3765 message was being sent as an SMTP response.
3767 PH/61 Add -l and -k options to exicyclog.
3769 PH/62 When verifying, if an address was redirected to one new address, so that
3770 verification continued, and the new address failed or deferred after
3771 having set something in $address_data, the value of $address_data was not
3772 passed back to the ACL. This was different to the case when no
3773 redirection occurred. The value is now passed back in both cases.
3775 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3776 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3777 use login_cap.h, so on its own it isn't the distinguishing feature. The
3778 new name refers directly to the setclassresources() function.
3780 PH/65 Added configuration files for NetBSD3.
3782 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3784 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3785 is preferred over IPv4.
3787 PH/68 The bounce_return_message and bounce_return_body options were not being
3788 honoured for bounces generated during the reception of non-SMTP messages.
3789 In particular, this applied to messages rejected by the ACL. This bug has
3790 been fixed. However, if bounce_return_message is true and bounce_return_
3791 body is false, the headers that are returned for a non-SMTP message
3792 include only those that have been read before the error was detected.
3793 (In the case of an ACL rejection, they have all been read.)
3795 PH/69 The HTML version of the specification is now built in a directory called
3796 spec_html instead of spec.html, because the latter looks like a path with
3797 a MIME-type, and this confuses some software.
3799 PH/70 Catch two compiler warnings in sieve.c.
3801 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3802 function verify_get_ident() calls ip_connect() to connect a socket, but
3803 if the "connect()" function timed out, ip_connect() used to close the
3804 socket. However, verify_get_ident() also closes the socket later, and in
3805 between Exim writes to the log, which may get opened at this point. When
3806 the socket was closed in ip_connect(), the log could get the same file
3807 descriptor number as the socket. This naturally causes chaos. The fix is
3808 not to close the socket in ip_connect(); the socket should be closed by
3809 the function that creates it. There was only one place in the code where
3810 this was missing, in the iplookup router, which I don't think anybody now
3811 uses, but I've fixed it anyway.
3813 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3814 well as to direct DNS lookups. Otherwise the handling of names in host
3815 lists is inconsistent and therefore confusing.
3821 PH/01 Two changes to the default runtime configuration:
3823 (1) Move the checks for relay_from_hosts and authenticated clients from
3824 after to before the (commented out) DNS black list checks.
3826 (2) Add control=submission to the relay_from_hosts and authenticated
3827 clients checks, on the grounds that messages accepted by these
3828 statements are most likely to be submissions.
3830 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3832 (1) Generate an error if the third argument for the ${prvs expansion is
3835 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3838 (3) Reset the variables that are obtained from the first argument of
3839 ${prvscheck and used in the second argument before leaving the code,
3840 because their memory is reclaimed, so using them afterwards may do
3843 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3844 one (it's much easier than Tom thought :-).
3846 (5) Because of (4), we can now allow for the use of $prvscheck_result
3847 inside the third argument.
3849 PH/03 For some reason, the default setting of PATH when running a command from
3850 a pipe transport was just "/usr/bin". I have changed it to
3853 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3854 anything to be listed in the output from -bV.
3856 PH/05 When a filter generated an autoreply, the entire To: header line was
3857 quoted in the delivery log line, like this:
3859 => >A.N.Other <ano@some.domain> <original@ddress> ...
3861 This has been changed so that it extracts the operative address. There
3862 may be more than one such address. If so, they are comma-separated, like
3865 => >ano@some.domain,ona@other.domain <original@ddress> ...
3867 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3868 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3869 not being looked up in the rDNS to get a host name, Exim was showing the
3870 IP address twice in Received: lines, even though the IP addresses were
3871 identical. For example:
3873 Received: from [1.2.3.4] (helo=[1.2.3.4])
3875 However, if the real host name was known, it was omitting the HELO data
3876 if it matched the actual IP address. This has been tidied up so that it
3877 doesn't show the same IP address twice.
3879 PH/07 When both +timestamp and +memory debugging was on, the value given by
3880 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3881 called by the debug printing, thereby overwriting the timestamp buffer.
3882 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3884 PH/08 When the original message was included in an autoreply transport, it
3885 always said "this is a copy of the message, including all the headers",
3886 even if body_only or headers_only was set. It now gives an appropriate
3889 PH/09 Applied a patch from the Sieve maintainer which:
3891 o fixes some comments
3892 o adds the (disabled) notify extension core
3893 o adds some debug output for the result of if/elsif tests
3894 o points to the current vacation draft in the documentation
3895 and documents the missing references header update
3899 o fixes a bug in processing the envelope test (when testing
3900 multiple envelope elements, the last element determined the
3903 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3904 Electronic Mail") by including:
3906 Auto-submitted: auto-generated
3908 in the messages that it generates (bounce messages and others, such as
3909 warnings). In the case of bounce messages for non-SMTP messages, there was
3910 also a typo: it was using "Auto_submitted" (underscore instead of
3911 hyphen). Since every message generated by Exim is necessarily in response
3912 to another message, thes have all been changed to:
3914 Auto-Submitted: auto-replied
3916 in accordance with these statements in the RFC:
3918 The auto-replied keyword:
3920 - SHOULD be used on messages sent in direct response to another
3921 message by an automatic process,
3923 - MUST NOT be used on manually-generated messages,
3925 - MAY be used on Delivery Status Notifications (DSNs) and Message
3926 Disposition Notifications (MDNs),
3928 - MUST NOT be used on messages generated by automatic or periodic
3929 processes, except for messages which are automatic responses to
3932 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3933 to the default Received: header definition.
3935 PH/12 Added log selector acl_warn_skipped (default on).
3937 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3938 variables because (a) they are in the wrong storage pool and (b) even if
3939 they were copied, it wouldn't work properly because of the caching.
3941 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3942 checking when decoding. Apparently there are clients that generate
3943 overlong encoded strings. Why am I not surprised?
3945 PH/15 If the first argument of "${if match_address" was not empty, but did not
3946 contain an "@" character, Exim crashed. Now it writes a panic log message
3947 and treats the condition as false.
3949 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3951 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3952 extension "envelope-auth". The code is finished and in agreement with
3953 other implementations, but there is no documentation so far and in fact,
3954 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3955 not changing the active code.
3957 Print executed "if" and "elsif" statements when debugging is used. This
3958 helps a great deal to understand what a filter does.
3960 Document more things not specified clearly in RFC3028. I had all this
3961 sorted out, when out of a sudden new issues came to my mind. Oops."
3963 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3966 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3967 canonical form (as documented). However, after a host name lookup from
3968 the IP address, check_host() was doing a simple string comparison with
3969 addresses acquired from the DNS when checking that the found name did
3970 have the original IP as one of its addresses. Since any found IPv6
3971 addresses are likely to be in abbreviated form, the comparison could
3972 fail. Luckily, there already exists a function for doing the comparison
3973 by converting both addresses to binary, so now that is used instead of
3974 the text comparison.
3976 PH/20 There was another similar case to PH/19, when a complete host name was
3977 given in a host list; looking up its IP address could give an abbreviated
3978 form, whereas the current host's name might or might not be abbreviated.
3979 The same fix has been applied.
3985 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3986 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3989 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3990 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3992 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3994 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3995 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3996 Also, for "Auto-Submitted", treat anything other than "no" as
3997 non-personal, in accordance with RFC 3834. (Previously it treated
3998 anything starting "auto-" as non-personal.)
4000 TF/01 The control=submission/name=... option had a problem with syntax
4001 errors if the name included a slash character. The /name= option
4002 now slurps the rest of the string, so it can include any characters
4003 but it must come last in the list of options (after /sender_retain
4006 PH/05 Some modifications to the interface to the fake nameserver for the new
4014 TK/01 Added the "success_on_redirect" address verification option. See
4015 NewStuff for rationale and an example.
4017 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4019 PH/02 Patch to exigrep to allow it to work on syslog lines.
4021 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4022 fread() to skip over the body file's header line, because in Cygwin the
4023 header line is locked and is inaccessible.
4025 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4026 co-exist for some time) to make it clear that it is the Exim ID that is
4027 referenced, not the Message-ID: header line.
4029 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4030 string_format() function, because snprintf() does not exist on all
4033 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4034 ${stat: expansion item.
4036 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4037 protocol synchronization error", to keep the pedants happy.
4039 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4040 well as for IRIX systems, when gcc is being used. See the host.c source
4043 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4045 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4048 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4049 transport and to the smtp transport in LMTP mode.
4051 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4053 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4054 run a filter in a subprocess. This could lead to confusion in subsequent
4055 lookups in the parent process. There should also be a search_tidyup() at
4056 the end of the subprocess.
4058 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4059 only if the host matched helo_try_verify_hosts, which caused the
4060 verification to occur when the EHLO/HELO command was issued. The ACL just
4061 tested the remembered result. Now, if a previous verification attempt has
4062 not happened, "verify = helo" does it there and then.
4064 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4066 TK/03 Fix log output including CR from clamd.
4068 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4069 exist provoked a memory error which could cause a segfault.
4071 PH/15 Installed PCRE 6.2
4073 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4075 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4076 of the problem. Specifically, suggested +O2 rather than +O1 for the
4079 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4081 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4082 record, the port was not correctly logged when the outgoing_port log
4083 selector was set (it logged the transort's default port).
4085 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4086 fallback_hosts, and "hosts" in the smtp transport.
4088 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4089 host errors such as "Connection refused".
4091 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4092 authentication with radiusclient 0.4.9:
4094 - Error returned from rc_read_config was caught wrongly
4095 - Username/password not passed on to radius server due to wrong length.
4097 The presumption is that some radiusclient API changes for 4.51/PH/17
4098 were not taken care of correctly. The code is still untested by me (my
4099 Linux distribution still has 0.3.2 of radiusclient), but it was
4100 contributed by a Radius user.
4102 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4103 expanding the "port" option of the smtp transport.
4105 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4106 while reading a MIME header. Thanks to Tom Hughes for a patch.
4108 PH/24 Include config.h inside local_scan.h so that configuration settings are
4111 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4112 that in an ACL for RCPT (for example), you can examine exactly what was
4115 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4116 commands, but it was not correctly comparing the address with the actual
4117 client host address. Thus, it would show the EHLO address in Received:
4118 header lines when this was not necessary.
4120 PH/27 Added the % operator to ${eval:}.
4122 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4123 it should be ignoring failures (because with -C, for example, it has lost
4124 privilege). It wasn't ignoring creation failures other than "already
4127 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4130 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4131 because an input line was too long, either on its own, or by virtue of
4132 too many continuations, the temporary file was not being removed, and the
4133 return code was incorrect.
4135 PH/31 Missing "BOOL" in function definition in filtertest.c.
4137 PH/32 Applied Sieve patches from the maintainer.
4139 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4141 PH/33 Added "verify = not_blind".
4143 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4144 Local/Makefile (with some defaults set). These are used in built scripts
4145 such as exicyclog, but they have never been used in the exim_install
4146 script (though there are many overriding facilities there). I have
4147 arranged that the exim_install script now takes note of these two
4150 PH/35 Installed configuration files for Dragonfly.
4152 PH/36 When a locally submitted message by a trusted user did not contain a
4153 From: header, and the sender address was obtained from -f or from an SMTP
4154 MAIL command, and the trusted user did not use -F to supply a sender
4155 name, $originator_name was incorrectly used when constructing a From:
4156 header. Furthermore, $originator_name was used for submission mode
4157 messages from external hosts without From: headers in a similar way,
4158 which is clearly wrong.
4160 PH/37 Added control=suppress_local_fixups.
4162 PH/38 When log_selector = +received_sender was set, and the addition of the
4163 sender made the log line's construction buffer exactly full, or one byte
4164 less than full, an overflow happened when the terminating "\n" was
4167 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4168 when the result of a list match is failure because a DNS lookup failed.
4170 PH/40 RM_COMMAND is now used in the building process.
4172 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4173 the "build-* directories that it finds.
4175 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4176 address in a domain literal was a prefix of an interface address.
4178 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4179 when verifying a sender address, unless rewrite_headers is false.
4181 PH/44 Wrote a long comment about why errors_to addresses are verified as
4182 recipients, not senders.
4184 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4185 the ratelimit ACL was added.
4187 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4189 PH/46 Added extra information about PostgreSQL errors to the error string.
4191 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4192 avoiding the need to install special zones in a real server. This is
4193 backwards compatible; if it can't find the fake resolver, it drops back.
4194 Thus, both old and new test suites can be run.
4196 TF/02 Added util/ratelimit.pl
4198 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4199 clock is set back in time.
4201 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4202 Candler <B.Candler@pobox.com>.
4204 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4205 OK for addresses that are the result of redirections.
4207 PH/48 A number of further additions for the benefit of the new test suite,
4208 including a fake gethostbyname() that interfaces to the fake DNS resolver
4211 TF/06 The fix for widen_domains has also been applied to qualify_single and
4212 search_parents which are the other dnslookup options that can cause
4215 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4218 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4220 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4221 service" if it encountered an MX record with an empty target hostname.
4222 The message is now "an MX or SRV record indicated no SMTP service".
4224 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4225 if the DNS of the sending site is misconfigured. This is quite a
4226 common situation. This change restores the behaviour of treating a
4227 helo verification defer as a failure.
4229 PH/51 If self=fail was set on a router, the bounce message did not include the
4230 actual error message.
4236 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4238 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4239 command itself ended in error, the underlying message about the transport
4240 filter timeout was being overwritten with the pipe command error. Now the
4241 underlying error message should be appended to the second error message.
4243 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4245 PH/02 Reduce the amount of output that "make" produces by default. Full output
4246 can still be requested.
4248 PH/03 The warning log line about a condition test deferring for a "warn" verb
4249 was being output only once per connection, rather than after each
4250 occurrence (because it was using the same function as for successful
4251 "warn" verbs). This seems wrong, so I have changed it.
4253 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4254 it should not have, which might have caused a crash in the right
4255 circumstances, but probably never did.
4257 PH/04 Installed a modified version of Tony Finch's patch to make submission
4258 mode fix the return path as well as the Sender: header line, and to
4259 add a /name= option so that you can make the user's friendly name appear
4262 TF/03 Added the control = fakedefer ACL modifier.
4264 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4265 Mark Lowes for thorough testing.
4267 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4269 TK/03 Merged latest SRS patch from Miles Wilton.
4271 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4272 with the definition in sysexits.h (which is #included earlier).
4273 Fortunately, Exim does not actually use EX_OK. The code used to try to
4274 preserve the sysexits.h value, by assuming that macro definitions were
4275 scanned for macro replacements. I have been disabused of this notion,
4276 so now the code just undefines EX_OK before #including unistd.h.
4278 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4279 in the smtp transport. When a block could not be written in a single
4280 write() function, the timeout was being re-applied to each part-write.
4281 This seems wrong - if the receiver was accepting one byte at a time it
4282 would take for ever. The timeout is now adjusted when this happens. It
4283 doesn't have to be particularly precise.
4285 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4286 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4288 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4289 without a domain if the check to <postmaster@domain> fails.
4291 SC/01 Eximstats: added -xls and the ability to specify output files
4292 (patch written by Frank Heydlauf).
