1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
12 side, to decode a TLS 1.3 packet capture.
14 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
15 Previously the default library behaviour applied, sending two, each in
18 JH/03 Debug output for ACL now gives the config file name and line number for
21 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
23 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
25 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
26 buffer overrun for (non-chunking) other transports.
28 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
29 TLS1.3, means that a server rejecting a client certificate is not visible
30 to the client until the first read of encrypted data (typically the
31 response to EHLO). Add detection for that case and treat it as a failed
32 TLS connection attempt, so that the normal retry-in-clear can work (if
35 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
36 and/or domain. Found and fixed by Jason Betts.
38 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
39 configuration). If a CNAME target was not a wellformed name pattern, a
42 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
43 the OS reports them interleaved with other addresses.
45 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
46 used both for input and for a verify callout, both encrypted, SMTP
47 responses being sent by the server could be lost. This resulted in
48 dropped connections and sometimes bounces generated by a peer sending
51 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
52 string. Previously it was possible to cause undefined behaviour in a
53 library routine (usually a crash). Found by "zerons".
55 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
58 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
59 API was removed, so update to use the newer ones.
61 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
62 any timeout set, is taking a long time. Previously we would hang on to a
63 rotated logfile "forever" if the input was arriving with long gaps
64 (a previous attempt to fix addressed lack, for a long time, of initial
67 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
68 shared (NFS) environment. The length of the tempfile name is now
69 4 + 16 ("hdr.$message_exim_id") which might break on file
70 systems which restrict the file name length to lower values.
73 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
74 shared (NFS) environment.
76 HS/02 Bug 2392: exigrep does case sensitive *option* processing (as it
77 did for all versions <4.90). Notably -M, -m, --invert, -I may be
80 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
81 on some platforms for bit 31.
83 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
84 to changes apparently associated with TLS1.3 handling some of the APIs
85 previously used were either nonfunctional or inappropriate. Strings
86 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
87 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
88 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
89 This affects log line X= elements, the $tls_{in,out}_cipher variables,
90 and the use of specific cipher names in the encrypted= ACL condition.
92 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
94 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
95 verification result was not updated unless hosts_require_ocsp applied.
97 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
98 queue_list_requires_admin set to false, non-admin users were denied the
101 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
102 directory-of-certs mode. Previously they were advertised despite the
105 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
106 A single TCP connection by a client will now hold a TLS connection open
107 for multiple message deliveries, by default. Previoud the default was to
110 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
111 default. If built with the facility, DANE will be used. The facility
112 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
114 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
115 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
116 must be defined and you must still, unless you define DISABLE_TLS, manage
117 the the include-dir and library-file requirements that go with that
118 choice. Non-TLS builds are still supported.
120 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
123 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
124 default. If the platfor supports and has the facility enabled, it will
125 be requested on all coneections.
131 JH/01 Remove code calling the customisable local_scan function, unless a new
132 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
134 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
135 non-signal-safe functions being used.
137 JH/03 Bug 2269: When presented with a received message having a stupidly large
138 number of DKIM-Signature headers, disable DKIM verification to avoid
139 a resource-consumption attack. The limit is set at twenty.
141 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
142 report of oldest_pass in ${authres } in consequence, and separate out
143 some descriptions of reasons for verification fail.
145 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
146 files in the spool were present and unlocked. A queue-runner could spot
147 them, resulting in a duplicate delivery. Fix that by doing the unlock
148 after the unlink. Investigation by Tim Stewart. Take the opportunity to
149 add more error-checking on spoolfile handling while that code is being
152 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
153 No known attacks, no CVE, this is defensive hardening.
155 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
156 a queue-runner could start a delivery while other operations were ongoing.
157 Cutthrough delivery was a common victim, resulting in duplicate delivery.
158 Found and investigated by Tim Stewart. Fix by using the open message data
159 file handle rather than opening another, and not locally closing it (which
160 releases a lock) for that case, while creating the temporary .eml format
161 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
163 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
164 $sender_verify_failure/$recipient_verify_failure to "random".
166 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
169 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
170 Previously this would segfault.
172 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
175 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
176 like zero, since the resolver should be doing this for us, But we need one
177 as a CNAME but no MX presence gets the CNAME returned; we need to check
178 that doesn't point to an MX to declare it "no MX returned" rather than
179 "error, loop". A new main option is added so the older capability of
180 following some limited number of chain links is maintained.
182 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
184 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
185 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
186 now-deprecated earlier definitions used only specified the range up to TLS
187 1.2 (in the older-version library docs).
189 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
191 JH/15 Rework TLS client-side context management. Stop using a global, and
192 explicitly pass a context around. This enables future use of TLS for
193 connections to service-daemons (eg. malware scanning) while a client smtp
194 connection is using TLS; with cutthrough connections this is quite likely.
196 JH/16 Fix ARC verification to do AS checks in reverse order.
198 JH/17 Support a "tls" option on the ${readsocket } expansion item.
200 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
201 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
202 Previously the "utf8" would be re-prepended for every additional message.
204 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
205 Previously thery were accepted, resulting in issues when attempting to
206 forward messages to a non-supporting MTA.
208 PP/02 Let -n work with printing macros too, not just options.
210 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
211 one parent address was copied, and bogus data was used at delivery-logging
212 time. Either a crash (after delivery) or bogus log data could result.
213 Discovery and analysis by Tim Stewart.
215 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
216 Previously if the string ended mid-character, we did not insert the
217 promised '?' replacement.
219 PP/04 Documentation: current string operators work on bytes, not codepoints.
221 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
222 should pack well giving a smaller memory footprint so better caching and
223 therefore performance. Group the declarations where this can't be done so
224 that the byte-sized flag variables are not interspersed among pointer
225 variables, giving a better chance of good packing by the compiler.
227 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
228 non-null, to avoid issues with sites running BATV. Previously reports were
229 sent with an empty envelope sender so looked like bounces.
231 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
232 The ignore_error flag wasn't being returned from the filter subprocess so
233 was not set for later routers. Investigation and fix by Matthias Kurz.
235 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
236 and a msg:complete for the whole, when a message is manually removed using
237 -Mrm. Developement by Matthias Kurz, hacked on by JH.
239 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
240 a "Gnu special" function, asprintf() in the DB utility binary builds; I
241 hope that is portable enough.
243 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
244 requiring a known-CA anchor certificate; make it now rely entirely on the
245 TLSA as an anchor. Checking the name on the leaf cert against the name
246 on the A-record for the host is still done for TA (but not for EE mode).
248 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
249 included in delivery lines for non-proxied connections, when compiled with
250 SUPPORT_SOCKS and running with proxy logging enabled.
252 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
253 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
254 move the existing event to fire before the normal logging of message
255 failure so that custom logging is bracketed by normal logging.
257 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
258 msg:fail:internal event. Developement by Matthias Kurz.
260 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
261 far too small for todays use of crypto signatures stored there. Go all
262 the way to the max DNS message size of 64kB, even though this might be
263 overmuch for IOT constrained device use.
265 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
266 copy a string over itself. The library routine is documented as not
267 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
269 JH/32 For main options check_spool_space and check_inode_space, where the
270 platform supports 64b integers, support more than the previous 2^31 kB
271 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
272 the previous G, M, k.
274 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
275 $authenticated_fail_id variable on authentication failure. Previously
278 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
279 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
280 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
281 GNUTLS_SEC_PARAM_MEDIUM.
283 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
284 side. Previously we would continue as if no SNI had been received.
286 JH/36 Harden the handling of string-lists. When a list consisted of a sole
287 "<" character, which should be a list-separator specification, we walked
288 off past the nul-terimation.
290 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
291 causes) even when the retry time is not yet met. Previously they were
292 not, meaning that when (say) an account was over-quota and temp-rejecting,
293 and multiple senders' messages were queued, only one sender would get
294 notified on each configured delay_warning cycle.
296 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
298 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
299 cipher-suites, an error can be left on the stack even for a succeeding
300 accept; this results in impossible error messages when a later operation
303 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
304 return error codes indicating retry. Under TLS1.3 this becomes required.
306 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
307 it only wrote the new authenticators, resulting in a lack of tracking of
308 peer changes of ESMTP extensions until the next cache flush.
310 JH/41 Fix the loop reading a message header line to check for integer overflow,
311 and more-often against header_maxsize. Previously a crafted message could
312 induce a crash of the recive process; now the message is cleanly rejected.
314 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
315 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
321 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
322 When redis_servers is set to a list of > 1 element, and the Redis servers
323 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
324 case of MOVED into a DEFER case instead, thus moving the query onto the
325 next server in the list. For a cluster of N elements, all N servers must
326 be defined in redis_servers.
328 GF/02 Catch and remove uninitialized value warning in exiqsumm
329 Check for existence of @ARGV before looking at $ARGV[0]
331 JH/01 Replace the store_release() internal interface with store_newblock(),
332 which internalises the check required to safely use the old one, plus
333 the allocate and data copy operations duplicated in both (!) of the
334 extant use locations.
336 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
337 modifier. This matches the restriction on the commandline.
339 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
340 Previously only the last row was returned.
342 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
343 we assumed that tags in the header were well-formed, and parsed the
344 element content after inspecting only the first char of the tag.
345 Assumptions at that stage could crash the receive process on malformed
348 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
349 While running the DKIM ACL we operate on the Permanent memory pool so that
350 variables created with "set" persist to the DATA ACL. Also (at any time)
351 DNS lookups that fail create cache records using the Permanent pool. But
352 expansions release any allocations made on the current pool - so a dnsdb
353 lookup expansion done in the DKIM ACL releases the memory used for the
354 DNS negative-cache, and bad things result. Solution is to switch to the
355 Main pool for expansions.
356 While we're in that code, add checks on the DNS cache during store_reset,
357 active in the testsuite.
358 Problem spotted, and debugging aided, by Wolfgang Breyha.
360 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
361 When none of the hosts presented to a transport match an already-open
362 connection, close it and proceed with the list. Previously we would
363 queue the message. Spotted by Lena with Yahoo, probably involving
366 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
367 Previously a spurious "250 OK id=" response was appended to the proper
370 JH/08 The "support for" informational output now, which built with Content
371 Scanning support, has a line for the malware scanner interfaces compiled
372 in. Interface can be individually included or not at build time.
374 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
375 by the template makefile "src/EDITME". The "STREAM" support for an older
376 ClamAV interface method is removed.
378 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
379 rows affected is given instead).
381 JH/11 The runtime Berkeley DB library version is now additionally output by
382 "exim -d -bV". Previously only the compile-time version was shown.
384 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
385 SMTP connection. Previously, when one had more receipients than the
386 first, an abortive onward connection was made. Move to full support for
387 multiple onward connections in sequence, handling cutthrough connection
388 for all multi-message initiating connections.
390 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
391 routers. Previously, a multi-recipient message would fail to match the
392 onward-connection opened for the first recipient, and cause its closure.
394 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
395 a timeout on read on a GnuTLS initiating connection, resulting in the
396 initiating connection being dropped. This mattered most when the callout
397 was marked defer_ok. Fix to keep the two timeout-detection methods
400 JH/15 Relax results from ACL control request to enable cutthrough, in
401 unsupported situations, from error to silently (except under debug)
402 ignoring. This covers use with PRDR, frozen messages, queue-only and
405 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
407 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
408 metadata, resulting in a crash in free().
410 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
411 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
412 Broken also in d185889f4, with init system revamp.
414 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
415 Previously we abruptly closed the connection after reading a malware-
416 found indication; now we go on to read the "scan ok" response line,
419 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
420 ACL. Previously, a crash would result.
422 JH/19 Speed up macro lookups during configuration file read, by skipping non-
423 macro text after a replacement (previously it was only once per line) and
424 by skipping builtin macros when searching for an uppercase lead character.
426 JH/20 DANE support moved from Experimental to mainline. The Makefile control
427 for the build is renamed.
429 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
430 was allocated for every new TLS startup, meaning one per message. Fix
431 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
433 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
434 reported the original. Fix to report (as far as possible) the ACL
435 result replacing the original.
437 JH/23 Fix memory leak during multi-message connections using STARTTLS under
438 OpenSSL. Certificate information is loaded for every new TLS startup,
439 and the resources needed to be freed.
441 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
443 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
446 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
447 DATA response info to the (existing) per-recipient response info for
448 the "C=" log element. It can have useful tracking info from the
449 destination system. Patch from Simon Arlott.
451 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
452 length value. Previously this would segfault.
454 HS/02 Support Avast multiline protoocol, this allows passing flags to
455 newer versions of the scanner.
457 JH/28 Ensure that variables possibly set during message acceptance are marked
458 dead before release of memory in the daemon loop. This stops complaints
459 about them when the debug_store option is enabled. Discovered specifically
460 for sender_rate_period, but applies to a whole set of variables.
461 Do the same for the queue-runner and queue-list loops, for variables set
462 from spool message files. Do the same for the SMTP per-message loop, for
463 certain variables indirectly set in ACL operations.
465 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
466 as a multi-recipient message from a mailinglist manager). The coding had
467 an arbitrary cutoff number of characters while checking for more input;
468 enforced by writing a NUL into the buffer. This corrupted long / fast
469 input. The problem was exposed more widely when more pipelineing of SMTP
470 responses was introduced, and one Exim system was feeding another.
471 The symptom is log complaints of SMTP syntax error (NUL chars) on the
472 receiving system, and refused recipients seen by the sending system
473 (propating to people being dropped from mailing lists).
474 Discovered and pinpointed by David Carter.
476 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
477 replaced by the ${authresults } expansion.
479 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
481 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
482 allows proper process termination in container environments.
484 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
485 Previously the "final dot" had a newline after it; ensure it is CR,LF.
487 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
488 and "err_perm", deprecated since 4.83 when the RFC-defined words
489 "temperror" and "permerror" were introduced.
491 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
492 transport-filters or DKIM-signing. The restriction was lost in the
493 consolidation of verify-callout and delivery SMTP handling.
494 Extend the restriction to also cover ARC-signing.
496 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
497 in defer=pass mode supply a 450 to the initiator. Previously the message
500 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
501 tls_require_ciphers is used as before.
503 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
504 "pass_unscanned". Only tmpfails from the scanner are written to
505 the paniclog, as they may require admin intervention (permission
506 denied, license issues). Other scanner errors (like decompression
507 bombs) do not cause a paniclog entry.
509 JH/36 Fix reinitialisation of DKIM logging variable between messages.
510 Previously it was possible to log spurious information in receive log
513 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
514 triggered odd behaviour from Outlook Express clients.
516 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
519 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
520 since the IETF WG has not yet settled on that versus the original
521 "bare" representation.
523 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
524 Previously the millisecond value corrupted the output.
525 Fix also for syslog_pid=no and log_selector +pid, for which the pid
526 corrupted the output.
532 JH/01 Rework error string handling in TLS interface so that the caller in
533 more cases is responsible for logging. This permits library-sourced
534 string to be attached to addresses during delivery, and collapses
535 pairs of long lines into single ones.
537 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
538 during configuration. Wildcards are allowed and expanded.
540 JH/02 Rework error string handling in DKIM to pass more info back to callers.
541 This permits better logging.
543 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
544 do not close it down and have the child transport start it up again on
545 the passed-on TCP connection. Instead, proxy the child (and any
546 subsequent ones) for TLS via a unix-domain socket channel. Logging is
547 affected: the continued delivery log lines do not have any DNSSEC, TLS
548 Certificate or OCSP information. TLS cipher information is still logged.
550 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
551 identical IP addresses on different listening ports. Will also affect
554 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
555 add noisy ifdef guards to special-case this sillyness.
556 Patch from Bernd Kuhls.
558 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
559 than 255 are no longer allowed.
561 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
562 Disable the session-cache too, which might reduce our load. Since we
563 currrectly use a new context for every connection, both as server and
564 client, there is no benefit for these.
565 GnuTLS appears to not support tickets server-side by default (we don't
566 call gnutls_session_ticket_enable_server()) but client side is enabled
567 by default on recent versions (3.1.3 +) unless the PFS priority string
570 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
571 <https://reproducible-builds.org/specs/source-date-epoch/>.
573 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
574 the check for any unsuccessful recipients did not notice the limit, and
575 erroneously found still-pending ones.
577 JH/08 Pipeline CHUNKING command and data together, on kernels that support
578 MSG_MORE. Only in-clear (not on TLS connections).
580 JH/09 Avoid using a temporary file during transport using dkim. Unless a
581 transport-filter is involved we can buffer the headers in memory for
582 creating the signature, and read the spool data file once for the
583 signature and again for transmission.
585 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
586 4.77 as the kernel support then wasn't solid, having issues in 64bit
587 mode. Now, it's been long enough. Add support for FreeBSD also.
589 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
590 case where the routing stage had gathered several addresses to send to
591 a host before calling the transport for the first, we previously failed
592 to close down TLS in the old transport process before passing the TCP
593 connection to the new process. The new one sent a STARTTLS command
594 which naturally failed, giving a failed delivery and bloating the retry
595 database. Investigation and fix prototype from Wolfgang Breyha.
597 JH/12 Fix check on SMTP command input synchronisation. Previously there were
598 false-negatives in the check that the sender had not preempted a response
599 or prompt from Exim (running as a server), due to that code's lack of
600 awareness of the SMTP input buffering.
602 PP/04 Add commandline_checks_require_admin option.
603 Exim drops privileges sanely, various checks such as -be aren't a
604 security problem, as long as you trust local users with access to their
605 own account. When invoked by services which pass untrusted data to
606 Exim, this might be an issue. Set this option in main configuration
607 AND make fixes to the calling application, such as using `--` to stop
610 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
611 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
612 responses to those, into a single TLS record each way (this usually means
613 a single packet). As a side issue, smtp_enforce_sync now works on TLS
616 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
617 affects you only if you're dancing at the edge of the param size limits.
618 If you are, and this message makes sense to you, then: raise the
619 configured limit or use OpenSSL 1.1. Nothing we can do for older
622 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
623 cmdline element to get the documented default one. Previously it was
626 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
627 are unsuable for non-message-specific reasons - eg. connection timeout,
628 banner-time rejection.
630 JH/16 Fix logging of delivery remote port, when specified by router, under
633 PP/06 Repair manualroute's ability to take options in any order, even if one
634 is the name of a transport.
637 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
639 JH/17 Change the list-building routines interface to use the expanding-string
640 triplet model, for better allocation and copying behaviour.
642 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
643 Previously it was constructed the first time a possibly-matching string
644 was met in the configuration file input during startup; now it is done
647 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
648 compatible one, to avoid the (poorly documented) possibility of a config
649 file in the working directory redirecting the DB files, possibly correpting
650 some existing file. CVE-2017-10140 assigned for BDB.
652 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
653 cache-hot. Previously, although the result was properly cached, the
654 initial verify call returned a defer.
656 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
657 the main verify for receipient in uncached-mode.
659 JH/22 Retire historical build files to an "unsupported" subdir. These are
660 defined as "ones for which we have no current evidence of testing".
662 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
663 if present. Previously it was ignored.
665 JH/24 Start using specified-initialisers in C structure init coding. This is
666 a C99 feature (it's 2017, so now considered safe).
668 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
669 if was a fixed-sized field and bitmask ops via macros; it is now more
672 PP/07 GitHub PR 56: Apply MariaDB build fix.
673 Patch provided by Jaroslav Škarvada.
675 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
676 during Coverity cleanups [4.87 JH/47]
677 Diagnosis and fix provided by Michael Fischer v. Mollard.
679 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
680 the right size to place the terminating semicolon on its own folded
681 line, the header hash was calculated to an incorrect value thanks to
682 the (relaxed) space the fold became.
684 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
685 and confused the parent.