4294 SC/02 Eximstats: use FileHandles for outputting results.
4296 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4298 SC/04 Eximstats: fixed display of large numbers with -nvr option
4300 SC/05 Eximstats: fixed merging of reports with empty tables.
4302 SC/06 Eximstats: added the -include_original_destination flag
4304 SC/07 Eximstats: removed tabs and trailing whitespace.
4306 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4308 TK/06 MBOX spool code: Add real "From " MBOX separator line
4309 so the .eml file is really in mbox format (even though
4310 most programs do not really care). Patch from Alex Miller.
4312 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4313 The latter is generated from $received_to and is only set if the
4314 message has one envelope recipient. SA can use these headers,
4315 obviously out-of-the-box. Patch from Alex Miller.
4317 PH/08 The ${def test on a variable was returning false if the variable's
4318 value was "0", contrary to what the specification has always said!
4319 The result should be true unless the variable is empty.
4321 PH/09 The syntax error of a character other than { following "${if
4322 def:variable_name" (after optional whitespace) was not being diagnosed.
4323 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4324 accidental colon was present, for example, could give incorrect results.
4326 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4327 result is used (not including appendfile, where other changes are about
4330 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4331 This involved changing a lot of size variables from int to off_t. It
4332 should work with maildirs and everything.
4334 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4335 spamd dying while we are connected to it.
4337 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4340 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4341 function for BDB 4.3.
4343 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4345 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4346 the sources for PCRE have been re-arranged and more files are now
4349 PH/15 The code I had for printing potentially long long variables in PH/11
4350 above was not the best (it lost precision). The length of off_t variables
4351 is now inspected at build time, and an appropriate printing format (%ld
4352 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4353 to be "long long int" or "long int". This is needed for the internal
4354 formatting function string_vformat().
4356 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4357 the configuration file to be ":syslog", then the script "guesses" where
4358 the logs files are, rather than using the compiled in default. In our
4359 case the guess is not the same as the compiled default, so the script
4360 suddenly stopped working when I started to use syslog. The patch checks
4361 to see if log_file_path is "". If so, it attempts to read it from exim
4362 with no configuration file to get the compiled in version, before it
4363 falls back to the previous guessing code."
4365 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4366 implementing BATV in an Exim configuration. See NewStuff for the gory
4369 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4370 Makefile that are specific to HP-UX.
4372 PH/18 If the "use_postmaster" option was set for a recipient callout together
4373 with the "random" option, the postmaster address was used as the MAIL
4374 FROM address for the random test, but not for the subsequent recipient
4375 test. It is now used for both.
4377 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4378 patch removes a few documentation additions to RFC 3028, because the
4379 latest draft now contains them. It adds the new en;ascii-case comparator
4380 and a new error check for 8bit text in MIME parts. Comparator and
4381 require names are now matched exactly. I enabled the subaddress
4382 extension, but it is not well tested yet (read: it works for me)."
4384 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4385 rework some of the code of TK/09 above to avoid the hardwired use of
4386 "%lld" and "long long". Replaced the call to snprintf() with a call to
4389 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4390 records point to non-existent hosts", "retry timeout exceeded", and
4391 "retry time not reached for any host after a long failure period".
4393 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4394 experimental DomainKeys support:
4396 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4397 (2) On an error such as an illegally used "control", the wrong name for
4398 the control was given.
4400 These problems did NOT occur unless DomainKeys support was compiled.
4402 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4404 PH/24 Added ${if match_ip condition.
4406 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4407 databases so that it will be absolutely obvious if a crash occurs in the
4408 DB library. This is a regular occurrence (often caused by mis-matched
4411 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4412 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4413 picked up on a user's system that detects such things. There doesn't seem
4414 to be a gcc warning option for this - only an attribute that has to be
4415 put on the function's prototype. It seems that in Fedora Core 4 they have
4416 set this on a number of new functions. No doubt there will be more in due
4419 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4420 specify a transport. However, if an address that was verified by such a
4421 router was the subject of a callout, Exim crashed because it tried to
4422 read the rcpt_include_affixes from the non-existent transport. Now it
4423 just assumes that the setting of that option is false. This bug was
4424 introduced by 4.51/PH/31.
4426 PH/28 Changed -d+all to exclude +memory, because that information is very
4427 rarely of interest, but it makes the output a lot bigger. People tend to
4428 do -d+all out of habit.
4430 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4431 code in os-type was giving problems when libc.so lives in lib64, like on
4434 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4435 aren't the modern standard, and it seems that some systems' include files
4436 don't always have them. Exim was already checking for some of the newer
4437 ones like T_AAAA, and defining it itself. I've added checks for all the
4438 record types that Exim uses.
4440 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4441 not automatically generating a new one, as it is supposed to. This
4442 prevented TLS from working. If the file did exist, but contained invalid
4443 data, a new version was generated, as expected. It was only the case of a
4444 non-existent file that was broken.
4446 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4447 with a change in libdomainkeys > 0.64.
4449 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4450 from DNS. If the selector record carries the flag, it now has
4451 precedence over the domain-wide flag.
4453 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4455 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4456 the use of a port name that isn't defined in /etc/services) occurred, the
4457 message was deferred as in a normal delivery, and thus remained on the
4458 spool, instead of being failed because of the mua_wrapper setting. This
4459 is now fixed, and I tidied up some of the mua_wrapper messages at the
4462 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4463 the messages in a hash of arrays rather than using individual hashes.
4464 This is a bit cleaner and results in dramatic memory savings, albeit
4465 at a slight CPU cost.
4467 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4468 as requested by Marc Sherman.
4470 SC/10 Eximstats: added histograms for user specified patterns as requested
4473 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4475 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4476 fopen() in the content-scanning modules that did not already have it.
4482 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4483 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4485 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4487 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4489 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4490 to test Sieve filters that use "vacation".
4492 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4493 that changes the way the GnuTLS parameters are stored in the cache file.
4494 The new format can be generated externally. For backward compatibility,
4495 if the data in the cache doesn't make sense, Exim assumes it has read an
4496 old-format file, and it generates new data and writes a new file. This
4497 means that you can't go back to an older release without removing the
4500 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4501 work if there are any delivery delays because "one_time" forces the
4502 parent to be marked "delivered", so its unseen clone is never tried
4503 again. For this reason, Exim now forbids the simultaneous setting of
4506 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4507 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4508 entry if you want to know the details. The fix, however, neglected to
4509 consider the case when local delivery batching is involved. The test for
4510 "previously delivered" was not happening when checking to see if an
4511 address could be batched with a previous (undelivered) one; under
4512 certain circumstances this could lead to multiple deliveries to the same
4515 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4516 in its include files, and this causes problems building Exim.
4518 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4519 header_syntax) but Exim was just ignoring anything given after a slash.
4520 In particular, this caused confusion with an attempt to use "verify =
4521 reverse_host_lookup/defer_ok". An error is now given when options are
4522 supplied for verify items that do not have them. (Maybe reverse_host_
4523 lookup should have a defer_ok option, but that's a different point.)
4525 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4526 defined by RFC 821) to 2048, because there were problems with some AUTH
4527 commands, and RFC 1869 says the size should be increased for extended
4528 SMTP commands that take arguments.
4530 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4533 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4534 "unknown" error; now it says that the functionality isn't in the binary.
4536 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4537 an address' error message when a string expansion fails (syntax or
4538 whatever). Otherwise the password may appear in the log. Following change
4539 PH/42 below, there is no longer a chance of it appearing in a bounce
4542 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4544 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4545 of its addresses was ever tried. (Bugzilla bug #2.)
4547 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4548 the result incorrectly in the debug output. (It correctly added a newline
4549 to what was transported.)
4551 TF/01 Added $received_time.
4553 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4554 commented out examples of how to interface to a virus scanner and to
4555 SpamAssassin. Also added commented examples of av_scanner and
4556 spamd_address settings.
4558 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4559 and controls are allowed in which ACLs. There were a couple of minor
4560 errors. Some of the entries in the conditions table (which is a table of
4561 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4562 negation of where the condition IS allowed.
4564 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4566 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4567 header file does not have a version number, so I've had to invent a new
4568 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4569 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4570 radiusclient), but it was contributed by a Radius user.
4572 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4573 files or directories, for OpenSSL.
4575 PH/19 When an Exim process that is running as root has to create an Exim log
4576 file, it does so in a subprocess that runs as exim:exim so as to get the
4577 ownership right at creation (otherwise, other Exim processes might see
4578 the file with the wrong ownership). There was no test for failure of this
4579 fork() call, which would lead to the process getting stuck as it waited
4580 for a non-existent subprocess. Forks do occasionally fail when resources
4581 run out. I reviewed all the other calls to fork(); they all seem to check
4584 PH/20 When checking for unexpected SMTP input at connect time (before writing
4585 the banner), Exim was not dealing correctly with a non-positive return
4586 from the read() function. If the client had disconnected by this time,
4587 the result was a log entry for a synchronization error with an empty
4588 string after "input=" when read() returned zero. If read() returned -1
4589 (an event I could not check), uninitialized data bytes were printed.
4590 There were reports of junk text (parts of files, etc) appearing after
4593 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4595 PH/22 Added support for macro redefinition, and (re)definition in between
4596 driver and ACL definitions.
4598 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4599 forgetting to use the resulting value; it was using the unexpanded value.
4601 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4602 hadn't been configured. The fix is from Juergen Kreileder, who
4603 understands it better than I do:
4605 "Here's what I see happening with three configured cyrus_sasl
4606 authenticators configured (plain, login, cram-md5):
4608 On startup auth_cyrus_sasl_init() gets called for each of these.
4609 This means three calls to sasl_listmech() without a specified mech_list.
4610 => SASL tests which mechs of all available mechs actually work
4611 => three warnings about OTP not working
4612 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4614 With the patch, sasl_listmech() also gets called three times. But now
4615 SASL's mech_list option is set to the server_mech specified in the the
4616 authenticator. Or in other words, the answer from sasl_listmech()
4617 gets limited to just the mech you're testing for (which is different
4619 => the return list contains just 'plain' or 'login', 'cram-md5' or
4620 nothing depending on the value of ob->server_mech.
4622 I've just tested the patch: Authentication still works fine,
4623 unavailable mechs specified in the exim configuration are still
4624 caught, and the auth.log warnings about OTP are gone."
4626 PH/25 When debugging is enabled, the contents of the command line are added
4627 to the debugging output, even when log_selector=+arguments is not
4630 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4631 answer is "GNU", and only if the return is "GNU/something" is the answer
4634 PH/27 $acl_verify_message is now set immediately after the failure of a
4635 verification in an ACL, and so is available in subsequent modifiers. In
4636 particular, the message can be preserved by coding like this:
4638 warn !verify = sender
4639 set acl_m0 = $acl_verify_message
4641 Previously, $acl_verify_message was set only while expanding "message"
4642 and "log_message" when a very denied access.
4644 PH/28 Modified OS/os.c-Linux with
4646 -#ifndef OS_LOAD_AVERAGE
4647 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4649 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4650 nomenclature these days.)
4652 PH/29 Installed patch from the Sieve maintainer that adds the options
4653 sieve_useraddress and sieve_subaddress to the redirect router.
4655 PH/30 In these circumstances:
4656 . Two addresses routed to the same list of hosts;
4657 . First host does not offer TLS;
4658 . First host accepts first address;
4659 . First host gives temporary error to second address;
4660 . Second host offers TLS and a TLS session is established;
4661 . Second host accepts second address.
4662 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4663 and peerdn, if requested) that were in fact used only for the second
4666 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4667 attention to any local part prefix or suffix that was matched by the
4668 router that accepted the address. It now behaves in the same way as it
4669 does for delivery: the affixes are removed from the local part unless
4670 rcpt_include_affixes is set on the transport.
4672 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4673 timeout during the DATA phase of an incoming message.
4675 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4676 applied a patch sanctioned by the Sieve maintainer.
4678 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4679 the uid or gid is negative. A case of a negative gid caused this to be
4680 noticed. The fix allows for either to be negative.
4682 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4683 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4686 PH/36 The change PH/12 above was broken. Fixed it.
4688 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4689 the grounds that routing the same address twice would always produce the
4690 same answer. This might have been true once, but it is certainly no
4691 longer true now. Routing a child address may depend on the previous
4692 routing that produced that child. Some complicated redirection strategies
4693 went wrong when messages had multiple recipients, and made Exim's
4694 behaviour dependent on the order in which the addresses were given.
4696 I have moved the duplicate checking until after the routing is complete.
4697 Exim scans the addresses that are assigned to local and remote
4698 transports, and removes any duplicates. This means that more work will be
4699 done, as duplicates will always all be routed, but duplicates are
4700 presumably rare, so I don't expect this is of any significance.
4702 For deliveries to pipes, files, and autoreplies, the duplicate checking
4703 still happens during the routing process, since they are not going to be
4706 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4707 It corrects a timeout issue with spamd. This is Ian's comment: "The
4708 background is that sometimes spamd either never reads data from a
4709 connection it has accepted, or it never writes response data. The exiscan
4710 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4711 blindly assumes that writes won't block so it may never time out."
4713 PH/39 Allow G after quota size as well as K and M.
4715 PH/40 The value set for $authenticated_id in an authenticator may not contain
4716 binary zeroes or newlines because the value is written to log lines and
4717 to spool files. There was no check on this. Now the value is run through
4718 the string_printing() function so that such characters are converted to
4719 printable escape sequences.
4721 PH/41 $message_linecount is a new variable that contains the total number of
4722 lines in the message. Compare $body_linecount, which is the count for the
4725 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4726 bounce and delay warning messages, except in certain special cases, which
4729 (a) An SMTP error message from a remote host;
4730 (b) A message specified in a :fail: redirection;
4731 (c) A message specified in a "fail" command in a system filter;
4732 (d) A message specified in a FAIL return from the queryprogram router;
4733 (e) A message specified by the cannot_route_message router option.
4735 In these cases only, Exim does include the error details in bounce and
4736 warning messages. There are also a few cases where bland messages such
4737 as "unrouteable address" or "local delivery error" are given.
4739 PH/43 $value is now also set for the "else" part of a ${run expansion.
4741 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4742 being worked on, but at least Exim now implements the latest version to
4745 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4746 process to complete was treated as a delivery failure, a timeout while
4747 writing the message to the pipe was logged, but erroneously treated as a
4748 successful delivery. Such timeouts include transport filter timeouts. For
4749 consistency with the overall process timeout, these timeouts are now
4750 treated as errors, giving rise to delivery failures by default. However,
4751 there is now a new Boolean option for the pipe transport called
4752 timeout_defer, which, if set TRUE, converts the failures into defers for
4753 both kinds of timeout. A transport filter timeout is now identified in
4756 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4757 systems where "make" and "gmake" are different, calling "gmake" at top
4758 level broke things. I've arranged for the value of $(MAKE) to be passed
4759 from the Makefile to this script so that it can call the same version of
4763 A note about Exim versions 4.44 and 4.50
4764 ----------------------------------------
4766 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4767 changes of various kinds. As a consequence, a big documentation update was
4768 needed. This delayed the release for rather longer than seemed good, especially
4769 in the light of a couple of (minor) security issues. Therefore, the changes
4770 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4771 release. So 4.44 and 4.50 are in effect two different branches that both start
4774 I have left the 4.50 change log unchanged; it contains all the changes since
4775 4.43. The change log for 4.44 is below; many of its items are identical to
4776 those for 4.50. This seems to be the most sensible way to preserve the
4777 historical information.
4783 1. Minor wording change to the doc/README.SIEVE file.
4785 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4786 computation of the current number of files was incorrect.
4788 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4789 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4792 4. Give more explanation in the error message when the command for a transport
4793 filter fails to execute.
4795 5. There are several places where Exim runs a non-Exim command in a
4796 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4797 was being done only for the command run by the queryprogram router. It is
4798 now done for all such subprocesses. The other cases are: ${run, transport
4799 filters, and the commands run by the lmtp and pipe transports.
4801 6. Added CONFIGURE_GROUP build-time option.
4803 7. Some older OS have a limit of 256 on the maximum number of file
4804 descriptors. Exim was using setrlimit() to set 1000 as a large value
4805 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4806 systems. I've change it so that if it can't get 1000, it tries for 256.
4808 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4809 was an oversight, and furthermore, ever since the addition of extra
4810 controls (e.g. 4.43/32), the checks on when to allow different forms of
4811 "control" were broken. There should now be diagnostics for all cases when a
4812 control that does not make sense is encountered.
4814 9. Added the /retain_sender option to "control=submission".
4816 10. $recipients is now available in the predata ACL (oversight).
4818 11. Tidy the search cache before the fork to do a delivery from a message
4819 received from the command line. Otherwise the child will trigger a lookup
4820 failure and thereby defer the delivery if it tries to use (for example) a
4821 cached ldap connection that the parent has called unbind on.
4823 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4824 of $address_data from the recipient verification was clobbered by the
4825 sender verification.
4827 13. The value of address_data from a sender verification is now available in
4828 $sender_address_data in subsequent conditions in the ACL statement.
4830 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4832 15. Added a new option "connect=<time>" to callout options, to set a different
4835 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4836 was its contents. (It was OK if the option was not defined at all.)
4838 17. A "Completed" log line is now written for messages that are removed from
4839 the spool by the -Mrm option.
4841 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4842 information about exactly what failed.
4844 19. Added -dd to debug only the daemon process.
4846 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4847 handles timeouts, both on the server side and network timeouts. Renamed the
4848 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4850 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4851 It is now set to "smtps".
4853 22. $host_address is now set to the target address during the checking of
4854 ignore_target_hosts.
4856 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4857 being passed; this would have caused $sender_host_name to have been used if
4858 matching the list had actually called for a host name (not very likely,
4859 since this list is usually IP addresses). A host name is now passed as
4862 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4863 code that specifies a non-restarting handler (typically sigaction() in
4864 modern systems) in an attempt to fix a rare and obscure crash bug.
4866 25. Narrowed the window for a race in the daemon that could cause it to ignore
4867 SIGCHLD signals. This is not a major problem, because they are used only to
4868 wake it up if nothing else does.
4870 26. A malformed maildirsize file could cause Exim to calculate negative values
4871 for the mailbox size or file count. Odd effects could occur as a result.
4872 The maildirsize information is now recalculated if the size or filecount
4875 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4876 support for a long time. Removed HAVE_SYS_VFS_H.
4878 28. Installed the latest version of exipick from John Jetmore.
4880 29. In an address list, if the pattern was not a regular expression, an empty
4881 subject address (from a bounce message) matched only if the pattern was an
4882 empty string. Non-empty patterns were not even tested. This was the wrong
4883 because it is perfectly reasonable to use an empty address as part of a
4884 database query. An empty address is now tested by patterns that are
4885 lookups. However, all the other forms of pattern expect the subject to
4886 contain a local part and a domain, and therefore, for them, an empty
4887 address still always fails if the pattern is not itself empty.
4889 30. Exim went into a mad DNS loop when attempting to do a callout where the
4890 host was specified on an smtp transport, and looking it up yielded more
4891 than one IP address.
4893 31. Re-factored the code for checking spool and log partition space into a
4894 function that finds that data and another that does the check. The former
4895 is then used to implement four new variables: $spool_space, $log_space,
4896 $spool_inodes, and $log_inodes.
4898 32. The RFC2047 encoding function was originally intended for short strings
4899 such as real names; it was not keeping to the 75-character limit for
4900 encoded words that the RFC imposes. It now respects the limit, and
4901 generates multiple encoded words if necessary. To be on the safe side, I
4902 have increased the buffer size for the ${rfc2047: expansion operator from
4905 33. It is now permitted to omit both strings after an "if" condition; if the
4906 condition is true, the result is "true". As before, when the second string
4907 is omitted, a false condition yields an empty string. This makes it less
4908 cumbersome to write custom ACL and router conditions.
4910 34. Failure to deliver a bounce message always caused it to be frozen, even if
4911 there was an errors_to setting on the router. The errors_to setting is now
4914 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4915 canonical form (fully expanded) before being placed in
4916 $sender_host_address.
4918 36. The table in the code that translates DNS record types into text (T_A to
4919 "A" for instance) was missing entries for NS and CNAME. It is just possible
4920 that this could have caused confusion if both these types were looked up
4921 for the same domain, because the text type is used as part of Exim's
4922 per-process caching. But the chance of anyone hitting this buglet seems
4925 37. The dnsdb lookup has been extended in a number of ways.
4927 (1) There is a new type, "zns", which walks up the domain tree until it
4928 finds some nameserver records. It should be used with care.
4930 (2) There is a new type, "mxh", which is like "mx" except that it returns
4931 just the host names, not the priorities.
4933 (3) It is now possible to give a list of domains (or IP addresses) to be
4934 looked up. The behaviour when one of the lookups defers can be
4935 controlled by a keyword.
4937 (4) It is now possible to specify the separator character for use when
4938 multiple records are returned.
4940 38. The dnslists ACL condition has been extended: it is now possible to supply
4941 a list of IP addresses and/or domains to be looked up in a particular DNS
4944 39. Added log_selector=+queue_time_overall.
4946 40. When running the queue in the test harness, wait just a tad after forking a
4947 delivery process, to get repeatability of debugging output.
4949 41. Include certificate and key file names in error message when GnuTLS fails
4950 to set them up, because the GnuTLS error message doesn't include the name
4951 of the failing file when there is a problem reading it.
4953 42. Allow both -bf and -bF in the same test run.
4955 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4957 44. The "Exiscan patch" is now merged into the mainline Exim source.
4959 45. Sometimes the final signoff response after QUIT could fail to get
4960 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4961 < 0 before doing a fflush(). This bug looks as though it goes back to the
4962 introduction of TLS in release 3.20, but "sometimes" must have been rare
4963 because the tests only now provoked it.
4965 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4966 (this can affect the format of dates).
4968 47. exim_tidydb, when checking for the continued existence of a message for
4969 which it has found a message-specific retry record, was not finding
4970 messages that were in split spool directories. Consequently, it was
4971 deleting retry records that should have stayed in existence.
4973 48. Steve fixed some bugs in eximstats.
4975 49. The SPA authentication driver was not abandoning authentication and moving
4976 on to the next authenticator when an expansion was forced to fail,
4977 contradicting the general specification for all authenticators. Instead it
4978 was generating a temporary error. It now behaves as specified.
4980 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4981 (the order specifies the preference for clients). The order is now AES256,
4982 AES128, 3DES, ARCFOUR128.
4984 51. Small patch to Sieve code - explicitly set From: when generating an
4987 52. Exim crashed if a remote delivery caused a very long error message to be
4988 recorded - for instance if somebody sent an entire SpamAssassin report back
4989 as a large number of 550 error lines. This bug was coincidentally fixed by
4990 increasing the size of one of Exim's internal buffers (big_buffer) that
4991 happened as part of the Exiscan merge. However, to be on the safe side, I
4992 have made the code more robust (and fixed the comments that describe what
4995 53. Now that there can be additional text after "Completed" in log lines (if
4996 the queue_time_overall log selector is set), a one-byte patch to exigrep
4997 was needed to allow it to recognize "Completed" as not the last thing in
5000 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5001 patch that reportedly fixes this has been added. I am not expert enough to
5002 create a test for it. This is what the patch creator wrote:
5004 "I found a little strange behaviour of ldap code when working with
5005 Windows 2003 AD Domain, where users was placed in more than one
5006 Organization Units. When I tried to give exim partial DN, the exit code
5007 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5008 But simultaneously result of request was absolutely normal ldap result,
5009 so I produce this patch..."
5011 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5012 REFERENCE, so I have modified the code to exclude the patch when that macro
5015 55. Some experimental protocols are using DNS PTR records for new purposes. The
5016 keys for these records are domain names, not reversed IP addresses. The
5017 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5018 leaves it alone. Component reversal etc. now happens only for IP addresses.
5021 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5023 57. Double the size of the debug message buffer (to 2048) so that more of very
5024 long debug lines gets shown.
5026 58. The exicyclog utility now does better if the number of log files to keep
5027 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5029 59. Two changes related to the smtp_active_hostname option:
5031 (1) $smtp_active_hostname is now available as a variable.
5032 (2) The default for smtp_banner uses $smtp_active_hostname instead
5033 of $primary_hostname.
5035 60. The host_aton() function is supposed to be passed a string that is known
5036 to be a valid IP address. However, in the case of IPv6 addresses, it was
5037 not checking this. This is a hostage to fortune. Exim now panics and dies
5038 if the condition is not met. A case was found where this could be provoked
5039 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5040 components; fortuitously, this particular loophole had already been fixed
5041 by change 4.50/55 above.
5043 If there are any other similar loopholes, the new check in host_aton()
5044 itself should stop them being exploited. The report I received stated that
5045 data on the command line could provoke the exploit when Exim was running as
5046 exim, but did not say which command line option was involved. All I could
5047 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5048 running as the user.
5051 61. There was a buffer overflow vulnerability in the SPA authentication code
5052 (which came originally from the Samba project). I have added a test to the
5053 spa_base64_to_bits() function which I hope fixes it.
5056 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5057 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5059 63. The daemon start-up calls getloadavg() while still root for those OS that
5060 need the first call to be done as root, but it missed one case: when
5061 deliver_queue_load_max is set with deliver_drop_privilege. This is
5062 necessary for the benefit of the queue runner, because there is no re-exec
5063 when deliver_drop_privilege is set.
5065 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5066 This has been fixed.
5068 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5069 was in use, was not putting the data itself into the right store pool;
5070 consequently, it could be overwritten for a subsequent message in the same
5071 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5074 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5076 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5077 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5078 the function were treating the return as a boolean value, which happened to
5079 work because 0=false and not-0=true, but is not correct code.
5081 68. The host_aton() function was not handling scoped IPv6 addresses (those
5082 with, for example, "%eth0" on the end) correctly.
5084 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5085 negated items (that is, ~something) in unsigned ints. Some compilers
5086 apparently mutter when there is no cast.
5088 70. If an address verification called from an ACL failed, and did not produce a
5089 user-specific message (i.e. there was only a "system" message), nothing was
5090 put in $acl_verify_message. In this situation, it now puts the system
5093 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5094 session; change 4.31/43 added the unwanted input to the log line - except
5095 that it did not do this in the start of session case. It now does.
5097 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5098 This is wrong and can cause the other end to generate a synchronization
5099 error if it is another Exim or anything else that does the synchronization
5100 check. A QUIT command is no longer sent after a timeout.
5102 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5103 during host lookups.
5105 74. The defer_ok option of callout verification was not working if it was used
5106 when verifying addresses in header lines, that is, for this case:
5108 verify = header_sender/callout=defer_ok
5110 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5111 those file descriptors could be used for SMTP connections. If anything
5112 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5113 could be sent to the SMTP client, causing chaos. The daemon now opens
5114 stdin, stdout, and stderr to /dev/null when it puts itself into the
5117 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5118 log by default. The user can override this with suitable Perl magic.
5120 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5121 the log message when discard triggers, was not working for the DATA ACL or
5122 for the non-SMTP ACL.
5124 78. Error message wording change in sieve.c.
5126 79. If smtp_accept_max_per_host was set, the number of connections could be
5127 restricted to fewer than expected, because the daemon was trying to set up
5128 a new connection before checking whether the processes handling previous
5129 connections had finished. The check for completed processes is now done
5130 earlier. On busy systems, this bug wouldn't be noticed because something
5131 else would have woken the daemon, and it would have reaped the completed
5134 80. If a message was submitted locally by a user whose login name contained one
5135 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5136 It caused a spool format error. I have fixed the spool reading code. A
5137 related problem was that the "from" clause in the Received: line became
5138 illegal because of the space(s). It is now covered by ${quote_local_part.
5140 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5143 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5145 83. Patch from Sieve maintainer to fix unterminated string problem in
5146 "vacation" handling.
5148 84. Some minor changes to the Linux configuration files to help with other
5149 OS variants using glibc.
5151 85. One more patch for Sieve to update vacation handling to latest spec.
5154 ----------------------------------------------------
5155 See the note above about the 4.44 and 4.50 releases.
5156 ----------------------------------------------------
5162 1. Change 4.43/35 introduced a bug that caused file counts to be
5163 incorrectly computed when quota_filecount was set in an appendfile
5166 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5167 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5170 3. Give more explanation in the error message when the command for a transport
5171 filter fails to execute.
5173 4. There are several places where Exim runs a non-Exim command in a
5174 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5175 was being done only for the command run by the queryprogram router. It is
5176 now done for all such subprocesses. The other cases are: ${run, transport
5177 filters, and the commands run by the lmtp and pipe transports.
5179 5. Some older OS have a limit of 256 on the maximum number of file
5180 descriptors. Exim was using setrlimit() to set 1000 as a large value
5181 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5182 systems. I've change it so that if it can't get 1000, it tries for 256.
5184 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5185 was an oversight, and furthermore, ever since the addition of extra
5186 controls (e.g. 4.43/32), the checks on when to allow different forms of
5187 "control" were broken. There should now be diagnostics for all cases when a
5188 control that does not make sense is encountered.
5190 7. $recipients is now available in the predata ACL (oversight).
5192 8. Tidy the search cache before the fork to do a delivery from a message
5193 received from the command line. Otherwise the child will trigger a lookup
5194 failure and thereby defer the delivery if it tries to use (for example) a
5195 cached ldap connection that the parent has called unbind on.
5197 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5198 of $address_data from the recipient verification was clobbered by the
5199 sender verification.
5201 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5202 was its contents. (It was OK if the option was not defined at all.)
5204 11. A "Completed" log line is now written for messages that are removed from
5205 the spool by the -Mrm option.
5207 12. $host_address is now set to the target address during the checking of
5208 ignore_target_hosts.
5210 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5211 being passed; this would have caused $sender_host_name to have been used if
5212 matching the list had actually called for a host name (not very likely,
5213 since this list is usually IP addresses). A host name is now passed as
5216 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5217 code that specifies a non-restarting handler (typically sigaction() in
5218 modern systems) in an attempt to fix a rare and obscure crash bug.