687 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
688 which could crash as a result. This could lead to undeliverable messages.
690 JH/28 Logging: "next input sent too soon" now shows where input was truncated
693 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
694 matters on fast-turnover and PID-randomising systems, which were getting
695 out-of-order delivery.
697 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
698 a possibly-overlapping copy. The symptom was that "Remote host closed
699 connection in response to HELO" was logged instead of the actual 4xx
702 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
703 Previously only that bufferd was discarded, resulting in SYMTP command
706 JH/32 DKIM: when a message has multiple signatures matching an identity given
707 in dkim_verify_signers, run the dkim acl once for each. Previously only
708 one run was done. Bug 2189.
710 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
711 "panic the current process" to "deliberately defer". The panic log is
712 still written with the problem list name; the mail and reject logs now
713 get a temp-reject line for the message that was being handled, saying
714 something like "domains check lookup or other defer". The SMTP 451
715 message is still "Temporary local problem".
717 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
718 A crafted sequence of BDAT commands could result in in-use memory beeing
719 freed. CVE-2017-16943.
721 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
722 from SMTP input. Previously it was always done; now only done for DATA
723 and not BDAT commands. CVE-2017-16944.
725 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
726 to the message (such as an overlong header line). Previously this was
727 not done and we did not exit BDAT mode. Followon from the previous item
728 though a different problem.
734 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
735 than -2003 did; needs libidn2 in addition to libidn.
737 JH/02 The path option on a pipe transport is now expanded before use.
739 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
740 Patch provided by "Björn", documentation fix added too.
742 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
743 missing a wire-to-host endian conversion.
745 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
746 close after a BDAT command line could be taken as a following command,
747 giving a synch failure. Fix by only checking for synch immediately
748 before acknowledging the chunk.
750 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
751 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
752 macro. Patches provided by Josh Soref.
754 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
755 Previously we did not; the RFC seems ambiguous and VRFY is not listed
756 by IANA as a service extension. However, John Klensin suggests that we
759 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
760 the dkim code may be unix-mode line endings rather than smtp wire-format
761 CRLF, so prepend a CR to any bare LF.
763 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
764 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
766 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
767 the first were themselves being wrongly included in the feed into dkim
768 processing; with most chunk sizes in use this resulted in an incorrect
769 body hash calculated value.
771 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
772 DKIM signature block, for verification. Although advised against by
773 standards it is specifically not ruled illegal.
775 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
777 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
778 missing a body hash (the bh= tag).
780 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
781 It seems that HAProxy sends the Proxy Protocol information in clear and
782 only then does a TLS startup, so do the same.
784 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
785 TCP connections (such as for Spamd) unless the daemon successfully set
786 Fast Open mode on its listening sockets. This fixes breakage seen on
787 too-old kernels or those not configured for Fast Open, at the cost of
788 requiring both directions being enabled for TFO, and TFO never being used
789 by non-daemon-related Exim processes.
791 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
792 endings, at least on the first header line. Try to canonify any that get
793 past that check, despite the cost.
795 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
796 now limited to an arbitrary five deep, while parsing addresses with the
797 strip_excess_angle_brackets option enabled.
799 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
800 instead leave the unprompted TLS handshake in socket buffer for the
801 TLS library to consume.
803 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
805 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
807 JH/16 Drop variables when they go out of scope. Memory management drops a whole
808 region in one operation, for speed, and this leaves assigned pointers
809 dangling. Add checks run only under the testsuite which checks all
810 variables at a store-reset and panics on a dangling pointer; add code
811 explicitly nulling out all the variables discovered. Fixes one known
812 bug: a transport crash, where a dangling pointer for $sending_ip_address
813 originally assigned in a verify callout, is re-used.
815 PP/06 Drop '.' from @INC in various Perl scripts.
817 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
819 PP/08 Reduce a number of compilation warnings under clang; building with
820 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
821 should be warning-free.
823 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
825 HS/01 Fix portability problems introduced by PP/08 for platforms where
826 realloc(NULL) is not equivalent to malloc() [SunOS et al].
828 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
829 chunk. This allows us to accept broken chunked messages. We need a more
830 general solution here.
832 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
833 already-broken messages in the queue.
835 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
837 JH/19 Fix reference counting bug in routing-generated-address tracking.
843 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
844 supports it and a size is available (ie. the sending peer gave us one).
846 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
847 years of being deprecated). The replacements are the ACLs
848 acl_smtp_mime and acl_not_smtp_mime.
850 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
851 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
852 or even an in-clear connection were permitted. Now, if the host lookup
853 was dnssec and dane was requested then the host is only used if the
854 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
855 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
856 if one fails this test.
857 This means that a poorly-configured remote DNS will make it incommunicado;
858 but it protects against a DNS-interception attack on it.
860 JH/04 Bug 1810: make continued-use of an open smtp transport connection
861 non-noisy when a race steals the message being considered.
863 JH/05 If main configuration option tls_certificate is unset, generate a
864 self-signed certificate for inbound TLS connections.
866 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
867 in rewrites and routers.
869 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
870 and logged a warning sing 4.83; now they are a configuration file error.
872 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
873 (lacking @domain). Apply the same qualification processing as RCPT.
875 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
877 JH/10 Support ${sha256:} applied to a string (as well as the previous
880 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
881 a cutthrough deliver is pending, as we always want to make a connection.
882 This also avoids re-routing the message when later placing the cutthrough
883 connection after a verify cache hit.
884 Do not update it with the verify result either.
886 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
887 when routing results in more than one destination address.
889 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
890 signing (which inhibits the cutthrough capability). Previously only
891 the presence of an option was tested; now an expansion evaluating as
892 empty is permissible (obviously it should depend only on data available
893 when the cutthrough connection is made).
895 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
896 the relevant preceding SMTP command did not note the pipelining mode.
898 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
899 Previously they were not counted.
901 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
902 as one having no matching records. Previously we deferred the message
903 that needed the lookup.
905 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
906 distinguished as "(=".
908 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
909 for missing MX records. Previously it only worked for missing A records.
911 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
913 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
914 after the data-go-ahead and data-ack. Patch from Jason Betts.
916 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
917 even for a "none" policy. Patch from Tony Meyer.
919 JH/22 Fix continued use of a connection for further deliveries. If a port was
920 specified by a router, it must also match for the delivery to be
923 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
924 When one of the recipients of a message was unsuitable for the connection
925 (has no matching addresses), we lost track of needing to mark it
926 deferred. As a result mail would be lost.
928 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
930 JH/25 Decoding ACL controls is now done using a binary search; the source code
931 takes up less space and should be simpler to maintain. Merge the ACL
932 condition decode tables also, with similar effect.
934 JH/26 Fix problem with one_time used on a redirect router which returned the
935 parent address unchanged. A retry would see the parent address marked as
936 delivered, so not attempt the (identical) child. As a result mail would
939 JH/27 Fix a possible security hole, wherein a process operating with the Exim
940 UID can gain a root shell. Credit to http://www.halfdog.net/ for
941 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
944 JH/28 Enable {spool,log} filesystem space and inode checks as default.
945 Main config options check_{log,spool}_{inodes,space} are now
946 100 inodes, 10MB unless set otherwise in the configuration.
948 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
949 Previously it only applied to the main-section connection policy
952 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
954 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
955 by me. Added RFC7919 DH primes as an alternative.
957 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
958 are not in the system include path.
960 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
961 GnuTLS, when a session startup failed (eg because the client disconnected)
962 Exim did stdio operations after fclose. This was exposed by a recent
963 change which nulled out the file handle after the fclose.
965 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
966 signed directly by the cert-signing cert, rather than an intermediate
967 OCSP-signing cert. This is the model used by LetsEncrypt.
969 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
971 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
972 an incoming connection.
974 HS/02 Bug 1802: Do not half-close the connection after sending a request
977 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
978 fallback to "prime256v1".
980 JH/34 SECURITY: Use proper copy of DATA command in error message.
981 Could leak key material. Remotely exploitable. CVE-2016-9963.
987 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
988 and 3.4.4 - once the server is enabled to respond to an OCSP request
989 it does even when not requested, resulting in a stapling non-aware
990 client dropping the TLS connection.
992 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
993 support variable-length bit vectors. No functional change.
995 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
996 The I= interface field on outgoing lines is now after the H= remote
997 host field, same as incoming lines. There is a separate
998 outgoing_interface log selector which allows you to disable the
1001 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1002 If not running log_selector +smtp_connection the mainlog would be held
1003 open indefinitely after a "too many connections" event, including to a
1004 deleted file after a log rotate. Leave the per net connection logging
1005 leaving it open for efficiency as that will be quickly detected by the
1006 check on the next write.
1008 HS/01 Bug 1671: Fix post transport crash.
1009 Processing the wait-<transport> messages could crash the delivery
1010 process if the message IDs didn't exist for some reason. When
1011 using 'split_spool_directory=yes' the construction of the spool
1012 file name failed already, exposing the same netto behaviour.
1014 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1015 mime_regex ACL conditions.
1017 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1018 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1019 to HELO, local diagnostic string.
1021 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1022 log line to debug. Even when configured with a tls authenticator many
1023 client connections are expected to not authenticate in this way, so
1024 an authenticate fail is not an error.
1026 HS/02 Add the Exim version string to the process info. This way exiwhat
1027 gives some more detail about the running daemon.
1029 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1030 matter for fast-change records such as DNSBLs.
1032 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1033 retry record, even if constant. There may be multiple transports with
1034 different interface settings and the retry behaviour needs to be kept
1037 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1039 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1041 JH/10 Bug 840: fix log_defer_output option of pipe transport
1043 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1044 hosts_randomize. This matters a lot when combined with mua_wrapper.
1046 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1047 ${quote_pgsql:<string>} operator.
1049 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1051 JH/14 Change header file name for freeradius-client. Relevant if compiling
1052 with Radius support; from the Gentoo tree and checked under Fedora.