5220 15. Narrowed the window for a race in the daemon that could cause it to ignore
5221 SIGCHLD signals. This is not a major problem, because they are used only to
5222 wake it up if nothing else does.
5224 16. A malformed maildirsize file could cause Exim to calculate negative values
5225 for the mailbox size or file count. Odd effects could occur as a result.
5226 The maildirsize information is now recalculated if the size or filecount
5229 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5230 support for a long time. Removed HAVE_SYS_VFS_H.
5232 18. Updated exipick to current release from John Jetmore.
5234 19. Allow an empty sender to be matched against a lookup in an address list.
5235 Previously the only cases considered were a regular expression, or an
5238 20. Exim went into a mad DNS lookup loop when doing a callout where the
5239 host was specified on the transport, if the DNS lookup yielded more than
5242 21. The RFC2047 encoding function was originally intended for short strings
5243 such as real names; it was not keeping to the 75-character limit for
5244 encoded words that the RFC imposes. It now respects the limit, and
5245 generates multiple encoded words if necessary. To be on the safe side, I
5246 have increased the buffer size for the ${rfc2047: expansion operator from
5249 22. Failure to deliver a bounce message always caused it to be frozen, even if
5250 there was an errors_to setting on the router. The errors_to setting is now
5253 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5254 canonical form (fully expanded) before being placed in
5255 $sender_host_address.
5257 24. Updated eximstats to version 1.33
5259 25. Include certificate and key file names in error message when GnuTLS fails
5260 to set them up, because the GnuTLS error message doesn't include the name
5261 of the failing file when there is a problem reading it.
5263 26. Expand error message when OpenSSL has problems setting up cert/key files.
5266 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5267 (this can affect the format of dates).
5269 28. exim_tidydb, when checking for the continued existence of a message for
5270 which it has found a message-specific retry record, was not finding
5271 messages that were in split spool directories. Consequently, it was
5272 deleting retry records that should have stayed in existence.
5274 29. eximstats updated to version 1.35
5275 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5276 1.35 - bugfix such that pie charts by volume are generated correctly
5278 30. The SPA authentication driver was not abandoning authentication and moving
5279 on to the next authenticator when an expansion was forced to fail,
5280 contradicting the general specification for all authenticators. Instead it
5281 was generating a temporary error. It now behaves as specified.
5283 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5284 (the order specifies the preference for clients). The order is now AES256,
5285 AES128, 3DES, ARCFOUR128.
5287 31. Small patch to Sieve code - explicitly set From: when generating an
5290 32. Exim crashed if a remote delivery caused a very long error message to be
5291 recorded - for instance if somebody sent an entire SpamAssassin report back
5292 as a large number of 550 error lines. This bug was coincidentally fixed by
5293 increasing the size of one of Exim's internal buffers (big_buffer) that
5294 happened as part of the Exiscan merge. However, to be on the safe side, I
5295 have made the code more robust (and fixed the comments that describe what
5298 33. Some experimental protocols are using DNS PTR records for new purposes. The
5299 keys for these records are domain names, not reversed IP addresses. The
5300 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5301 leaves it alone. Component reversal etc. now happens only for IP addresses.
5304 34. The host_aton() function is supposed to be passed a string that is known
5305 to be a valid IP address. However, in the case of IPv6 addresses, it was
5306 not checking this. This is a hostage to fortune. Exim now panics and dies
5307 if the condition is not met. A case was found where this could be provoked
5308 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5309 components; fortuitously, this particular loophole had already been fixed
5310 by change 4.50/55 or 4.44/33 above.
5312 If there are any other similar loopholes, the new check in host_aton()
5313 itself should stop them being exploited. The report I received stated that
5314 data on the command line could provoke the exploit when Exim was running as
5315 exim, but did not say which command line option was involved. All I could
5316 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5317 running as the user.
5320 35. There was a buffer overflow vulnerability in the SPA authentication code
5321 (which came originally from the Samba project). I have added a test to the
5322 spa_base64_to_bits() function which I hope fixes it.
5325 36. The daemon start-up calls getloadavg() while still root for those OS that
5326 need the first call to be done as root, but it missed one case: when
5327 deliver_queue_load_max is set with deliver_drop_privilege. This is
5328 necessary for the benefit of the queue runner, because there is no re-exec
5329 when deliver_drop_privilege is set.
5331 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5332 was in use, was not putting the data itself into the right store pool;
5333 consequently, it could be overwritten for a subsequent message in the same
5334 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5337 38. Sometimes the final signoff response after QUIT could fail to get
5338 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5339 < 0 before doing a fflush(). This bug looks as though it goes back to the
5340 introduction of TLS in release 3.20, but "sometimes" must have been rare
5341 because the tests only now provoked it.
5347 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5348 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5349 Now it returns an integer. A number of calls were still expecting a T/F
5350 return. Fortuitously, in all cases, the tests worked in OK situations,
5351 which is the norm. However, things would have gone wrong on any write
5352 failures on the smtp file descriptor. This function is used when sending
5353 messages over SMTP and also when doing verify callouts.
5355 2. When Exim is called to do synchronous delivery of a locally submitted
5356 message (the -odf or -odi options), it no longer closes stderr before doing
5359 3. Implemented the mua_wrapper option.
5361 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5363 5. Implemented the functions header_remove(), header_testname(),
5364 header_add_at_position(), and receive_remove_recipient(), and exported them
5367 6. If an ACL "warn" statement specified the addition of headers, Exim already
5368 inserted X-ACL-Warn: at the start if there was no header name. However, it
5369 was not making this test for the second and subsequent header lines if
5370 there were newlines in the string. This meant that an invalid header could
5371 be inserted if Exim was badly configured.
5373 7. Allow an ACL "warn" statement to add header lines at the start or after all
5374 the Received: headers, as well as at the end.
5376 8. Added the rcpt_4xx retry error code.
5378 9. Added postmaster_mailfrom=xxx to callout verification option.
5380 10. Added mailfrom=xxxx to the callout verification option, for verify=
5383 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5384 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5386 12. Inserted some casts to stop certain compilers warning when using pointer
5387 differences as field lengths or precisions in printf-type calls (mostly
5388 affecting debugging statements).
5390 13. Added optional readline() support for -be (dynamically loaded).
5392 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5393 same clock tick as a message's arrival, so that its received time was the
5394 same as the "first fail" time on the retry record, and that message
5395 remained on the queue past the ultimate address timeout, every queue runner
5396 would try a delivery (because it was past the ultimate address timeout) but
5397 after another failure, the ultimate address timeout, which should have then
5398 bounced the address, did not kick in. This was a "< instead of <=" error;
5399 in most cases the first failure would have been in the next clock tick
5400 after the received time, and all would be well.
5402 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5403 being recognized when the domain list was tested by the match_domain
5404 condition in an expansion string.
5406 16. Added the ${str2b64: operator.
5408 17. Exim was always calling setrlimit() to set a large limit for the number of
5409 processes, without checking whether the existing limit was already
5410 adequate. (It did check for the limit on file descriptors.) Furthermore,
5411 errors from getrlimit() and setrlimit() were being ignored. Now they are
5412 logged to the main and panic logs, but Exim does carry on, to try to do its
5413 job under whatever limits there are.
5415 18. Imported PCRE 5.0.
5417 19. Trivial typo in log message " temporarily refused connection" (the leading
5420 20. If the log selector return_path_on_delivery was set and an address was
5421 redirected to /dev/null, the delivery process crashed because it assumed
5422 that a return path would always be set for a "successful" delivery. In this
5423 case, the whole delivery is bypassed as an optimization, and therefore no
5426 21. Internal re-arrangement: the function for sending a challenge and reading
5427 a response while authentication was assuming a zero-terminated challenge
5428 string. It's now changed to take a pointer and a length, to allow for
5429 binary data in such strings.
5431 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5433 23. Exim was not respecting finduser_retries when seeking the login of the
5434 uid under which it was called; it was always trying 10 times. (The default
5435 setting of finduser_retries is zero.) Also, it was sleeping after the final
5436 failure, which is pointless.
5438 24. Implemented tls_on_connect_ports.
5440 25. Implemented acl_smtp_predata.
5442 26. If the domain in control=submission is set empty, Exim assumes that the
5443 authenticated id is a complete email address when it generates From: or
5444 Sender: header lines.
5446 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5447 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5448 chown and chgrp in /bin and hostname in /usr/bin.
5450 28. Exim was keeping the "process log" file open after each use, just as it
5451 does for the main log. This opens the possibility of it remaining open for
5452 long periods when the USR1 signal hits a daemon. Occasional processlog
5453 errors were reported, that could have been caused by this. Anyway, it seems
5454 much more sensible not to leave this file open at all, so that is what now
5457 29. The long-running daemon process does not normally write to the log once it
5458 has entered its main loop, and it closes the log before doing so. This is
5459 so that log files can straightforwardly be renamed and moved. However,
5460 there are a couple of unusual error situations where the daemon does write
5461 log entries, and I had neglected to close the log afterwards.
5463 30. The text of an SMTP error response that was received during a remote
5464 delivery was being truncated at 512 bytes. This is too short for some of
5465 the long messages that one sometimes sees. I've increased the limit to
5468 31. It is now possible to make retry rules that apply only when a message has a
5469 specific sender, in particular, an empty sender.
5471 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5472 it possible to be selective about when SMTP synchronization is enforced.
5474 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5476 32. Implemented hosts_connection_nolog.
5478 33. Added an ACL for QUIT.
5480 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5483 35. Added mailbox_size and mailbox_filecount to appendfile.
5485 36. Added control = no_multiline_responses to ACLs.
5487 37. There was a bug in the logic of the code that waits for the clock to tick
5488 in the case where the clock went backwards by a substantial amount such
5489 that the microsecond fraction of "now" was more than the microsecond
5490 fraction of "then" (but the whole seconds number was less).
5492 38. Added support for the libradius Radius client library this is found on
5493 FreeBSD (previously only the radiusclient library was supported).
5499 1. When certain lookups returned multiple values in the form name=value, the
5500 quoting of the values was not always being done properly. Specifically:
5501 (a) If the value started with a double quote, but contained no whitespace,
5503 (b) If the value contained whitespace other than a space character (i.e.
5504 tabs or newlines or carriage returns) it was not quoted.
5505 This fix has been applied to the mysql and pgsql lookups by writing a
5506 separate quoting function and calling it from the lookup code. The fix
5507 should probably also be applied to nisplus, ibase and oracle lookups, but
5508 since I cannot test any of those, I have not disturbed their existing code.
5510 2. A hit in the callout cache for a specific address caused a log line with no
5511 reason for rejecting RCPT. Now it says "Previous (cached) callout
5512 verification failure".
5514 3. There was an off-by-one bug in the queryprogram router. An over-long
5515 return line was truncated at 256 instead of 255 characters, thereby
5516 overflowing its buffer with the terminating zero. As well as fixing this, I
5517 have increased the buffer size to 1024 (and made a note to document this).
5519 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5520 when Exim is waiting for an SMTP response from a remote server, Exim
5521 restarts its select() call on the socket, thereby resetting its timeout.
5522 This is not a problem when such interrupts are rare. Somebody set up a cron
5523 job to run exiwhat every 2 minutes, which is less than the normal select()
5524 timeout (5 or 10 minutes). This meant that the select() timeout never
5525 kicked in because it was always reset. I have fixed this by comparing the
5526 time when an interrupt arrives with the time at the start of the first call
5527 to select(). If more time than the timeout has elapsed, the interrupt is
5528 treated as a timeout.
5530 5. Some internal re-factoring in preparation for the addition of Sieve
5531 extensions (by MH). In particular, the "personal" test is moved to a
5532 separate function, and given an option for scanning Cc: and Bcc: (which is
5533 not set for Exim filters).
5535 6. When Exim created an email address using the login of the caller as the
5536 local part (e.g. when creating a From: or Sender: header line), it was not
5537 quoting the local part when it contained special characters such as @.
5539 7. Installed new OpenBSD configuration files.
5541 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5542 try to make them clearer.
5544 9. Callout options, other than the timeout value, were being ignored when
5545 verifying sender addresses in header lines. For example, when using
5547 verify = header_sender/callout=no_cache
5549 the cache was (incorrectly) being used.
5551 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5552 only the Cygwin environment.
5554 11. When return_path_on_delivery was set as a log selector, if different remote
5555 addresses in the same message used different return paths and parallel
5556 remote delivery occurred, the wrong values would sometimes be logged.
5557 (Whenever a remote delivery process finished, the return path value from
5558 the most recently started remote delivery process was logged.)
5560 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5561 lines when AUTH and/or TLS are in use. This is the "received protocol"
5562 field. Exim used to use "asmtp" for authenticated SMTP, without any
5563 indication (in the protocol name) for TLS use. Now it follows the RFC and
5564 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5565 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5566 names appear in log lines as well as in Received: header lines.
5568 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5569 extensions, and comparison tests, and to fix some bugs.
5571 14. Changes to the "personal" filter test:
5573 (1) The test was buggy in that it was just doing the equivalent of
5574 "contains" tests on header lines. For example, if a user's address was
5575 anne@some.where, the "personal" test would incorrectly be true for
5577 To: susanne@some.where
5579 This test is now done by extracting each address from the header in turn,
5580 and checking the entire address. Other tests that are part of "personal"
5581 are now done using regular expressions (for example, to check local parts
5582 of addresses in From: header lines).
5584 (2) The list of non-personal local parts in From: addresses has been
5585 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5586 taken from the Sieve specification recommendations.
5588 (3) If the message contains any header line starting with "List-" it is
5589 treated as non-personal.
5591 (4) The test for "circular" in the Subject: header line has been removed
5592 because it now seems ill-conceived.
5594 15. Minor typos in src/EDITME comments corrected.
5596 16. Installed latest exipick from John Jetmore.
5598 17. If headers_add on a router specified a text string that was too long for
5599 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5600 of string_sprintf() is now avoided.
5602 18. $message_body_size was not set (it was always zero) when running the DATA
5603 ACL and the local_scan() function.
5605 19. For the "mail" command in an Exim filter, no default was being set for
5606 the once_repeat time, causing a random time value to be used if "once" was
5607 specified. (If the value happened to be <= 0, no repeat happened.) The
5608 default is now 0s, meaning "never repeat". The "vacation" command was OK
5609 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5610 (I found it when inspecting the code).
5612 20. There is now an overall timeout for performing a callout verification. It
5613 defaults to 4 times the callout timeout, which applies to individual SMTP
5614 commands during the callout. The overall timeout applies when there is more
5615 than one host that can be tried. The timeout is checked before trying the
5616 next host. This prevents very long delays if there are a large number of
5617 hosts and all are timing out (e.g. when the network connections are timing
5618 out). The value of the overall timeout can be changed by specifying an
5619 additional sub-option for "callout", called "maxwait". For example:
5621 verify = sender/callout=5s,maxwait=20s
5623 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5624 to the end before writing, but this should make it even safer).
5626 22. Exim was forgetting that it had advertised PIPELINING for the second and
5627 subsequent messages on an SMTP connection. It was also not resetting its
5628 memory on STARTTLS and an internal HELO.
5630 23. When Exim logs an SMTP synchronization error within a session, it now
5631 records whether PIPELINING has been advertised or not.
5633 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5634 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5635 rather than long int.
5637 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5639 26. Added the never_mail option to autoreply.
5645 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5646 crash if the getsockname() call failed; this can happen if a connection is
5647 closed very soon after it is established. The problem was simply in the
5648 order in which certain operations were done, causing Exim to try to write
5649 to the SMTP stream before it had set up the file descriptor. The bug has
5650 been fixed by making things happen in the correct order.
5656 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5657 before the connection was closed, thus losing the rejection response.
5659 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5660 some early Solaris releases, but causes trouble in current releases where
5661 socklen_t is defined.
5663 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5666 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5669 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5670 that releasing the top of it at the end releases what was used for sub-
5671 expansions (unless the block got too big). However, discard this block if
5672 the first thing is a variable or header, so that we can use its block when
5673 it is dynamic (useful for very large $message_headers, for example).
5675 6. Lookups now cache *every* query, not just the most recent. A new, separate
5676 store pool is used for this. It can be recovered when all lookup caches are
5677 flushed. Lookups now release memory at the end of their result strings.
5678 This has involved some general refactoring of the lookup sources.
5680 7. Some code has been added to the store_xxx() functions to reduce the amount
5681 of flapping under certain conditions.
5683 8. log_incoming_interface used to affect only the <= reception log lines. Now
5684 it causes the local interface and port to be added to several more SMTP log
5685 lines, for example "SMTP connection from", and rejection lines.
5687 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5689 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5691 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5692 could be overwritten at the end of the current message (or the start of a
5693 new message if it was set in a HELO ACL). The value is now preserved for
5694 the duration of the SMTP connection.
5696 12. If a transport had a headers_rewrite setting, and a matching header line
5697 contained an unqualified address, that address was qualified, even if it
5698 did not match any rewriting rules. The underlying bug was that the values
5699 of the flags that permit the existence of unqualified sender and recipient
5700 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5701 non-local messages, and by -bnq for local messages) were not being
5702 preserved with the message after it was received.
5704 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5705 "next input=" as part of the text comprising the host identity instead of
5706 the correct text. The code was using the same buffer for two different
5707 strings. However, depending on which order the printing function evaluated
5708 its arguments, the bug did not always show up. Under Linux, for example, my
5709 test suite worked just fine.
5711 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5712 doesn't work with some older versions of Perl. It has been changed to "my",
5713 which in any case is probably the better facility to use.
5715 15. A really picky compiler found some instances of statements for creating
5716 error messages that either had too many or two few arguments for the format
5719 16. The size of the buffer for calls to the DNS resolver has been increased
5720 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5721 for addresses that have a lot of PTR records. This alleviates a problem; it
5722 does not fully solve it.
5724 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5725 buffer now truncates the list and logs the incident, which is the same
5726 action as happens when Exim is looking up a host name and its aliases.
5727 Previously in this situation something unpredictable would happen;
5728 sometimes it was "internal error: store_reset failed".
5730 18. If a server dropped the connection unexpectedly when an Exim client was
5731 using GnuTLS and trying to read a response, the client delivery process
5732 crashed while trying to generate an error log message.
5734 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5735 string, for example:
5737 warn message = H1: something\nH2: something
5739 the text was added as a single header line from Exim's point of view
5740 though it ended up OK in the delivered message. However, searching for the
5741 second and subsequent header lines using $h_h2: did not work. This has been
5742 fixed. Similarly, if a system filter added multiple headers in this way,
5743 the routers could not see them.
5745 20. Expanded the error message when iplsearch is called with an invalid key to
5746 suggest using net-iplsearch in a host list.
5748 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5749 ACLs are no longer actually imposed (and a message to that effect is
5752 22. If a "gecos" field in a passwd entry contained escaped characters, in
5753 particular, if it contained a \" sequence, Exim got it wrong when building
5754 a From: or a Sender: header from that name. A second bug also caused
5755 incorrect handling when an unquoted " was present following a character
5756 that needed quoting.
5758 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5759 was not being matched caselessly.
5761 24. Arranged for all hyphens in the exim.8 source to be escaped with
5764 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5765 properly. Recipient callout cache records were still being keyed to include
5766 the sender, even when use_sender was set false. This led to far more
5767 callouts that were necessary. The sender is no longer included in the key
5768 when use_sender is false.
5770 26. Added "control = submission" modifier to ACLs.
5772 27. Added the ${base62d: operator to decode base 62 numbers.
5774 28. dnsdb lookups can now access SRV records.
5776 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5777 the configuration file.
5779 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5780 (-v) mode. This makes the output for a verbose queue run more intelligible.
5782 31. Added a use_postmaster feature to recipient callouts.
5784 32. Added the $body_zerocount variable, containing the number of binary zero
5785 bytes in the message body.
5787 33. The time of last modification of the "new" subdirectory is now used as the
5788 "mailbox time last read" when there is a quota error for a maildir
5791 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5793 35. Added +ignore_unknown as a special item in host lists.
5795 36. Code for decoding IPv6 addresses in host lists is now included, even if
5796 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5797 address was recognized as an IP address, but was then not correctly decoded
5798 into binary, causing unexpected and incorrect effects when compared with
5805 1. Very minor rewording of debugging text in manualroute to say "list of
5806 hosts" instead of "hostlist".
5808 2. If verify=header_syntax was set, and a header line with an unqualified
5809 address (no domain) and a large number of spaces between the end of the
5810 name and the colon was received, the reception process suffered a buffer
5811 overflow, and (when I tested it) crashed. This was caused by some obsolete
5812 code that should have been removed. The fix is to remove it!
5814 3. When running in the test harness, delay a bit after writing a bounce
5815 message to get a bit more predictability in the log output.
5817 4. Added a call to search_tidyup() just before forking a reception process. In
5818 theory, someone could use a lookup in the expansion of smtp_accept_max_
5819 per_host which, without the tidyup, could leave open a database connection.
5821 5. Added the variables $recipient_data and $sender_data which get set from a
5822 lookup success in an ACL "recipients" or "senders" condition, or a router
5823 "senders" option, similar to $domain_data and $local_part_data.
5825 6. Moved the writing of debug_print from before to after the "senders" test
5828 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5829 problems for message scanning, either using a data ACL, or using
5830 local_scan() because the Received: header was not generated till after they
5831 were called (in order to set the time as the time of reception completion).
5832 I have revised the way this works. The header is now generated after the
5833 body is received, but before the ACL or local_scan() are called. After they
5834 are run, the timestamp in the header is updated.
5840 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5841 before starting a queue runner without re-exec. This happened only when
5842 deliver_drop_privilege was set or when the Exim user was set to root. The
5843 effect of the bug was that timeouts during subsequent deliveries caused
5844 crashes instead of being properly handled. The handler is now left at its
5845 default (and expected) setting.
5847 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5848 message, again when deliver_drop_privilege is set or Exim is run as root.
5849 The bug described in (1) was not present in this case, but the tidying up
5850 of the other signals was missing. I have made the two cases consistent.
5852 3. The ignore_target_hosts setting on a manualroute router was being ignored
5853 for hosts that were looked up using the /MX notation.
5855 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5858 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5859 operated on the sender address. After changing the $sender_address to <>
5860 for the sender address verify, Exim was re-instated it as the original
5861 (before rewriting) address, but remembering that it had rewritten it, so it
5862 wasn't rewriting it again. This bug also had the effect of breaking the
5863 sender address verification caching when the sender address was rewritten.
5865 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5866 This has been changed so that if the ip literal address matches
5867 ignore_target_hosts, the router declines.
5869 7. Added expansion conditions match_domain, match_address, and match_local_
5870 part (NOT match_host).
5872 8. The placeholder for the Received: header didn't have a length field set.
5874 9. Added code to Exim itself and to exim_lock to test for a specific race
5875 condition that could lead to file corruption when using MBX delivery. The
5876 issue is with the lockfile that is created in /tmp. If this file is removed
5877 after a process has opened it but before that process has acquired a lock,
5878 there is the potential for a second process to recreate the file and also
5879 acquire a lock. This could lead to two Exim processes writing to the file
5880 at the same time. The added code performs the same test as UW imapd; it
5881 checks after acquiring the lock that its file descriptor still refers to
5882 the same named file.
5884 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5885 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5888 11. Added the smtp_active_hostname option. If used, this will typically be made
5889 to depend on the incoming interface address. Because $interface_address is
5890 not set up until the daemon has forked a reception process, error responses
5891 that can happen earlier (such as "too many connections") no longer contain
5894 12. If an expansion in a condition on a "warn" statement fails because a lookup
5895 defers, the "warn" statement is abandoned, and the next ACL statement is
5896 processed. Previously this caused the whole ACL to be aborted.
5898 13. Added the iplsearch lookup type.
5900 14. Added ident_timeout as a log selector.
5902 15. Added tls_certificate_verified as a log selector.
5904 16. Added a global option tls_require_ciphers (compare the smtp transport
5905 option of the same name). This controls incoming TLS connections.
5907 17. I finally figured out how to make tls_require_ciphers do a similar thing
5908 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5909 before starting the TLS session.
5911 18. Tabs are now shown as \t in -bP output.
5913 19. If the log selector return_path_on_delivery was set, Exim crashed when
5914 bouncing a message because it had too many Received: header lines.
5916 20. If two routers both had headers_remove settings, and the first one included
5917 a superfluous trailing colon, the final name in the first list and the
5918 first name in the second list were incorrectly joined into one item (with a
5919 colon in the middle).
5925 1. Added -C and -D options to the exinext utility, mainly to make it easier
5926 to include in the automated testing, but these could be helpful when
5927 multiple configurations are in use.
5929 2. The exinext utility was not formatting the output nicely when there was
5930 an alternate port involved in the retry record key, nor when there was a
5931 message id as well (for retries that were specific to a specific message
5932 and a specific host). It was also confused by IPv6 addresses, because of
5933 the additional colons they contain. I have fixed the IPv4 problem, and
5934 patched it up to do a reasonable job for IPv6.
5936 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5937 delivery, the log line now contains "pipelined" if PIPELINING was used.
5939 4. An SMTP transport process used to panic and die if the bind() call to set
5940 an explicit outgoing interface failed. This has been changed; it is now
5941 treated in the same way as a connect() failure.
5943 5. A reference to $sender_host_name in the part of a conditional expansion
5944 that was being skipped was still causing a DNS lookup. This no longer
5947 6. The def: expansion condition was not recognizing references to header lines
5948 that used bh_ and bheader_.
5950 7. Added the _cache feature to named lists.
5952 8. The code for checking quota_filecount in the appendfile transport was
5953 allowing one more file than it should have been.
5955 9. For compatibility with Sendmail, the command line option
5963 and sets the incoming protocol and host name (for trusted callers). The
5964 host name and its colon can be omitted when only the protocol is to be set.
5965 Note the Exim already has two private options, -pd and -ps, that refer to
5966 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5967 "s", but I don't think that's a major issue.
5969 10. A number of refactoring changes to the code, none of which should affect
5972 (a) The number of logging options was getting close to filling up the
5973 32-bit word that was used as a bit map. I have split them into two classes:
5974 those that are passed in the argument to log_write(), and those that are
5975 only ever tested independently outside of that function. These are now in
5976 separate 32-bit words, so there is plenty of room for expansion again.
5977 There is no change in the user interface or the logging behaviour.
5979 (b) When building, for example, log lines, the code previously used a
5980 macro that called string_cat() twice, in order to add two strings. This is
5981 not really sufficiently general. Furthermore, there was one instance where
5982 it was actually wrong because one of the argument was used twice, and in
5983 one call a function was used. (As it happened, calling the function twice
5984 did not affect the overall behaviour.) The macro has been replaced by a
5985 function that can join an arbitrary number of extra strings onto a growing
5988 (c) The code for expansion conditions now uses a table and a binary chop
5989 instead of a serial search (which was left over from when there were very
5990 few conditions). Also, it now recognizes conditions like "pam" even when
5991 the relevant support is not compiled in: a suitably worded error message is
5992 given if an attempt is made to use such a condition.
5994 11. Added ${time_interval:xxxxx}.
5996 12. A bug was causing one of the ddress fields not to be passed back correctly
5997 from remote delivery subprocesses. The field in question was not being
5998 subsequently used, so this caused to problems in practice.
6000 13. Added new log selectors queue_time and deliver_time.
6002 14. Might have fixed a bug in maildirsizefile handling that threw up
6003 "unexpected character" debug warnings, and recalculated the data
6004 unnecessarily. In any case, I expanded the warning message to give more
6007 15. Added the message "Restricted characters in address" to the statements in
6008 the default ACL that block characters like @ and % in local parts.
6010 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6011 Three changes have been made:
6013 (a) There was a serious bug; a negative response to MAIL caused the whole
6014 recipient domain to be cached as invalid, thereby blocking all messages
6015 to all local parts at the same domain, from all senders. This bug has
6016 been fixed. The domain is no longer cached after a negative response to
6017 MAIL if the sender used is not empty.
6019 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6022 (c) A new callout option, "use_sender" has been added for people who want
6023 the modified behaviour.
6029 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6032 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6033 indeed breaks things for older releases.
6035 3. Added additional logging to the case where there is a problem reading data
6036 from a filter that is running in a subprocess using a pipe, in order to
6037 try to track down a specific problem.
6039 4. Testing facility fudge: when running in the test harness and attempting
6040 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6041 getting "No route to host". Convert this to a timeout.
6043 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6046 6. Some OS don't have socklen_t but use size_t instead. This affects the
6047 fifth argument of getsockopt() amongst other things. This is now
6048 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6049 can be set for individual OS. I have set it for SunOS5, OSF1, and
6050 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6051 some earlier ones do not.
6053 7. Change 4.30/15 was not doing the test caselessly.
6055 8. The standard form for an IPv6 address literal was being rejected by address
6056 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6057 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6058 this, as well as the form without the "IPv6" on the front (but only when
6059 address literals are enabled, of course).
6061 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6063 10. Exim crashed if a message with an empty sender address specified by -f
6064 encountered a router with an errors_to setting. This could be provoked only
6065 by a command such as
6069 where an empty string was supplied; "<>" did not hit this bug.
6071 11. Installed PCRE release 4.5.
6073 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6074 remained set. It is now erased.
6076 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6077 times from message ids (which are base 36 rather than the normal 62).
6079 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6080 were being counted as actual protocol errors, and logged if the log
6081 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6082 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6083 and DATA following a set of rejected RCPTs do not count as protocol errors.
6084 In other words, Exim assumes they were pipelined, though this may not
6085 actually be the case. Of course, in all cases the client gets an
6086 appropriate error code.
6088 15. If a lookup fails in an ACL condition, a message about the failure may
6089 be available; it is used if testing the ACL cannot continue, because most
6090 such messages specify what the cause of the deferral is. However, some
6091 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6092 that caused an old message to be retained and used if a later statement
6093 caused a defer, replacing the real cause of the deferral.
6095 16. If an IP address had so many PTR records that the DNS lookup buffer
6096 was not large enough to hold them, Exim could crash while trying to process
6097 the truncated data. It now detects and logs this case.