1054 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1056 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1057 option transport_filter, meaning no filtering.
1059 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1061 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1062 defaults to "*" (all hosts). The variable is now available when not built
1063 with TLS, default unset, mainly to enable keeping the testsuite sane.
1064 If a server certificate is not supplied (via tls_certificate) an error is
1065 logged, and clients will find TLS connections fail on startup. Presumably
1066 they will retry in-clear.
1067 Packagers of Exim are strongly encouraged to create a server certificate
1068 at installation time.
1070 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1071 with the $config_file variable.
1073 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1074 in transport context, after the attempt, and per-recipient. The latter type
1075 is per host attempted. The event data is the error message, and the errno
1076 information encodes the lookup type (A vs. MX) used for the (first) host,
1077 and the trailing two digits of the smtp 4xx response.
1079 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1080 to write to mainlog (or rejectlog, paniclog) in the window between file
1081 creation and permissions/ownership being changed. Particularly affects
1082 installations where exicyclog is run as root, rather than exim user;
1083 result is that the running daemon panics and dies.
1085 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1087 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1088 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1089 "pri" and "weight". Note that the previous implicit priority given by the
1090 list order is no longer honoured.
1092 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1093 for DKIM processing.
1095 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1096 by defining SUPPORT_SOCKS.
1098 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1099 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1100 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1101 variables are renamed to proxy_{local,external}_{address,port}.
1103 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1104 for a build by defining SUPPORT_I18N
1106 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1107 of the query string, and make ${quote_redis:} do that quoting.
1109 JH/29 Move Events support from Experimental to mainline, enabled by default
1110 and removable for a build by defining DISABLE_EVENT.
1112 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1114 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1115 cached by the daemon.
1117 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1118 by defining LOOKUP_REDIS. The libhiredis library is required.
1120 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1121 keys are given for lookup.
1123 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1124 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1125 only supported when built with TLS support. The PolarSSL SHA routines
1126 are still used when the TLS library is too old for convenient support.
1128 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1129 openssl_options), for security. OpenSSL forces this from version 1.1.0
1130 server-side so match that on older versions.
1132 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1133 allocation for $value could be released as the expansion processing
1134 concluded, but leaving the global pointer active for it.
1136 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1137 and to use the domains and local_parts ACL conditions.
1139 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1140 incorrectly not doubled on cutthrough transmission, hence seen as a
1141 body-termination at the receiving system - resulting in truncated mails.
1142 Commonly the sender saw a TCP-level error, and retransmitted the message
1143 via the normal store-and-forward channel. This could result in duplicates
1144 received - but deduplicating mailstores were liable to retain only the
1145 initial truncated version.
1147 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1149 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1151 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1152 we're in there, support oversigning also; bug 1309.
1154 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1156 HS/04 Add support for keep_environment and add_environment options.
1158 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1159 either intentional arithmetic overflow during PRNG, or testing config-
1162 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1163 delivery resulted in actual delivery. Cancel cutthrough before DATA
1166 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1167 rejects a recipient: pass the reject to the originator.
1169 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1170 Many were false-positives and ignorable, but it's worth fixing the
1173 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1174 for the new environment-manipulation done at startup. Move the routines
1175 from being local to tls.c to being global via the os.c file.
1177 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1178 an extract embedded as result-arg for a map, the first arg for extract
1179 is unavailable so we cannot tell if this is a numbered or keyed
1180 extraction. Accept either.
1186 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1189 JH/02 The smtp transport option "multi_domain" is now expanded.
1191 JH/03 The smtp transport now requests PRDR by default, if the server offers
1194 JH/04 Certificate name checking on server certificates, when exim is a client,
1195 is now done by default. The transport option tls_verify_cert_hostnames
1196 can be used to disable this per-host. The build option
1197 EXPERIMENTAL_CERTNAMES is withdrawn.
1199 JH/05 The value of the tls_verify_certificates smtp transport and main options
1200 default to the word "system" to access the system default CA bundle.
1201 For GnuTLS, only version 3.0.20 or later.
1203 JH/06 Verification of the server certificate for a TLS connection is now tried
1204 (but not required) by default. The verification status is now logged by
1205 default, for both outbound TLS and client-certificate supplying inbound
1208 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1211 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1212 Status Notification (bounce) messages are now MIME format per RFC 3464.
1213 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1214 under the control of the dsn_advertise_hosts option, and routers may
1215 have a dsn_lasthop option.
1217 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1218 default, modifiable by a malware= option. The list separator for
1219 the options can now be changed in the usual way. Bug 68.
1221 JH/10 The smtp_receive_timeout main option is now expanded before use.
1223 JH/11 The incoming_interface log option now also enables logging of the
1224 local interface on delivery outgoing connections.
1226 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1227 if the interface and destination host and port all match.
1229 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1232 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1233 Patch from Andrew Lewis.
1235 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1236 now supports optional time-restrictions, weighting, and priority
1237 modifiers per server. Patch originally by <rommer@active.by>.
1239 JH/16 The spamd_address main option now supports a mixed list of local
1240 and remote servers. Remote servers can be IPv6 addresses, and
1241 specify a port-range.
1243 JH/17 Bug 68: The spamd_address main option now supports an optional
1244 timeout value per server.
1246 JH/18 Bug 1581: Router and transport options headers_add/remove can
1247 now have the list separator specified.
1249 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1252 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1255 JH/21 Support for the A6 type of dns record is withdrawn.
1257 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1258 rather than the verbs used.
1260 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1261 from 255 to 1024 chars.
1263 JH/24 Verification callouts now attempt to use TLS by default.
1265 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1266 are generic router options now. The defaults didn't change.
1268 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1269 Original patch from Alexander Shikoff, worked over by JH.
1271 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1272 files if ZCAT_COMMAND is not executable.
1274 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1276 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1278 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1279 Normally benign, it bites when the pair was led to by a CNAME;
1280 modern usage is to not canonicalize the domain to a CNAME target
1281 (and we were inconsistent anyway for A-only vs AAAA+A).
1283 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1285 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1286 when evaluating $sender_host_dnssec.
1288 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1289 $sender_helo_dnssec variable.
1291 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1293 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1295 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1297 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1298 documented as working, but never had. Support all but $spam_report.
1300 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1301 added for tls authenticator.
1303 HS/03 Add perl_taintmode main config option
1309 TL/01 When running the test suite, the README says that variables such as
1310 no_msglog_check are global and can be placed anywhere in a specific
1311 test's script, however it was observed that placement needed to be near
1312 the beginning for it to behave that way. Changed the runtest perl
1313 script to read through the entire script once to detect and set these
1314 variables, reset to the beginning of the script, and then run through
1315 the script parsing/test process like normal.
1317 TL/02 The BSD's have an arc4random API. One of the functions to induce
1318 adding randomness was arc4random_stir(), but it has been removed in
1319 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1320 function when detected.
1322 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1323 cause callback expansion.
1325 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1326 syntax errors in an expansion can be treated as a string instead of
1327 logging or causing an error, due to the internal use of bool_lax
1328 instead of bool when processing it.
1330 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1331 server certificates when making smtp deliveries.
1333 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1335 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1337 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1339 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1340 Merged patch from Sebastian Wiedenroth.
1342 JH/05 Fix results-pipe from transport process. Several recipients, combined
1343 with certificate use, exposed issues where response data items split
1344 over buffer boundaries were not parsed properly. This eventually
1345 resulted in duplicates being sent. This issue only became common enough
1346 to notice due to the introduction of connection certificate information,
1347 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1349 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1350 size buffer was used, resulting in syntax errors when an expansion
1353 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1354 version 3.3.6 or later.
1356 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1357 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1358 both become "event_action", the variables become $event_name, $event_data
1359 and $event_defer_errno. There is a new variable $verify_mode, usable in
1360 routers, transports and related events. The tls:cert event is now also
1361 raised for inbound connections, if the main configuration event_action
1364 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1365 early OCSP support, but no stapling (appears to be less than 1.0.0).
1367 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1368 server certificate names available under the smtp transport option
1369 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1372 JH/10 Time-related extraction expansions from certificates now use the main
1373 option "timezone" setting for output formatting, and are consistent
1374 between OpenSSL and GnuTLS compilations. Bug 1541.
1376 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1377 encoded parameter in the incoming message. Bug 1558.
1379 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1380 include certificate info, eximon was claiming there were spoolfile
1383 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1385 JH/14 Log delivery-related information more consistently, using the sequence
1386 "H=<name> [<ip>]" wherever possible.
1388 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1389 are problematic for Debian distribution, omit them from the release
1392 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1394 JH/16 Fix string representation of time values on 64bit time_t architectures.
1397 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1398 output list separator was used.
1403 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1404 checkers that were complaining about end of non-void function with no
1407 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1408 This was a regression introduced in 4.83 by another bugfix.
1410 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1412 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1413 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1419 TF/01 Correctly close the server side of TLS when forking for delivery.
1421 When a message was received over SMTP with TLS, Exim failed to clear up
1422 the incoming connection properly after forking off the child process to
1423 deliver the message. In some situations the subsequent outgoing
1424 delivery connection happened to have the same fd number as the incoming
1425 connection previously had. Exim would try to use TLS and fail, logging
1426 a "Bad file descriptor" error.
1428 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1429 utilities have not been installed.
1431 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1432 temporary space as the ACL may create new global variables.
1434 TL/01 LDAP support uses per connection or global context settings, depending
1435 upon the detected version of the libraries at build time.
1437 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1438 to extract and use the src ip:port in logging and expansions as if it
1439 were a direct connection from the outside internet. PPv2 support was
1440 updated based on HAProxy spec change in May 2014.
1442 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1444 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1445 Properly escape header and check for NULL return.
1447 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1450 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1452 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1453 characters in header names, implemented as a verify condition.
1454 Contributed by Michael Fischer v. Mollard.
1456 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1457 results permerror and temperror. Previous values are deprecated but
1458 still accepted. In a future release, err_perm and err_temp will be
1459 completely removed, which will be a backward incompatibility if the
1460 ACL tests for either of these two old results. Patch contributed by
1461 user bes-internal on the mailing list.
1463 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1465 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1466 selectors, in both main and reject logs.
1468 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1471 JH/07 Add malware type "sock" for talking to simple daemon.
1473 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1475 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1476 routers/transports under cutthrough routing.
1478 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1479 numbers. Touch up "bool" conditional to keep the same definition.
1481 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1483 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1485 JH/12 Expand items in router/transport headers_add or headers_remove lists
1486 individually rather than the list as a whole. Bug 1452.
1488 Required for reasonable handling of multiple headers_ options when
1489 they may be empty; requires that headers_remove items with embedded
1490 colons must have them doubled (or the list-separator changed).
1492 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1493 view the policy declared in the DMARC record. Currently, $dmarc_status
1494 is a combined value of both the record presence and the result of the
1497 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1499 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1500 dnslookup router and the smtp transport (applying to the forward
1503 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1504 of ldap servers used for a specific lookup. Patch provided by Heiko
1507 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1508 New variable $lookup_dnssec_authenticated for observability.
1510 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1511 Patch submitted by Lars Timman.
1513 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1515 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1516 Requires trusted mode and valid format message id, aborts otherwise.
1517 Patch contributed by Heiko Schlichting.
1519 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1520 certextract with support for various fields. Bug 1358.
1522 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1523 is requested by default, modifiable by smtp transport option
1526 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1527 operate on certificate variables to give certificate fingerprints
1528 Also new ${sha256:cert_variable}.
1530 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1532 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1533 Christian Aistleitner.
1535 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1537 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1538 file. Patch from Wolfgang Breyha.
1540 JH/25 Expand the coverage of the delivery $host and $host_address to
1541 client authenticators run in verify callout. Bug 1476.
1543 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1544 align with daemon_smtp_ports. Bug 72.
1546 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1547 support and error reporting did not work properly.
1549 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1550 and is readable. Patch from Andrew Colin Kissa.
1552 TL/14 Enhance documentation of ${run expansion and how it parses the
1553 commandline after expansion, particularly in the case when an
1554 unquoted variable expansion results in an empty value.
1556 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1558 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1559 by renaming away from it. Addresses GH issue 15, reported by
1562 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1564 TL/15 SECURITY: prevent double expansion in math comparison functions
1565 (can expand unsanitized data). Not remotely exploitable.
1572 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1574 PP/02 Make -n do something, by making it not do something.
1575 When combined with -bP, the name of an option is not output.
1577 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1580 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1581 $sender_host_name and config options to manage this, and basic check
1584 PP/05 DSCP support for outbound connections and control modifier for inbound.
1586 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1587 (Only plugin which currently uses this is kerberos4, which nobody should
1588 be using, but we should make it available and other future plugins might
1589 conceivably use it, even though it would break NAT; stuff *should* be
1590 using channel bindings instead).
1592 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1593 name; added for Sendmail compatibility; requires admin caller.
1594 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1595 just ignore it); requires trusted caller.
1596 Also parse but ignore: -Ac -Am -X<logfile>
1599 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1601 TL/02 Add +smtp_confirmation as a default logging option.
1603 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1604 Patch by Magnus Holmgren from 2007-02-20.
1606 TL/04 Bugzilla 1281 - Spec typo.
1607 Bugzilla 1283 - Spec typo.
1608 Bugzilla 1290 - Spec grammar fixes.
1610 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1612 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1614 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1615 mailing list by by Dmitry Isaikin.
1617 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1619 JH/02 Support "G" suffix to numbers in ${if comparisons.
1621 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1623 NM/01 Bugzilla 1197 - Spec typo
1624 Bugzilla 1196 - Spec examples corrections
1626 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1628 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1629 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1632 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1633 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1635 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1636 "acl {{name}{arg}...}", and optional args on acl condition
1639 JH/05 Permit multiple router/transport headers_add/remove lines.
1641 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1643 JH/07 Avoid using a waiting database for a single-message-only transport.
1644 Performance patch from Paul Fisher. Bugzilla 1262.
1646 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1649 JH/09 Add $headers_added variable, with content from use of ACL modifier
1650 add_header (but not yet added to the message). Bugzilla 199.
1652 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1653 Pulled from Bugzilla 817 by Wolfgang Breyha.
1655 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1657 (nb: this is the same fix as in Exim 4.80.1)
1659 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1662 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1665 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1667 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1668 advertise SMTP AUTH mechanism to us, instead of a generic
1669 protocol violation error. Also, make Exim more robust to bad
1670 data from the Dovecot auth socket.
1672 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1674 When a queue runner is handling a message, Exim first routes the
1675 recipient addresses, during which it prunes them based on the retry
1676 hints database. After that it attempts to deliver the message to
1677 any remaining recipients. It then updates the hints database using
1680 So if a recipient address works intermittently, it can get repeatedly
1681 deferred at routing time. The retry hints record remains fresh so the
1682 address never reaches the final cutoff time.
1684 This is a fairly common occurrence when a user is bumping up against
1685 their storage quota. Exim had some logic in its local delivery code
1686 to deal with this. However it did not apply to per-recipient defers
1687 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1689 This change adds a proper retry rule check during routing so that the
1690 final cutoff time is checked against the message's age. We only do
1691 this check if there is an address retry record and there is not a
1692 domain retry record; this implies that previous attempts to handle
1693 the address had the retry_use_local_parts option turned on. We use
1694 this as an approximation for the destination being like a local
1695 delivery, as in LMTP.
1697 I suspect this new check makes the old local delivery cutoff check
1698 redundant, but I have not verified this so I left the code in place.
1700 TF/02 Correct gecos expansion when From: is a prefix of the username.
1702 Test 0254 submits a message to Exim with the header
1706 When I ran the test suite under the user fanf2, Exim expanded
1707 the header to contain my full name, whereas it should have added
1708 a Resent-Sender: header. It erroneously treats any prefix of the
1709 username as equal to the username.
1711 This change corrects that bug.
1713 GF/01 DCC debug and logging tidyup
1714 Error conditions log to paniclog rather than rejectlog.
1715 Debug lines prefixed by "DCC: " to remove any ambiguity.
1717 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1719 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1720 Bug spotted by Jeremy Harris; was flawed since initial commit.
1721 Would have resulted in OCSP responses post-SNI triggering an Exim
1722 NULL dereference and crash.
1724 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1726 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1727 Bug detection, analysis and fix by Samuel Thibault.
1728 Bugzilla 1331, Debian bug #698092.
1730 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1732 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1733 Server implementation by Todd Lyons, client by JH.
1734 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1735 config variable "prdr_enable" controls whether the server
1736 advertises the facility. If the client requests PRDR a new
1737 acl_data_smtp_prdr ACL is called once for each recipient, after
1738 the body content is received and before the acl_smtp_data ACL.
1739 The client is controlled by both of: a hosts_try_prdr option
1740 on the smtp transport, and the server advertisement.
1741 Default client logging of deliveries and rejections involving
1742 PRDR are flagged with the string "PRDR".
1744 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1745 fclose(). Diagnosis by Todd Lyons.
1747 PP/17 Update configure.default to handle IPv6 localhost better.
1748 Patch by Alain Williams (plus minor tweaks).
1751 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1752 This is now consistent with GnuTLS, and is now documented: the
1753 previous undocumented portable approach to treating the option as
1754 unset was to force an expansion failure. That still works, and
1755 an empty string is now equivalent.
1757 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1758 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1759 not performing validation itself.
1761 PP/20 Added force_command boolean option to pipe transport.
1762 Patch from Nick Koston, of cPanel Inc.
1764 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1767 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1769 PP/21 Fix eximon continuous updating with timestamped log-files.
1770 Broken in a format-string cleanup in 4.80, missed when I repaired the
1771 other false fix of the same issue.
1772 Report and fix from Heiko Schlichting.
1775 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1776 Report from Prashanth Katuri.
1778 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1779 It's SecureTransport, so affects any MacOS clients which use the
1780 system-integrated TLS libraries, including email clients.
1782 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1783 using a MIME ACL for non-SMTP local injection.
1784 Report and assistance in diagnosis by Warren Baker.
1786 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1788 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1790 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1791 last id that failed so it may be referenced in subsequent ACL's.
1793 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1796 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1797 ldap library initialization, allowing self-signed CA's to be
1798 used. Also properly sets require_cert option later in code by
1799 using NULL (global ldap config) instead of ldap handle (per
1800 session). Bug diagnosis and testing by alxgomz.
1802 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1803 the src/util/ subdirectory.
1805 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1806 renamed to Transport Post Delivery Action by Jeremy Harris, as
1809 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1810 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1811 redis_servers = needs to be configured which will be used by the redis
1812 lookup. Patch from Warren Baker, of The Packet Hub.
1814 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1816 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1817 hostname or reverse DNS when processing a host list. Used suggestions
1818 from multiple comments on this bug.
1820 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1822 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1823 Missed a few lines, added it to make the runtest require no keyboard
1826 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1827 contains upper case chars. Make router use caseful_local_part.
1829 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1830 support when GnuTLS has been built with p11-kit.
1836 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1838 This, or similar/improved, will also be change PP/11 of 4.82.
1844 PP/01 Handle short writes when writing local log-files.
1845 In practice, only affects FreeBSD (8 onwards).
1846 Bugzilla 1053, with thanks to Dmitry Isaikin.
1848 NM/01 Bugzilla 949 - Documentation tweak
1850 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1853 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1855 PP/02 Implemented gsasl authenticator.
1857 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1859 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1860 `pkg-config foo` for cflags/libs.
1862 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1863 with rest of GSASL and with heimdal_gssapi.
1865 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1866 `pkg-config foo` for cflags/libs for the TLS implementation.
1868 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1869 properties get this fed in as external SSF. A number of robustness
1870 and debugging improvements to the cyrus_sasl authenticator.