6099 17. Further to 4.21/58, another change has been made: if (and only if) the
6100 first line of a message (the first header line) ends with CRLF, a bare LF
6101 in a subsequent header line has a space inserted after it, so as not to
6102 terminate the header.
6104 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6105 unnecessarily, used atoi() instead of strtol(), and didn't check the
6106 termination when getting file sizes from file names by regex.
6108 19. Completely re-implemented the support for maildirsize files, in the light
6109 of a number of problems with the previous contributed implementation
6110 (4.30/29). In particular:
6112 . If the quota is zero, the maildirsize file is maintained, but no quota is
6115 . If the maildir directory does not exist, it is created before any attempt
6116 to write a maildirsize file.
6118 . The quota value in the file is just a cache; if the quota is changed in
6119 the transport, the new value overrides.
6121 . A regular expression is available for excluding directories from the
6124 20. The autoreply transport checks the characters in options that define the
6125 message's headers; it allows continued headers, but it was checking with
6126 isspace() after an embedded newline instead of explicitly looking for a
6129 21. If all the "regular" hosts to which an address was routed had passed their
6130 expiry times, and had not reached their retry times, the address was
6131 bounced, even if fallback hosts were defined. Now Exim should go on to try
6134 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6135 equivalent code in the SMTP transport. Some hosts send humungous responses
6136 to HELO/EHLO, more than 1024 it seems.
6138 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6139 gives compiler warnings in some environments. I've now done it "properly",
6142 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6143 (because of problems with the built-in one) was declared to return uschar *
6144 instead of char *, causing compiler failure.
6146 25. Fixed a file descriptor leak when processing alias/forward files.
6148 26. Fixed a minor format string issue in dbfn.c.
6150 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6152 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6153 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6156 29. When a sender address is verified, it is cached, to save repeating the test
6157 when there is more than one recipient in a message. However, when the
6158 verification involves a callout, it is possible for different callout
6159 options to be set for different recipients. It is too complicated to keep
6160 track of this in the cache, so now Exim always runs a verification when a
6161 callout is required, relying on the callout cache for the optimization.
6162 The overhead is duplication of the address routing, but this should not be
6165 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6166 to be verified with callout=postmaster, and the main callout worked but the
6167 postmaster check failed, the verification correctly failed. However, if a
6168 subsequent RCPT command asked for sender verification *without* the
6169 postmaster check, incorrect caching caused this verification also to fail,
6172 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6173 it was not caching the DNS options (qualify_single, search_parents) that
6174 were used when the lookup failed. A subsequent lookup with different
6175 options therefore always gave the same answer, though there were cases
6176 where it should not have. (Example: a "domains = !$mx_any" option on a
6177 dnslookup router: the "domains" option is always processed without any
6178 widening, but the router might have qualify_single set.) Now Exim uses the
6179 cached value only when the same options are set.
6181 32. Added John Jetmore's "exipick" utility to the distribution.
6183 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6184 than a timeout (e.g. a certificate is required, and is not provided), an
6185 Exim server now closes the connection immediately. Previously it waited for
6186 the client to close - but if the client is SSL, it seems that they each
6187 wait for each other, leading to a delay before one of them times out.
6189 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6190 maintained 0.8.x compatibility because I don't think many are using it, and
6191 it is clearly obsolete.
6193 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6196 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6197 client certificate was expired. A simple patch fixes this, though I don't
6198 understand the full logic of why the verify callback is called multiple
6201 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6202 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6203 which causes problems with some clients (such as the Certicom SSL Plus
6204 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6205 disables the coutermeasure allowing Eudora to connect."
6207 38. Exim was not checking that a write() to a log file succeeded. This could
6208 lead to Bad Things if a log got too big, in particular if it hit a file
6209 size limit. Exim now panics and dies if it cannot write to a log file, just
6210 as it does if it cannot open a log file.
6212 39. Modified OS/Makefile-Linux so that it now contains
6214 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6216 The two -D definitions ensure that Exim is compiled with large file
6217 support, which makes it possible to handle log files that are bigger than
6220 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6221 instance) a domain was checked against a named list that involved a lookup,
6222 causing $domain_data to be set, then another domain was checked against the
6223 same list, then the first domain was re-checked, the value of $domain_data
6224 after the final check could be wrong. In particular, if the second check
6225 failed, it could be set empty. This bug probably also applied to
6228 41. The strip_trailing_dot option was not being applied to the address given
6229 with the -f command-line option.
6231 42. The code for reading a message's header from the spool was incrementing
6232 $received_count, but never initializing it. This meant that the value was
6233 incorrect (doubled) while delivering a message in the same process in which
6234 it was received. In the most common configuration of Exim, this never
6235 happens - a fresh exec is done - but it can happen when
6236 deliver_drop_privilege is set.
6238 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6239 it now includes up to 150 characters of the unexpected data in the log
6242 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6243 and building data strings. The size of both of these buffers was 10 000
6244 bytes - far larger than anybody would *ever* want, thought I. Needless to
6245 say, somebody hit the limit. I have increased the maximum line length to
6246 20 000 and the maximum data length of concatenated lines to 100 000. I have
6247 also fixed two bugs, because there was no checking on these buffers. Tsk,
6248 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6249 buffer is too small.
6251 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6252 lsearch lookups. Now it does.
6254 46. When parsing a route_list item in a manualroute router, a fixed-length
6255 buffer was used for the list of hosts. I made this 1024 bytes long,
6256 thinking that nobody would ever have a list of hosts that long. Wrong.
6257 Somebody had a whole pile of complicated expansion conditions, and the
6258 string was silently truncated, leading to an expansion error. It turns out
6259 that it is easier to change to an unlimited length (owing to other changes
6260 that have happened since this code was originally written) than to build
6261 structure for giving a limitation error. The length of the item that
6262 expands into the list of hosts is now unlimited.
6264 47. The lsearch lookup could not handle data where the length of text line was
6265 more than 4095 characters. Such lines were truncated, leading to shortened
6266 data being returned. It should now handle lines of any length.
6268 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6269 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6272 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6273 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6274 no longer happens in comment lines. A list of replacements is now placed
6275 at the head of all of the source files, except those whose only change is
6276 to replace PERL_COMMAND in the very first #! line.
6278 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6279 messages on the queue, with a bottom-up merge sort, using code contributed
6280 by Michael Haardt. This should make operations like -bp somewhat faster on
6281 large queues. It won't affect queue runners, except when queue_run_in_order
6284 51. Installed eximstats 1.31 in the distribution.
6286 52. Added support for SRV lookups to the dnslookup router.
6288 53. If an ACL referred to $message_body or $message_body_end, the value was not
6289 reset for any messages that followed in the same SMTP session.
6291 54. The store-handling optimization for building very long strings was not
6292 differentiating between the different store pools. I don't think this
6293 actually made any difference in practice, but I've tidied it.
6295 55. While running the routers to verify a sender address, $sender_address
6296 was still set to the sender address. This is wrong, because when routing to
6297 send a bounce to the sender, it would be empty. Therefore, I have changed
6298 it so that, while verifying a sender address, $sender_address is set to <>.
6299 (There is no change to what happens when verifying a recipient address.)
6301 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6302 target A or AAAA records (if not already returned) without resetting the
6303 qualify_single or search_parents options of the DNS resolver. These are
6304 inappropriate in this case because the targets of MX and SRV records must
6305 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6306 target that, when qualified, matched something in the local domain. These
6307 two options are now turned off when doing these lookups.
6309 57. It seems that at least some releases of Reiserfs (which does not have the
6310 concept of a fixed number of inodes) returns zero and not -1 for the
6311 number of available inodes. This interacted badly with check_spool_inodes,
6312 which assumed that -1 was the "no such thing" setting. What I have done is
6313 to check that the total number of inodes is greater than zero before doing
6314 the test of how many are available.
6316 58. When a "warn" ACL statement has a log_message modifier, the message is
6317 remembered, and not repeated. This is to avoid a lot of repetition when a
6318 message has many recipients that cause the same warning to be written.
6319 However, Exim was preserving the list of already written lines for an
6320 entire SMTP session, which doesn't seem right. The memory is now reset if a
6321 new message is started.
6323 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6324 debugging output unless log_rewrite was also set.
6326 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6327 of (int)(handle) when we know that handle contains (void *)(-1).
6329 61. The Exim daemon panic-logs an error return when it closes the incoming
6330 connection. However "connection reset by peer" seems to be common, and
6331 isn't really an error worthy of noting specially, so that particular error
6334 62. When Exim is trying to find all the local interfaces, it used to panic and
6335 die if the ioctl to get the interface flags failed. However, it seems that
6336 on at least one OS (Solaris 9) it is possible to have an interface that is
6337 included in the list of interfaces, but for which you get a failure error
6338 for this call. This happens when the interface is not "plumbed" into a
6339 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6340 failure of the "get flags" call assumes that the interface is down.
6342 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6343 makes life easier for people who are doing arithmetic on fields extracted
6344 from dates, where you often get leading zeros that should not be
6345 interpreted as octal.
6347 64. Added qualify_domain to the redirect router, to override the global
6350 65. If a pathologically long header line contained very many addresses (the
6351 report of this problem mentioned 10 000) and each of them was rewritten,
6352 Exim could use up a very large amount of memory. (It kept on making new
6353 copies of the header line as it rewrote, and never released the old ones.)
6354 At the expense of a bit more processing, the header rewriting function has
6355 been changed so that it no longer eats memory in this way.
6357 66. The generation of the Received: header has been moved from the time that a
6358 message starts to be received, to the time that it finishes. The timestamp
6359 in the Received: header should now be very close to that of the <= log
6360 line. There are two side-effects of this change:
6362 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6363 logged header lines no longer include the local Received: line, because
6364 it has not yet been created. The same applies to a copy of the message
6365 that is returned to a non-SMTP sender when a message is rejected.
6367 (b) When a filter file is tested using -bf, no additional Received: header
6368 is added to the test message. After some thought, I decided that this
6371 This change does not affect the value of $received_for. It is still set
6372 after address rewriting, but before local_scan() is called.
6374 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6376 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6377 gave an unhelpful panic error message, and a defer error. I have managed to
6378 change this behaviour so that it now rejects any supplied certificate,
6379 which seems right, as the list of acceptable certificates is empty.
6381 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6382 gave an unhelpful defer error. I have not managed to make this reject any
6383 supplied certificates, but the error message it gives is "no certificate
6384 supplied", which is not helpful.
6386 70. exigrep's output now also includes lines that are not associated with any
6387 message, but which match the given pattern. Implemented by a patch from
6388 Martin Sluka, which also tidied up the Perl a bit.
6390 71. Recipient callout verification, like sender verification, was using <> in
6391 the MAIL FROM command. This isn't really the right thing, since the actual
6392 sender may affect whether the remote host accepts the recipient or not. I
6393 have changed it to use the actual sender in the callout; this means that
6394 the cache record is now keyed on a recipient/sender pair, not just the
6395 recipient address. There doesn't seem to be a real danger of callout loops,
6396 since a callout by the remote host to check the sender would use <>.
6397 [SEE ABOVE: changed after hitting problems.]
6399 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6400 temporary errors. However, in the case of such a code being given after
6401 the end of a data transmission (i.e. after ".") Exim was failing to write
6402 a retry record for the message. (Yes, there was some broken host that was
6403 actually sending 8xx at this point.)
6405 73. An unknown lookup type in a host list could cause Exim to panic-die when
6406 the list was checked. (An example that provoked this was putting <; in the
6407 middle of a list instead of at the start.) If this happened during a DATA
6408 ACL check, a -D file could be left lying around. This kind of configuration
6409 error no longer causes Exim to die; instead it causes a defer error. The
6410 incident is still logged to the main and panic logs.
6412 74. Buglet left over from Exim 3 conversion. The message "too many messages
6413 in one connection" was written to the rejectlog but not the mainlog, except
6414 when address rewriting (yes!) was being logged.
6416 75. Added write_rejectlog option.
6418 76. When a system filter was run not as root (that is, when system_filter_user
6419 was set), the values of the $n variables were not being returned to the
6420 main process; thus, they were not subsequently available in the $sn
6423 77. Added +return_path_on_delivery log selector.
6425 78. A connection timeout was being treated differently from recipients deferred
6426 when testing hosts_max_try with a message that was older than the host's
6427 retry timeout. (The host should not be counted, thus allowing all hosts to
6428 be tried at least once before bouncing.) This may have been the cause of an
6429 occasionally reported bug whereby a message would remain on the queue
6430 longer than the retry timeout, but would be bounced if a delivery was
6431 forced. I say "may" because I never totally pinned down the problem;
6432 setting up timeout/retry tests is difficult. See also the next item.
6434 79. The ultimate address timeout was not being applied to errors that involved
6435 a combination of host plus message (for example, a timeout on a MAIL
6436 command). When an address resolved to a number of possible hosts, and they
6437 were not all tried for each delivery (e.g. because of hosts_max_try), a
6438 message could remain on the queue longer than the retry timeout.
6440 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6443 81. Fixed an obscure SMTP outgoing bug which required at least the following
6444 conditions: (a) there was another message waiting for the same server;
6445 (b) the server returned 5xx to all RCPT commands in the first message so
6446 that the message was not completed; (c) the server dropped the connection
6447 or gave a negative response to the RSET that Exim sends to abort the
6448 transaction. The observed case was a dropped connection after DATA that had
6449 been sent in pipelining mode. That is, the server had advertised PIPELINING
6450 but was not implementing it correctly. The effect of the bug was incorrect
6451 behaviour, such as trying another host, and this could lead to a crash.
6457 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6458 and daemon.c were passed as pointers to ints; they should have been
6459 pointers to socklen_t variables (which are typically unsigned ints).
6461 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6464 3. Fixed a really odd bug that affected only the testing scheme; patching a
6465 certain fixed string in the binary changed the value of another string that
6466 happened to be identical to the end of the original first string.
6468 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6469 name", it returns that address as the IP address. On some operating
6470 systems (e.g. Solaris), it also passes back the IP address string as the
6471 "host name". However, on others (e.g. Linux), it passes back an empty
6472 string. Exim wasn't checking for this, and was changing the host name to an
6473 empty string, assuming it had been canonicalized.
6475 5. Although rare, it is permitted to have more than one PTR record for a given
6476 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6477 all the names associated with an address, because they do in Solaris.
6478 However, it seems that they do not in Linux for data that comes from the
6479 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6480 I found this out when I moved to a new Linux workstation and tried to run
6481 the Exim test suite.
6483 To get round this problem I have changed the code so that it now does its
6484 own call to the DNS to look up PTR records when searching for a host name.
6485 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6486 addresses that are only in /etc/hosts are still found.
6488 This behaviour is, however, controlled by an option called host_lookup_
6489 order, which defaults to "bydns:byaddr". If people want to use the other
6490 order, or indeed, just use one or the other means of lookup, they can
6491 specify it in this variable.
6493 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6494 some operating systems, this comes back from gethostbyaddr() as an empty
6495 string, and this is what Exim used to test for. However, it seems that in
6496 other systems, "." is yielded. Exim now tests for this case too.