1872 PP/08 cyrus_sasl server now expands the server_realm option.
1874 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1875 Patch by Jeremy Harris.
1877 PP/10 Added dbmjz lookup type.
1879 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1881 PP/12 MAIL args handles TAB as well as SP, for better interop with
1882 non-compliant senders.
1883 Analysis and variant patch by Todd Lyons.
1885 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1886 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1887 Patch from Dirk Mueller <dmueller@suse.com>
1889 PP/13 tls_peerdn now print-escaped for spool files.
1890 Observed some $tls_peerdn in wild which contained \n, which resulted
1891 in spool file corruption.
1893 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1894 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1895 or write after TLS renegotiation, which otherwise led to messages
1898 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1899 as a tracking header (ie: a signed header comes before the signature).
1900 Patch from Wolfgang Breyha.
1902 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1903 comma-sep list; embedded commas doubled.
1905 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1907 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1909 Report and patch from Dmitry Banschikov.
1911 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1912 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1913 failures. We appear to now support TLS1.1+ with Exim.
1915 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1916 lets Exim select keys and certificates based upon TLS SNI from client.
1917 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1918 before an outbound SMTP session. New log_selector, +tls_sni.
1920 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1921 NULL dereference. Report and patch from Alun Jones.
1923 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1924 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1925 Not seeing resolver debug output on NetBSD, but suspect this is a
1926 resolver implementation change.
1928 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1929 Left warnings. Added "eximon gdb" invocation mode.
1931 PP/21 Defaulting "accept_8bitmime" to true, not false.
1933 PP/22 Added -bw for inetd wait mode support.
1935 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1936 locate the relevant includes and libraries. Made this the default.
1938 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1939 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1941 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1942 This may cause build issues on older platforms.
1944 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1945 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1946 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1947 Added SNI support via GnuTLS too.
1948 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1950 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1952 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1953 Applied second patch from Janne, implementing suggestion to default
1954 multiple-strings-in-record handling to match SPF spec.
1956 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1958 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1959 read-only, out of scope).
1960 Patch from Wolfgang Breyha, report from Stuart Northfield.
1962 PP/29 Fix three issues highlighted by clang analyser static analysis.
1963 Only crash-plausible issue would require the Cambridge-specific
1964 iplookup router and a misconfiguration.
1965 Report from Marcin Mirosław.
1967 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1969 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1970 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1971 As part of this, removing so much warning spew let me fix some minor
1972 real issues in debug logging.
1974 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1975 assignment on my part. Fixed.
1977 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1978 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1979 Janne Snabb (who went above and beyond: thank you).
1981 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1982 string otherwise requires a connection and a bunch more work and it's
1983 relatively easy to get wrong. Should also expose TLS library linkage
1986 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1987 64-bit ${eval} (JH/03).
1989 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1990 GNU libc to support some of the 64-bit stuff, should not lead to
1991 conflicts. Defined before os.h is pulled in, so if a given platform
1992 needs to override this, it can.
1994 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1995 protection layer was required, which is not implemented.
1996 Bugzilla 1254, patch from Wolfgang Breyha.
1998 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1999 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2000 tls_dhparam take prime identifiers. Also unbreak combination of
2001 OpenSSL+DH_params+TLSSNI.
2003 PP/39 Disable SSLv2 by default in OpenSSL support.
2009 PP/01 Solaris build fix for Oracle's LDAP libraries.
2010 Bugzilla 1109, patch from Stephen Usher.
2012 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2014 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2017 TF/02 Fix a couple more cases where we did not log the error message
2018 when unlink() failed. See also change 4.74-TF/03.
2020 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2021 lock up or crash if it happened to be inside a call to libc when it
2022 got a SIGUSR1 from exiwhat.
2024 The SIGUSR1 handler appends the current process status to the process
2025 log which is later printed by exiwhat. It used to use the general
2026 purpose logging code to do this, but several functions it calls are
2027 not safe for signals.
2029 The new output code in the SIGUSR1 handler is specific to the process
2030 log, and simple enough that it's easy to inspect for signal safety.
2031 Removing some special cases also simplifies the general logging code.
2032 Removing the spurious timestamps from the process log simplifies
2035 TF/04 Improved ratelimit ACL condition.
2037 The /noupdate option has been deprecated in favour of /readonly which
2038 has clearer semantics. The /leaky, /strict, and /readonly update modes
2039 are mutually exclusive. The update mode is no longer included in the
2040 database key; it just determines when the database is updated. (This
2041 means that when you upgrade Exim will forget old rate measurements.)
2043 Exim now checks that the per_* options are used with an update mode that
2044 makes sense for the current ACL. For example, when Exim is processing a
2045 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2046 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2047 must specify per_mail/readonly. If you omit the update mode it defaults to
2048 /leaky where that makes sense (as before) or /readonly where required.
2050 The /noupdate option is now undocumented but still supported for
2051 backwards compatibility. It is equivalent to /readonly except that in
2052 ACLs where /readonly is required you may specify /leaky/noupdate or
2053 /strict/noupdate which are treated the same as /readonly.
2055 A useful new feature is the /count= option. This is a generalization
2056 of the per_byte option, so that you can measure the throughput of other
2057 aggregate values. For example, the per_byte option is now equivalent
2058 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2060 The per_rcpt option has been generalized using the /count= mechanism
2061 (though it's more complicated than the per_byte equivalence). When it is
2062 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2063 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2064 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2065 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2066 Note that using per_rcpt with a non-readonly update mode in more than
2067 one ACL will cause the recipients to be double-counted. (The per_mail
2068 and per_byte options don't have this problem.)
2070 The handling of very low rates has changed slightly. If the computed rate
2071 is less than the event's count (usually one) then this event is the first
2072 after a long gap. In this case the rate is set to the same as this event's
2073 count, so that the first message of a spam run is counted properly.
2075 The major new feature is a mechanism for counting the rate of unique
2076 events. The new per_addr option counts the number of different
2077 recipients that someone has sent messages to in the last time period. It
2078 behaves like per_rcpt if all the recipient addresses are different, but
2079 duplicate recipient addresses do not increase the measured rate. Like
2080 the /count= option this is a general mechanism, so the per_addr option
2081 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2082 example, measure the rate that a client uses different sender addresses
2083 with the options per_mail/unique=$sender_address. There are further
2084 details in the main documentation.
2086 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2088 TF/06 Removed a few PCRE remnants.
2090 TF/07 Automatically extract Exim's version number from tags in the git
2091 repository when doing development or release builds.
2093 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2094 Bugzilla 879. Patch from Paul Fisher.
2096 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2097 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2100 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2102 PP/05 Variable $av_failed, true if the AV scanner deferred.
2103 Bugzilla 1078. Patch from John Horne.
2105 PP/06 Stop make process more reliably on build failure.
2106 Bugzilla 1087. Patch from Heiko Schlittermann.
2108 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2109 Bugzilla 1089. Patch from Heiko Schlittermann.
2111 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2112 Bugzilla 1131. Patch from Holger Weiß.
2114 PP/09 Handle IPv6 addresses with SPF.
2115 Bugzilla 860. Patch from Wolfgang Breyha.
2117 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2119 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2122 PP/11 match_* no longer expand right-hand-side by default.
2123 New compile-time build option, EXPAND_LISTMATCH_RHS.
2124 New expansion conditions, "inlist", "inlisti".
2126 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2128 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2130 PP/14 fix log_write() format string regression from TF/03.
2131 Bugzilla 1152. Patch from Dmitry Isaikin.
2137 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2139 PP/02 Harmonised TLS library version reporting; only show if debugging.
2140 Layout now matches that introduced for other libraries in 4.74 PP/03.
2142 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2144 PP/04 New "dns_use_edns0" global option.
2146 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2149 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2150 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2152 TK/01 Updated PolarSSL code to 0.14.2.
2153 Bugzilla 1097. Patch from Andreas Metzler.
2155 PP/07 Catch divide-by-zero in ${eval:...}.
2156 Fixes bugzilla 1102.
2158 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2161 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2162 format-string attack -- SECURITY: remote arbitrary code execution.
2164 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2165 time unintentionally subject to list matching rules, letting the header
2166 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2167 arbitrary string expansion). This allowed for information disclosure.
2169 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2170 INT_MIN/-1 -- value coerced to INT_MAX.
2176 NM/01 Workaround for PCRE version dependency in version reporting
2179 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2180 This fixes portability to compilers other than gcc, notably
2181 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2183 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2184 makefiles for portability to HP-UX and POSIX correctness.
2186 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2187 Also via indented variable definition in the Makefile.
2188 (Debugging by Oliver Heesakkers).
2190 PP/02 Restore caching of spamd results with expanded spamd_address.
2191 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2193 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2194 Improves build reliability. Fix from: Frank Elsner
2196 NM/02 Fix wide character breakage in the rfc2047 coding
2197 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2199 NM/03 Allow underscore in dnslist lookups
2200 Fixes bug 1026. Patch from Graeme Fowler
2202 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2203 Code patches from Adam Ciarcinski of NetBSD.
2205 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2208 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2209 is logged, to avoid truncation. Patch from John Horne.
2211 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2212 Patch from Jakob Hirsch.
2214 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2215 SQL string expansion failure details.
2216 Patch from Andrey Oktyabrski.
2218 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2219 Patch from Simon Arlott.
2221 PP/09 New lookups functionality failed to compile on old gcc which rejects
2222 extern declarations in function scope.
2223 Patch from Oliver Fleischmann
2225 PP/10 Use sig_atomic_t for flags set from signal handlers.
2226 Check getgroups() return and improve debugging.
2227 Fixed developed for diagnosis in bug 927 (which turned out to be
2230 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2231 Patch from Mark Zealey.
2233 PP/12 Bugzilla 1056: Improved spamd server selection.
2234 Patch from Mark Zealey.
2236 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2237 Based on patch from Heiko Schlittermann.
2239 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2240 Patch from Uwe Doering, sign-off by Michael Haardt.
2242 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2243 variable declaration deep within a block. Bug and patch from
2246 PP/15 lookups-Makefile IRIX compatibility coercion.
2248 PP/16 Make DISABLE_DKIM build knob functional.
2250 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2251 Patch by Simon Arlott
2253 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2254 variable argument macros. Our copy now differs from upstream.
2260 TF/01 Failure to get a lock on a hints database can have serious
2261 consequences so log it to the panic log.
2263 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2264 controlled using the smtp_confirmation log selector.
2266 TF/03 Include the error message when we fail to unlink a spool file.
2268 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2269 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2270 for maintaining out-of-tree patches for some time.
2272 PP/01 Bugzilla 139: Documentation and portability issues.
2273 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2274 Handle per-OS dynamic-module compilation flags.
2276 PP/02 Let /dev/null have normal permissions.
2277 The 4.73 fixes were a little too stringent and complained about the
2278 permissions on /dev/null. Exempt it from some checks.
2279 Reported by Andreas M. Kirchwitz.
2281 PP/03 Report version information for many libraries, including
2282 Exim version information for dynamically loaded libraries. Created
2283 version.h, now support a version extension string for distributors
2284 who patch heavily. Dynamic module ABI change.
2286 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2287 privilege escalation vulnerability whereby the Exim run-time user
2288 can cause root to append content of the attacker's choosing to
2291 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2294 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2295 If dropping privileges for untrusted macros, we disabled normal logging
2296 on the basis that it would fail; for the Exim run-time user, this is not
2297 the case, and it resulted in successful deliveries going unlogged.
2298 Fixed. Reported by Andreas Metzler.
2304 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2305 only prepend for the Resent-* case. Fixes regression introduced in
2306 Exim 4.70 by NM/22 for Bugzilla 607.
2308 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2309 increasing numbers of administrators be bitten by this.
2311 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2313 PP/03 Bugzilla 994: added openssl_options main configuration option.
2315 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2317 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2319 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2321 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2322 without a peer certificate, leading to a segfault because of an
2323 assumption that peers always have certificates. Be a little more
2324 paranoid. Problem reported by Martin Tscholak.
2326 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2327 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2328 NB: ClamAV planning to remove STREAM in "middle of 2010".
2329 CL also introduces -bmalware, various -d+acl logging additions and
2330 more caution in buffer sizes.
2332 PP/09 Implemented reverse_ip expansion operator.
2334 PP/10 Bugzilla 937: provide a "debug" ACL control.
2336 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2338 PP/12 Bugzilla 973: Implement --version.
2340 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2342 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2344 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2346 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2347 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2348 ignore trailing whitespace.
2350 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2352 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2355 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2356 Notification from Dr Andrew Aitchison.
2358 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2359 ExtendedDetectionInfo response format.
2360 Notification from John Horne.
2362 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2365 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2366 XSL and documented dependency on system catalogs, with examples of how
2369 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2372 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2373 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2374 configuration file which is writeable by the Exim user or group.
2376 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2377 of configuration files to cover files specified with the -C option if
2378 they are going to be used with root privileges, not just the default
2381 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2382 option (effectively making it always true).
2384 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2385 files to be used while preserving root privileges.
2387 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2388 that rogue child processes cannot use them.
2390 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2391 run-time user, instead of root.
2393 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2394 Exim run-time user without dropping privileges.
2396 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2397 result string, instead of calling string_vformat() twice with the same
2400 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2401 for other users. Others should always drop root privileges if they use
2402 -C on the command line, even for a whitelisted configure file.
2404 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2406 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2412 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2413 $data_path, and $header_path variables; fixed documentation bugs and
2416 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2417 exipick to access non-standard spools, including the "frozen" queue
2420 NM/01 Bugzilla 965: Support mysql stored procedures.
2421 Patch from Alain Williams
2423 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2425 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2426 Patch from Andreas Metzler
2428 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2429 Patch from Kirill Miazine
2431 NM/05 Bugzilla 671: Added umask to procmail example.
2433 JJ/03 installed exipick 20100323.0, fixing doc bug
2435 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2436 directory. Notification and patch from Dan Rosenberg.
2438 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2440 TK/02 Improve log output when DKIM signing operation fails.
2442 MH/01 Treat the transport option dkim_domain as a colon separated
2443 list, not as a single string, and sign the message with each element,
2444 omitting multiple occurences of the same signer.
2446 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2447 Bugzilla 985, 986. Patch by Simon Arlott
2449 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2450 Patch by Simon Arlott
2452 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2453 MBX locking. Notification from Dan Rosenberg.
2459 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2461 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2463 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2465 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2467 NM/04 Bugzilla 915: Flag broken perl installation during build.
2473 TK/01 Added patch by Johannes Berg that expands the main option
2474 "spamd_address" if it starts with a dollar sign.
2476 TK/02 Write list of recipients to X-Envelope-Sender header when building
2477 the mbox-format spool file for content scanning (suggested by Jakob
2480 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2481 (http://www.dcc-servers.net/) support via dccifd. Activated by
2482 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2484 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2485 by Mark Daniel Reidel <mr@df.eu>.
2487 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2488 When building exim an external PCRE library is now needed -
2489 PCRE is a system library on the majority of modern systems.
2490 See entry on PCRE_LIBS in EDITME file.
2492 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2493 conversation. Added nologin parameter to request.
2494 Patch contributed by Kirill Miazine.
2496 TF/01 Do not log submission mode rewrites if they do not change the address.
2498 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2500 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2501 log files in place. Contributed by Roberto Lima.
2503 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2505 TF/03 Bugzilla 615: When checking the local_parts router precondition
2506 after a local_part_suffix or local_part_prefix option, Exim now
2507 does not use the address's named list lookup cache, since this
2508 contains cached lookups for the whole local part.
2510 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2511 Robert Millan. Documentation is in experimental-spec.txt.
2513 TF/04 Bugzilla 668: Fix parallel build (make -j).
2515 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2517 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2518 Patch provided by Jan Srzednicki.
2520 TF/05 Leading white space used to be stripped from $spam_report which
2521 wrecked the formatting. Now it is preserved.
2523 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2524 that they are available at delivery time.
2526 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2528 TF/08 TLS error reporting now respects the incoming_interface and
2529 incoming_port log selectors.
2531 TF/09 Produce a more useful error message if an SMTP transport's hosts
2532 setting expands to an empty string.
2534 NM/06 Bugzilla 744: EXPN did not work under TLS.
2535 Patch provided by Phil Pennock.
2537 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2538 Patch provided by Richard Godbee.
2540 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2541 acl_smtp_notquit, added index entry.
2543 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2544 Patch provided by Eugene Bujak.
2546 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2547 accept(). Patch provided by Maxim Dounin.
2549 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2550 Patch provided by Phil Pennock.
2552 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2554 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2555 Patch provided by Brad "anomie" Jorsch.
2557 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2558 Patch provided by Dean Brooks.
2560 TK/05 Add native DKIM support (does not depend on external libraries).
2562 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2563 Patch provided by Graeme Fowler.
2565 NM/16 Bugzilla 851: Documentation example syntax fix.
2567 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2569 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2572 NM/19 Bugzilla 745: TLS version reporting.
2573 Patch provided by Phil Pennock.
2575 NM/20 Bugzilla 167: bool: condition support.
2576 Patch provided by Phil Pennock.
2578 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2579 clients. Patch provided by Phil Pennock.
2581 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2582 Patch provided by Brad "anomie" Jorsch.
2584 NM/23 Bugzilla 687: Fix misparses in eximstats.
2585 Patch provided by Heiko Schlittermann.
2587 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2588 Patch provided by Heiko Schlittermann.
2590 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2591 plus update to original patch.
2593 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2595 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2596 Patch provided by David Brownlee.
2598 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2600 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2602 NM/30 Bugzilla 888: TLS documentation bugfixes.
2604 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2606 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2607 Unlike the original bugzilla I have changed all shell scripts in src tree.
2609 NM/33 Bugzilla 898: Transport filter timeout fix.
2610 Patch by Todd Rinaldo.
2612 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2613 Patch by Serge Demonchaux.
2615 NM/35 Bugzilla 39: Base64 decode bug fixes.
2616 Patch by Jakob Hirsch.
2618 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2620 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2622 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2624 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2630 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2631 ALT-N's libdkim that I have put here:
2632 http://duncanthrax.net/exim-experimental/
2634 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2635 were called 'true' and it seems that C99 defines that as a reserved
2636 keyword to be used with 'bool' variable types. That means you could
2637 not include C99-style headers which use bools without triggering
2638 build errors in sieve.c.
2640 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2641 as mailq or other aliases. Changed the --help handling significantly
2642 to do whats expected. exim_usage() emits usage/help information.
2644 SC/01 Added the -bylocaldomain option to eximstats.
2646 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2648 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2650 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2656 PH/01 Another patch from the Sieve maintainer.
2658 PH/02 When an IPv6 address is converted to a string for single-key lookup
2659 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2660 dots are used instead of colons so that keys in lsearch files need not
2661 contain colons. This was done some time before quoting was made available
2662 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2663 (notated using the quote facility) so as to distinguish them from IPv4
2664 keys. This meant that lookups for IP addresses in host lists did not work
2665 for iplsearch lookups.
2667 This has been fixed by arranging for IPv6 addresses to be expressed with
2668 colons if the lookup type is iplsearch. This is not incompatible, because
2669 previously such lookups could never work.
2671 The situation is now rather anomalous, since one *can* have colons in
2672 ordinary lsearch keys. However, making the change in all cases is
2673 incompatible and would probably break a number of configurations.
2675 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2678 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2679 conversion specification without a maximum field width, thereby enabling
2680 a rogue spamd server to cause a buffer overflow. While nobody in their
2681 right mind would setup Exim to query an untrusted spamd server, an
2682 attacker that gains access to a server running spamd could potentially
2683 exploit this vulnerability to run arbitrary code as the Exim user.