6498 7. The values of check_spool_space and check_log_space are now held internally
6499 as a number of kilobytes instead of an absolute number of bytes. If a
6500 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6501 kilobyte. This means that much larger values can be stored.
6503 8. Exim monitor: an attempt to get the action menu when not actually pointing
6504 at a message produces an empty menu entitled "No message selected". This
6505 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6506 no entries in it ("Shell widget menu has zero width and/or height"). So I
6507 have added a single, blank menu entry in this case.
6509 9. Added ${quote_local_part.
6511 10. MIME decoding is now applied to the contents of Subject: header lines when
6514 11. Now that a reference to $sender_host_address automatically causes a reverse
6515 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6516 host lookup before query-style lookups in lists that might use this
6517 variable. This has therefore been abolished, and the "net-" prefix is no
6518 longer necessary for query-style lookups.
6520 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6521 have been a typo for LFLAGS, so it has been changed.
6523 13. The install script calls Exim with "-C /dev/null" in order to find the
6524 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6525 to be output. However, since Exim outputs its version number before the
6526 error, it didn't break the script. It just looked ugly. I fixed this by
6527 always allowing "-C /dev/null" if the caller is root.
6529 14. Ignore overlarge ACL variable number when reading spool file - insurance
6530 against a later release with more variables having written the file.
6532 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6533 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6534 this, as well as the form without the "IPv6" on the front.
6536 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6537 OS/Makefile-Darwin file.
6539 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6540 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6542 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6543 variable. However, some OS use ints for inodes. I've added cast to long int
6544 to get rid of the compiler warning.
6546 19. I had forgotten to lock out "/../" in configuration file names when
6547 ALT_CONFIG_PREFIX was set.
6549 20. Routers used for verification do not need to specify transports. However,
6550 if such a router generated a host list, and callout was configured, Exim
6551 crashed, because it could not find a port number from the (non-existent)
6552 transport. It now assumes port 25 in this circumstance.
6554 21. Added the -t option to exigrep.
6556 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6557 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6558 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6559 them is set, LOOKUP_LSEARCH is forced.
6561 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6563 24. Added sender and host information to the "rejected by local_scan()" log
6564 line; previously there was no indication of these.
6566 25. Added .include_if_exists.
6568 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6569 Linux. It turns out that not all file systems support this. Apparently some
6570 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6571 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6572 not supported on the file descriptor, is now ignored when Exim is trying to
6573 sync a directory. This applies only to Linux.
6575 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6577 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6578 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6581 29. Added support for maildirsize files from supplied patch (modified a bit).
6583 30. The use of :fail: followed by an empty string could lead Exim to respond to
6584 sender verification failures with (e.g.):
6586 550 Verification failed for <xxx>
6587 550 Sender verify failed
6589 where the first response line was missing the '-' that indicates it is not
6590 the final line of the response.
6592 31. The loop for finding the name of the user that called Exim had a hardwired
6593 limit of 10; it now uses the value of finduser_retries, which is used for
6594 all other user lookups.
6596 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6599 33. Exim was neglecting to zero errno before one call of strtol() when
6600 expanding a string and expecting an integer value. On some systems this
6601 resulted in spurious "integer overflow" errors. Also, it was casting the
6602 result into an int without checking.
6604 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6605 did not work. The code looks as if it has *never* worked, though it appears
6606 to have been documented since at least release 1.62. I have made it work.
6608 35. The "timeout_DNS" error in retry rules, also documented since at least
6609 1.62, also never worked. As it isn't clear exactly what this means, and
6610 clearly it isn't a major issue, I have abolished the feature by treating it
6611 as "timeout", and writing a warning to the main and panic logs.
6613 36. The display of retry rules for -brt wasn't always showing the error code
6616 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6617 timeout_connect_A, timeout_connect_MX.
6619 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6620 to the empty sender.
6622 39. The daemon was not analysing the content of -oX till after it had closed
6623 stderr and disconnected from the controlling terminal. This meant that any
6624 syntax errors were only noted on the panic log, and the return code from
6625 the command was 0. By re-arranging the code a little, I've made the
6626 decoding happen first, so such errors now appear on stderr, and the return
6627 code is 1. However, the actual setting up of the sockets still happens in
6628 the disconnected process, so errors there are still only recorded on the
6631 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6632 connections (as happens on some IP stacks) was logged at start up time as
6633 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6634 it from "IPv6 and IPv4", which means that two separate sockets are being
6637 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6638 says whether AF_INET or AF_INET6 was passed as an argument.
6640 42. Exiwhat output was messed up when time zones were included in log
6643 43. Exiwhat now gives more information about the daemon's listening ports,
6644 and whether -tls-on-connect was used.
6646 44. The "port" option of the smtp transport is now expanded.
6648 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6649 silently ignored. Now an error message is written to the main and panic
6652 46. There's a new ACL modifier called "logwrite" which writes to a log file
6653 as soon as it is encountered.
6655 47. Added $local_user_uid and $local_user_gid at routing time.
6657 48. Exim crashed when trying to verify a sender address that was being
6660 49. Exim was recognizing only a space character after ".include". It now also
6661 recognizes a tab character.
6663 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6664 extracting the relevant information from the specification. The man page no
6665 longer contains scrambled data for the -d option, and I've added a section
6666 at the front about calling Exim under different names.
6668 51. Added "extra_headers" argument to the "mail" command in filter files.
6670 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6673 53. Installed eximstats 1.29.
6675 54. Added transport_filter_timeout as a generic transport option.
6677 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6678 Cc: header lines. This was required by RFC 822, but it not required by RFC
6681 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6682 incoming messages that did not have them. Now it does so only if the
6683 message originates locally, that is, if there is no associated remote host
6684 address. When Resent- header lines are present, this applies to the Resent-
6685 lines rather than the non-Resent- lines.
6687 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6688 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6690 58. Messages for configuration errors now include the name of the main
6691 configuration file - useful now that there may be more than one file in a
6692 list (.included file names were always shown).
6694 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6695 for those rare installations that do not start the daemon as root or run it
6696 setuid root. I've cut out the call to initgroups() if the daemon is not
6699 60. The Exim user and group can now be bound into the binary as text strings
6700 that are looked up at the start of Exim's processing.
6702 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6704 62. Added $mailstore_basename variable.
6706 63. Installed patch to sieve.c from Michael Haardt.
6708 64. When Exim failed to open the panic log after failing to open the main log,
6709 the original message it was trying to log was written to stderr and debug
6710 output, but if they were not available (the usual case in production), it
6711 was lost. Now it is written to syslog before the two lines that record the
6712 failures to open the logs.
6714 65. Users' Exim filters run in subprocesses under the user's uid. It is
6715 possible for a "deliver" command or an alias in a "personal" command to
6716 provoke an address rewrite. If logging of address rewriting is configured,
6717 this fails because the process is not running as root or exim. There may be
6718 a better way of dealing with this, but for the moment (because 4.30 needs
6719 to be released), I have disabled address rewrite logging when running a
6720 filter in a non-root, non-exim process.
6726 1. The buildconfig auxiliary program wasn't quoting the value set for
6727 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6728 not defined. This bug was masked in 4.22 by the effect that was fixed in
6731 2. Some messages that were rejected after a message id was allocated were
6732 shown as "incomplete" by exigrep. It no longer does this for messages that
6733 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6735 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6736 have allow_domain_literals set, the ID did not get logged in the <= line.
6737 Domain literals are now always recognized in Message-ID: header lines.
6739 4. The first argument for a ${extract expansion item is the key name or field
6740 number. Leading and trailing spaces in this item were not being ignored,
6741 causing some misleading effects.
6743 5. When deliver_drop_privilege was set, single queue runner processes started
6744 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6745 same command in the process it spins off) were not dropping privilege.
6747 6. When the daemon running as "exim" started a queue runner, it always
6748 re-executed Exim in the spun-off process. This is a waste of effort when
6749 deliver_drop_privilege is set. The new process now just calls the
6750 queue-runner function directly.
6756 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6759 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6760 ignored. Though the use of -oP was forcing the writing of a pid file, it
6761 was always written to the default place.
6763 3. If the message "no IP address found for host xxxx" is generated during
6764 incoming verification, it is now followed by identification of the incoming
6765 connection (so you can more easily find what provoked it).
6767 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6769 5. Added some features to "harden" Exim a bit more against certain attacks:
6771 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6772 be put in Local/Makefile. This is like the never_users runtime option,
6773 but it cannot be overridden. The default setting is "root".
6775 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6776 prefix string with which any file named in a -C command line option
6779 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6780 is retained for -C and -D only if the caller of Exim is root. Without
6781 it, the exim user may also use -C and -D and retain privilege.
6783 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6784 command line option is disabled.
6786 6. Macro names set by the -D option must start with an upper case letter, just
6787 like macro names defined in the configuration file.
6789 7. Added "dereference=" facility to LDAP.
6791 8. Two instances of the typo "uknown" in the source files are fixed.
6793 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6794 the Configure-Makefile script screwed up while processing it.
6796 10. Incorporated PCRE 4.4.
6798 11. The SMTP synchronization check was not operating right at the start of an
6799 SMTP session. For example, it could not catch a HELO sent before the client
6800 waited for the greeting. There is now a check for outstanding input at the
6801 point when the greeting is written. Because of the duplex, asynchronous
6802 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6803 way, but not yet received, when the check is performed.
6805 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6806 on TCP/IP sockets, because this apparently causes some broken clients to
6809 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6810 unchanged) from the Cygwin maintainer.
6812 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6813 was supported instead of testing for mbx. Effectively a typo.
6815 15. The spa authenticator server code was not checking that the input it
6816 received was valid base64.
6818 16. The debug output line for the "set" modifier in ACLs was not showing the
6819 name of the variable that was being set.
6821 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6823 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6824 lookup on its own. Something else was needed to trigger the lookup. For
6825 example, a match in host_lookup or the need for a host name in a host list.
6826 Now, if $sender_host_name is referenced and the host name has not yet been
6827 looked up, a lookup is performed. If the lookup fails, the variable remains
6828 empty, and $host_lookup_failed is set to "1".
6830 19. Added "eqi" as a case-independent comparison operator.
6832 20. The saslauthd authentication condition could segfault if neither service
6833 nor realm was specified.
6835 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6836 error message that was logged was misleading, and incoming SMTP
6837 connections were dropped. The message is now more accurate, and temporary
6838 errors are given to SMTP connections.
6840 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6841 (except RSET) with a 421 error, until QUIT is received. However, it was
6842 failing to send a response to QUIT.
6844 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6845 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6846 ACL is now run after the helo_try_verify_hosts code.
6848 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6849 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6850 case-independent, so other case variants are also recognized). Apparently
6851 some systems use these upper case variants.
6853 25. If more than two messages were waiting for the same host, and a transport
6854 filter was specified for the transport, Exim sent two messages over the
6855 same TCP/IP connection, and then failed with "socket operation on non-
6856 socket" when it tried to send the third.
6858 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6860 27. The extern definition of crypt16() in expand.c was not being excluded when
6861 the OS had its own crypt16() function.
6863 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6864 as a preferred synonym for return_size_limit, both as an option and as an
6867 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6869 30. Changed the default configuration ACL to relax the local part checking rule
6870 for addresses that are not in any local domains. For these addresses,
6871 slashes and pipe symbols are allowed within local parts, but the sequence
6872 /../ is explicitly forbidden.
6874 31. SPA server authentication was not clearing the challenge buffer before
6877 32. log_message in a "warn" ACL statement was writing to the reject log as
6878 well as to the main log, which contradicts the documentation and doesn't
6879 seem right (because no rejection is happening). So I have stopped it.
6881 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6882 However, I am unable to do any testing of this.
6884 34. Fixed an infelicity in the appendfile transport. When checking directories
6885 for a mailbox, to see if any needed to be created, it was accidentally
6886 using path names with one or more superfluous leading slashes; tracing
6887 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6889 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6890 contents are added to the log line that is written for every discarded
6891 recipient. (Previously a log_message setting was ignored.)
6893 36. The ${quote: operator now quotes the string if it is empty.
6895 37. The install script runs exim in order to find its version number. If for
6896 some reason other than non-existence or emptiness, which it checks, it
6897 could not run './exim', it was installing it with an empty version number,
6898 i.e. as "exim-". This error state is now caught, and the installation is
6901 38. An argument was missing from the function that creates an error message
6902 when Exim fails to connect to the socket for saslauthd authentication.
6903 This could cause Exim to crash, or give a corrupted message.
6905 39. Added isip, isip4, and isip6 to ${if conditions.
6907 40. The ACL variables $acl_xx are now saved with the message, and can be
6908 accessed later in routers, transports, and filters.
6910 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6911 strings in the file are not string-expanded.
6913 42. If a MAIL command specified a SIZE value that was too large to fit into an
6914 int variable, the check against message_size_limit failed. Such values are
6915 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6916 day this will have to be increased, but I don't think I want to be around
6917 when emails are that large.
6924 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6925 iconv() is not standard in FreeBSD.
6927 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6928 IPv6 enabled. The observed symptom was a segmentation fault on return from
6929 the function os_common_find_running_interfaces() in src/os.c.
6931 3. In the check_special_case() function in daemon.c I had used "errno" as an
6932 argument name, which causes warnings on some systems. This was basically a
6933 typo, since it was named "eno" in the comments!
6935 4. The code that waits for the clock to tick (at a resolution of some fraction
6936 of a second) so as to ensure message-id uniqueness was always waiting for
6937 at least one whole tick, when it could have waited for less. [This is
6938 almost certainly not relevant at current processor speeds, where it is
6939 unlikely to ever wait at all. But we try to future-proof.]
6941 5. The function that sleeps for a time interval that includes fractions of a
6942 second contained a race. It did not block SIGALRM between setting the
6943 timer, and suspending (a couple of lines later). If the interval was short
6944 and the sigsuspend() was delayed until after it had expired, the suspension
6945 never ended. On busy systems this could lead to processes getting stuck for
6948 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6949 process, before it forks any delivery processes. The open lookup caching
6950 mechanism meant that the open file or database connection was passed into
6951 the delivery process. The problem was that delivery processes always tidy
6952 up cached lookup data. This could cause a problem for the next delivery
6953 process started by the queue runner, because the external queue runner
6954 process does not know about the closure. So the next delivery process
6955 still has data in the lookup cache. In the case of a file lookup, there was
6956 no problem because closing a file descriptor in a subprocess doesn't affect
6957 the parent. However, if the lookup was caching a connection to a database,
6958 the connection was closed, and the second delivery process was likely to
6959 see errors such as "PGSQL: query failed: server closed the connection
6960 unexpectedly". The problem has been fixed by closing all cached lookups
6961 in a queue runner before running a delivery process.
6963 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6964 seem to have the "const" qualifier which it has on other OS. I've
6967 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6968 *of the same type* (client or server) with the same public name that an
6969 error should be diagnosed.
6971 9. When Exim looked up a host name for an IP address, but failed to find the
6972 original IP address when looking up the host name (a safety check), it
6973 output the message "<ip address> does not match any IP for NULL", which was
6974 confusing, to say the least. The bug was that the host name should have
6975 appeared instead of "NULL".