2685 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2686 $primary_hostname instead of what libspf2 thinks the hosts name is.
2688 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2689 a directory entry by the name of the lookup key. Previously, if a
2690 symlink pointed to a non-existing file or a file in a directory that
2691 Exim lacked permissions to read, a lookup for a key matching that
2692 symlink would fail. Now it is enough that a matching directory entry
2693 exists, symlink or not. (Bugzilla 503.)
2695 PH/03 The body_linecount and body_zerocount variables are now exported in the
2698 PH/04 Added the $dnslist_matched variable.
2700 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2701 This means they are set thereafter only if the connection becomes
2704 PH/06 Added the client_condition to authenticators so that some can be skipped
2705 by clients under certain conditions.
2707 PH/07 The error message for a badly-placed control=no_multiline_responses left
2708 "_responses" off the end of the name.
2710 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2712 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2713 (without spaces) instead of just copying the configuration text.
2715 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2717 PH/11 Added $max_received_linelength.
2719 PH/12 Added +ignore_defer and +include_defer to host lists.
2721 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2722 way in which PCRE > 7.0 is built.
2724 PH/14 Implemented queue_only_load_latch.
2726 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2727 MAIL command. The effect was to mangle the value on 64-bit systems.
2729 PH/16 Another patch from the Sieve maintainer.
2731 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2733 PH/18 If a system quota error occurred while trying to create the file for
2734 a maildir delivery, the message "Mailbox is full" was not appended to the
2735 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2736 only to a quota excession during the actual writing of the file.
2738 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2739 characters?) which causes problems in log lines. The DN values are now
2740 passed through string_printing() before being added to log lines.
2742 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2743 and InterBase are left for another time.)
2745 PH/21 Added message_body_newlines option.
2747 PH/22 Guard against possible overflow in moan_check_errorcopy().
2749 PH/23 POSIX allows open() to be a macro; guard against that.
2751 PH/24 If the recipient of an error message contained an @ in the local part
2752 (suitably quoted, of course), incorrect values were put in $domain and
2753 $local_part during the evaluation of errors_copy.
2759 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2760 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2763 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2764 issue a MAIL command.
2766 PH/02 In an ACL statement such as
2768 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2770 if a client was not listed at all, or was listed with a value other than
2771 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2772 the condition was not true (as it should be), so access was not denied.
2773 The bug was that the ! inversion was incorrectly passed on to the second
2774 item. This has been fixed.
2776 PH/03 Added additional dnslists conditions == and =& which are different from
2777 = and & when the dns lookup returns more than one IP address.
2779 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2780 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2782 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2783 FSYNC, which compiles an option called disable_fsync that allows for
2784 bypassing fsync(). The documentation is heavily laced with warnings.
2786 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2788 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2789 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2790 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2791 including adding "make clean"; (3) Added -fPIC when compiling the test
2792 dynamically loaded module, to get rid of a warning.
2794 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2795 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2796 = 0s. The bug is otherwise harmless.
2798 PH/07 There was a bug in the dovecot authenticator such that the value of
2799 $auth1 could be overwritten, and so not correctly preserved, after a
2800 successful authentication. This usually meant that the value preserved by
2801 the server_setid option was incorrect.
2803 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2805 PH/09 Installed PCRE release 7.0.
2807 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2808 run for batched SMTP input. It is now run at the start of every message
2809 in the batch. While fixing this I discovered that the process information
2810 (output by running exiwhat) was not always getting set for -bs and -bS
2811 input. This is fixed, and it now also says "batched" for BSMTP.
2813 PH/11 Added control=no_pipelining.
2815 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2816 patch, slightly modified), and move the expansion of helo_data till after
2817 the connection is made in the smtp transport (so it can use these
2820 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2822 PH/14 Added log_selector = +pid.
2824 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2826 PH/16 Add ${if forany and ${if forall.
2828 PH/17 Added dsn_from option to vary the From: line in DSNs.
2830 PH/18 Flush SMTP output before performing a callout, unless control =
2831 no_callout_flush is set.
2833 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2834 was true (the default) a successful delivery failed to delete the retry
2835 item, thus causing premature timeout of the address. The bug is now
2838 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2840 PH/21 Long custom messages for fakedefer and fakereject are now split up
2841 into multiline responses in the same way that messages for "deny" and
2842 other ACL rejections are.
2844 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2845 with slight modification.
2847 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2848 draft, changing the syntax and factoring some duplicate code".
2850 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2851 for deliveries of the second and subsequent messages over the same SMTP
2854 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2855 ${reduce, with only minor "tidies".
2857 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2859 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2860 expansion side effects.
2862 PH/27 When a message times out after an over-quota error from an Exim-imposed
2863 quota, the bounce message says "mailbox is full". This message was not
2864 being given when it was a system quota that was exceeded. It now should
2867 MH/03 Made $recipients available in local_scan(). local_scan() already has
2868 better access to the recipient list through recipients_list[], but
2869 $recipients can be useful in postmaster-provided expansion strings.
2871 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2872 in the case of a MAIL command with additional options following the
2873 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2874 were accidentally chopped off.
2876 PH/29 SMTP synchronization checks are implemented when a command is read -
2877 there is a check that no more input is waiting when there shouldn't be
2878 any. However, for some commands, a delay in an ACL can mean that it is
2879 some time before the response is written. In this time, more input might
2880 arrive, invalidly. So now there are extra checks after an ACL has run for
2881 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2882 pipelining has not been advertised.
2884 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2886 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2887 returned in the SMTP response when smtp_return_error_details was false.
2888 This has been fixed.
2890 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2891 instead of the C I/O that was originally supplied, because problems were
2892 reported on Solaris.
2894 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2895 Exim which did not show up earlier: it was assuming that a call to
2896 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2897 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2898 was a macro that became an assignment, so it seemed to work. This has
2899 changed to a proper function call with a void return, hence the compile
2900 error. Exim's code has been fixed.
2902 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2905 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2907 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2909 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2910 because it turns out to be too big or there is a timeout) there is no
2911 "Completed" line in the log. When some messages of this type were
2912 selected by exigrep, they were listed as "not completed". Others were
2913 picked up by some special patterns. I have improved the selection
2914 criteria to be more general.
2916 PH/38 The host_find_failed option in the manualroute router can now be set
2917 to "ignore", to completely ignore a host whose IP address cannot be
2918 found. If all hosts are ignored, the behaviour is controlled by the new
2919 host_all_ignored option.
2921 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2922 homing or because of multiple MX records with /mx) generated more than
2923 one IP address, and the following item turned out to be the local host,
2924 all the secondary addresses of the first item were incorrectly removed
2925 from the list, along with the local host and any following hosts (which
2926 is what is supposed to happen).
2928 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2929 whoever called Exim into the -H file. In the case of the daemon it was
2930 behaving confusingly. When first started, it used values for whoever
2931 started the daemon, but after a SIGHUP it used the Exim user (because it
2932 calls itself on a restart). I have changed the code so that it now always
2935 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2936 message are rejected with the same error (e.g. no authentication or bad
2937 sender address), and a DATA command is nevertheless sent (as can happen
2938 with PIPELINING or a stupid MUA), the error message that was given to the
2939 RCPT commands is included in the rejection of the DATA command. This is
2940 intended to be helpful for MUAs that show only the final error to their
2943 PH/42 Another patch from the Sieve maintainer.
2945 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2946 Eximstats - Fixed some broken HTML links and added missing column headers
2948 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2949 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2951 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2953 PH/43 Yet another patch from the Sieve maintainer.
2955 PH/44 I found a way to check for a TCP/IP connection going away before sending
2956 the response to the final '.' that terminates a message, but only in the
2957 case where the client has not sent further data following the '.'
2958 (unfortunately, this is allowed). However, in many cases there won't be
2959 any further data because there won't be any more messages to send. A call
2960 to select() can be used: if it shows that the input is "ready", there is
2961 either input waiting, or the socket has been closed. An attempt to read
2962 the next input character can distinguish the two cases. Previously, Exim
2963 would have sent an OK response which the client would never have see.
2964 This could lead to message repetition. This fix should cure that, at
2965 least in a lot of common cases.
2967 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2968 advertised in response to EHLO.
2974 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2975 fixed by 4.65/MH/01 (is this a record?) are fixed:
2977 (i) An empty string was always treated as zero by the numeric comparison
2978 operators. This behaviour has been restored.
2980 (ii) It is documented that the numeric comparison operators always treat
2981 their arguments as decimal numbers. This was broken in that numbers
2982 starting with 0 were being interpreted as octal.
2984 While fixing these problems I realized that there was another issue that
2985 hadn't been noticed. Values of message_size_limit (both the global option
2986 and the transport option) were treated as octal if they started with 0.
2987 The documentation was vague. These values are now always treated as
2988 decimal, and I will make that clear in the documentation.
2994 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2995 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2998 MH/01 Don't check that the operands of numeric comparison operators are
2999 integers when their expansion is in "skipping" mode (fixes bug
3000 introduced by 4.64-PH/07).
3002 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3003 child addresses, Exim now panics and dies. Previously, because the count
3004 is held in a short int, deliveries were likely to be lost. As such a
3005 large number of recipients for a single message is ridiculous
3006 (performance will be very, very poor), I have chosen to impose a limit
3007 rather than extend the field.
3013 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3014 leftover -K file (the existence of which was triggered by #402).
3015 While we were at it, introduced process PID as part of the -K
3016 filename. This should rule out race conditions when creating
3019 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3020 processing considerably. Previous code took too long for large mails,
3021 triggering a timeout which in turn triggers #401.
3023 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3024 in the DK code in transports.c. sendfile() is not really portable,
3025 hence the _LINUX specificness.
3027 TF/01 In the add_headers option to the mail command in an Exim filter,
3028 there was a bug that Exim would claim a syntax error in any
3029 header after the first one which had an odd number of characters
3032 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3033 callout verification, Exim cached a "reject" for the entire domain. This
3034 is correct for most verifications, but it is not