6977 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6978 user that is built into the binary, and the -C or -D options is used, root
6979 privilege is dropped before the configuration file is read. In addition,
6980 logging is switched to stderr instead of the normal log files. If the
6981 configuration then re-defines the Exim user, the unprivileged environment
6982 is probably not what is expected, so Exim logs a panic warning message (but
6985 However, if deliver_drop_privilege is set, the unprivileged state may well
6986 be exactly what is intended, so the warning has been cut out in that case,
6987 and Exim is allowed to try to write to its normal log files.
6993 1. smtp_return_error_details was not giving details for temporary sender
6994 or receiver verification errors.
6996 2. Diagnose a configuration error if two authenticators have the same public
6999 3. Exim used not to create the message log file for a message until the first
7000 delivery attempt. This could be confusing when incoming messages were held
7001 for policy or load reasons. The message log file is now created at the time
7002 the message is received, and an initial "Received" line is written to it.
7004 4. The automatically generated man page for command line options had a minor
7005 bug that caused no ill effects; however, a more serious problem was that
7006 the procedure for building the man page automatically didn't always
7007 operate. Consequently, release 4.20 contains an out-of-date version. This
7008 shouldn't happen again.
7010 5. When building Exim with embedded Perl support, the script that builds the
7011 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7012 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7014 6. The freeze_tell option was not being used for messages that were frozen on
7015 arrival, either by an ACL or by local_scan().
7017 7. Added the smtp_incomplete_transaction log selector.
7019 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7020 was accepting AUTH without a new EHLO.
7022 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7023 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7026 10. Logging of TCP/IP connections (when configured) now happens in the main
7027 daemon process instead of the child process, so that the TCP/IP connection
7028 count is more accurate (but it can never be perfect).
7030 11. The use of "drop" in a nested ACL was not being handled correctly in the
7031 outer ACL. Now, if condition failure induced by the nested "drop" causes
7032 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7033 or "require"), the connection is dropped.
7035 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7036 that yield "discard" can now be used with an "accept" or a "discard" verb,
7037 but an error is generated for any others (because I can't see a useful way
7038 to define what should happen).
7040 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7041 are now processed in the same way as lines in the Exim configuration file.
7042 In particular, continuation lines are supported.
7044 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7046 15. Added -ti meaning -t -i.
7048 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7049 domains, and warn by logging if others are found.
7051 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7052 returned from GIFCONF when Exim is trying to find the list of interfaces on
7053 a host. The code in os.c has been modified to copy each ifreq to an aligned
7054 structure in all cases.
7056 Also, in some cases, the returned ifreq's were being copied to a 'struct
7057 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7058 means the last couple of bytes of an IPv6 address could be chopped if the
7059 ifreq contained only a normal sockaddr (14 bytes storage).
7061 18. Named domain lists were not supported in the hosts_treat_as_local option.
7062 An entry such as +xxxx was not recognized, and was treated as a literal
7065 19. Ensure that header lines added by a DATA ACL are included in the reject log
7066 if the ACL subsequently rejects the message.
7068 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7069 MD5 (which is deprecated).
7071 21. When testing a filter file using -bf, Exim was writing a message when it
7072 took the sender from a "From " line in the message, but it was not doing so
7073 when it took $return_path from a Return-Path: header line. It now does.
7075 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7076 with a valid header line field name (a series of printing characters
7077 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7079 23. Changed "disc" in the source to "disk" to conform to the documentation and
7080 the book and for uniformity.
7082 24. Ignore Sendmail's -Ooption=value command line item.
7084 25. When execve() failed while trying to run a command in a pipe transport,
7085 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7086 could be confused with a return value of 69 from the command itself. This
7087 has been changed to 127, the value the shell returns if it is asked to run
7088 a non-existent command. The wording for the related log line suggests a
7089 non-existent command as the problem.
7091 26. If received_header_text expands to an empty string, do not add a Received:
7092 header line to the message. (Well, it adds a token one on the spool, but
7093 marks it "old" so that it doesn't get used or transmitted.)
7095 27. Installed eximstats 1.28 (addition of -nt option).
7097 28. There was no check for failure on the call to getsockname() in the daemon
7098 code. This can fail if there is a shortage of resources on the system, with
7099 ENOMEM, for example. A temporary error is now given on failure.
7101 29. Contrary to the C standard, it seems that in some environments, the
7102 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7103 program. Exim now does this explicitly; it affects the formatting of
7104 timestamps using strftime().
7106 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7107 complaints. I've now initialized all the variables, to avoid this.
7109 32. Header lines added by a system filter were not being "seen" during
7110 transport-time rewrites.
7112 33. The info_callback() function passed to OpenSSL is set up with type void
7113 (*)(SSL *, int, int), as described somewhere. However, when calling the
7114 function (actually a macro) that sets it up, the type void(*)() is
7115 expected. I've put in a cast to prevent warnings from picky compilers.
7117 34. If a DNS black list lookup found a CNAME record, but there were no A
7118 records associated with the domain it pointed at, Exim crashed.
7120 35. If a DNS black list lookup returned more than one A record, Exim ignored
7121 all but the first. It now scans all returned addresses if a particular IP
7122 value is being sought. In this situation, the contents of the
7123 $dnslist_value variable are a list of all the addresses, separated by a
7126 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7127 to accept a host name and all its aliases if the forward lookup for any of
7128 them yielded the IP address of the incoming connection. Now it accepts only
7129 those names whose forward lookup yields the correct IP address. Any other
7130 names are discarded. This closes a loophole whereby a rogue DNS
7131 administrator could create reverse DNS records to break through a
7132 wildcarded host restriction in an ACL.
7134 37. If a user filter or a system filter that ran in a subprocess used any of
7135 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7136 the wrong values were passed to the pipe command ($thisaddress had the
7137 value of $0, $0 had the value of $1, etc). This bug was introduced by
7138 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7140 38. Improved the line breaking for long SMTP error messages from ACLs.
7141 Previously, if there was no break point between 40 and 75 characters, Exim
7142 left the rest of the message alone. Two changes have been made: (a) I've
7143 reduced the minimum length to 35 characters; (b) if it can't find a break
7144 point between 35 and 75 characters, it looks ahead and uses the first one
7145 that it finds. This may give the occasional overlong line, but at least the
7146 remaining text gets split now.
7148 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7149 file descriptors might be low, and that setting 1000 would always raise it.
7150 It turns out that in some environments, the limit is already over 1000 and
7151 that lowering it causes trouble. So now Exim takes care not to decrease it.
7153 40. When delivering a message, the value of $return_path is set to $sender_
7154 address at the start of routing (routers may change the value). By an
7155 oversight, this default was not being set up when an address was tested by
7156 -bt or -bv, which affected the outcome if any router or filter referred to
7159 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7160 log only when "message" or "log_message" are set. However, if one of the
7161 conditions was an address verification, or a call to a nested ACL, the
7162 messages generated by the underlying test were being passed through. This
7163 no longer happens. The underlying message is available in $acl_verify_
7164 message for both "message" and "log_message" expansions, so it can be
7165 passed through if needed.
7167 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7168 new expansion $bh_ to give the encoded byte string without charset
7169 translation. Translation happens only if iconv() is available; HAVE_ICONV
7170 indicates this at build time. HEADERS_CHARSET gives the charset to
7171 translate to; headers_charset can change it in the configuration, and
7172 "headers charset" can change it in an individual filter file.
7174 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7175 that creates RFC 2047 encoded "words" labels them as that charset instead
7176 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7177 expansion operator; (ii) when Exim creates a From: line for a local
7178 message; (iii) when a header line is rewritten to include a "phrase" part.
7180 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7181 buggy, causing it to skip the first lines of messages whose message ID
7182 ended in 'D'. This would not have bitten before Exim release 4.14, because
7183 message IDs were unlikely to end in 'D' before then. The effect was to have
7184 incorrect size information for certain domains.
7186 45. #include "config.h" was missing at the start of the crypt16.c module. This
7187 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7190 46. If there was a timeout during a "random" callout check, Exim treated it as
7191 a failure of the random address, and carried on sending RSET and the real
7192 address. If the delay was just some slowness somewhere, the response to the
7193 original RCPT would be taken as a response to RSET and so on, causing
7194 mayhem of various kinds.
7196 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7197 when I implemented it. It didn't allow for the fact that some option values
7198 may legitimately be negative (e.g. size_addition), and it didn't even do
7199 the right test for positive values.
7201 48. Domain names in DNS records are case-independent. Exim always looks them up
7202 in lower case. Some resolvers return domain names in exactly the case they
7203 appear in the zone file, that is, they may contain uppercase letters. Not
7204 all resolvers do this - some return always lower case. Exim was treating a
7205 change of case by a resolver as a change of domain, similar to a widening
7206 of a domain abbreviation. This triggered its re-routing code and so it was
7207 trying to route what was effectively the same domain again. This normally
7208 caused routing to fail (because the router wouldn't handle the domain
7209 twice). Now Exim checks for this case specially, and just changes the
7210 casing of the domain that it ultimately uses when it transmits the message
7213 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7216 50. If a filter generated a file delivery with a non-absolute name (possible if
7217 no home directory exists for the router), the forbid_file option was not
7220 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7221 the existing equality matching.
7223 52. Exim was using ints instead of ino_t variables in some places where it was
7224 dealing with inode numbers.
7226 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7227 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7228 called TMPDIR, and if it finds it is different, it changes its value.
7230 54. The smtp_printf() function is now made available to local_scan() so
7231 additional output lines can be written before returning. There is also an
7232 smtp_fflush() function to enable the detection of a dropped connection.
7233 The variables smtp_input and smtp_batched_input are exported to
7236 55. Changed the default runtime configuration: the message "Unknown user"
7237 has been removed from the ACL, and instead placed on the localuser router,
7238 using the cannot_route_message feature. This means that any verification
7239 failures that generate their own messages won't get overridden. Similarly,
7240 the "Unrouteable address" message that was in the ACL for unverifiable
7241 relay addresses has also been removed.
7243 56. Added hosts_avoid_esmtp to the smtp transport.
7245 57. The exicyclog script was not checking for the esoteric option
7246 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7247 will work only if exicyclog is run under the appropriate euid.
7249 58. Following a discussion on the list, the rules by which Exim recognises line
7250 endings on incoming messages have been changed. The -dropcr and drop_cr
7251 options are now no-ops, retained only for backwards compatibility. The
7252 following line terminators are recognized: LF CRLF CR. However, special
7253 processing applies to CR:
7255 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7256 nor a local message in the state where . is a terminator.
7258 (ii) If a bare CR is encountered in a header line, an extra space is added
7259 after the line terminator so as not to end the header. The reasoning
7260 behind this is that bare CRs in header lines are most likely either
7261 to be mistakes, or people trying to play silly games.
7263 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7264 was being incorrectly given as 18 bytes larger than it should have been.
7265 This is a VOB (very old bug).
7267 60. This may never have affected anything current, but just in case it has:
7268 When the local host is found other than at the start of a list of hosts,
7269 the local host, those with the same MX, and any that follow, are discarded.
7270 When the list in question was part of a longer list of hosts, the following
7271 hosts (not currently being processed) were also being discarded. This no
7272 longer happens. I'm not sure if this situation could ever has previously
7275 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7278 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7279 header. This is something that is recommended in a new Internet Draft, and
7280 is something that is documented as being done by Sendmail. There are two
7281 possible values. For messages generated by the autoreply transport, Exim
7284 Auto-Submitted: auto-replied
7286 whereas for all other generated messages (e.g. bounces) it adds
7288 Auto-Submitted: auto-generated
7290 63. The "personal" condition in filters now includes a test for the
7291 Auto-Submitted: header. If it contains the string "auto-" the message it
7292 not considered personal.
7294 64. Added rcpt_include_affixes as a generic transport option.
7296 65. Added queue_only_override (default true).
7298 66. Added the syslog_duplication option.
7300 67. If what should have been the first header line of a message consisted of
7301 a space followed by a colon, Exim was mis-interpreting it as a header line.
7302 It isn't of course - it is syntactically invalid and should therefore be
7303 treated as the start of the message body. The misbehaviour could have
7304 caused a number of strange effects, including loss of data in subsequent
7305 header lines, and spool format errors.
7307 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7308 client host had authenticated. This control can now be exercised by an ACL
7309 for more flexibility.
7311 69. By default, callouts do not happen when testing with -bh. There is now a
7312 variant, -bhc, which does actually run the callout code, including
7313 consulting and updating the callout cache.
7315 70. Added support for saslauthd authentication, courtesy of Alexander
7318 71. If statvfs() failed on the spool or log directories while checking their
7319 size for availability, Exim confusingly gave the error "space shortage".
7320 Furthermore, in debugging mode it crashed with a floating point exception.
7321 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7322 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7323 serious problem, Exim now writes to the main and panic logs when this
7324 happens, with details of the failure. It then refuses to accept the
7325 incoming message, giving the message "spool directory problem" or "log
7326 directory problem" with a 421 code for SMTP messages.
7328 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7329 0, 1, and 2 exist, because some OS complain for execs without them (see
7330 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7331 descriptors. However, the code omitted to check that the open succeeded,
7332 causing mysterious errors if for some reason the permissions on /dev/null
7333 got screwed. Now Exim writes a message to the main and panic logs, and
7334 bombs out if it can't open /dev/null.
7336 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7337 interact so that it is all more flexible. It is supposed to remain
7338 backwards compatible. Also added extra_local_interfaces.
7340 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7341 to bomb out with an assertion failure - to the client this appears as a
7342 connection drop. This problem occurs in the part of the code that was taken
7343 from the Samba project. Fortunately, the assertion is in a very simple
7344 function, so I have fixed this by reproducing the function inline in the
7345 one place where it is called, and arranging for authentication to fail
7346 instead of killing the process with assert().
7348 75. The SPA client code was not working when the server requested OEM rather
7349 than Unicode encoding.
7351 76. Added code to make require_files with a specific uid setting more usable in
7352 the case where statting the file as root fails - usually a non-root-mounted
7353 NFS file system. When this happens and the failure is EACCES, Exim now
7354 forks a subprocess and does the per-uid checking as the relevant uid.
7356 77. Added process_log_path.
7358 78. If log_file_path was not explicitly set, a setting of check_log_space or
7359 check_log_inodes was ignored.
7361 79. If a space check for the spool or log partitions fails, the incident is now
7362 logged. Of course, in the latter case the data may get lost...
7364 80. Added the %p formatting code to string_format() so that it can be used to
7365 print addresses in debug_print(). Adjusted all the address printing in the
7366 debugging in store.c to use %p rather than %d.
7368 81. There was a concern that a line of code in smtp_in.c could overflow a
7369 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7370 initially expressed, the concern was not well-founded, because trailing
7371 spaces are removed early. However, if the trailing spaces were followed by
7372 a NULL, they did not get removed, so the overflow was possible. Two fixes
7375 (a) I re-wrote the offending code in a cleaner fashion.
7376 (b) If an incoming SMTP command contains a NULL character, it is rejected
7379 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7380 runs initgroups(), so that if the Exim user is in any additional groups,
7381 they will be used during message reception.
7387 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob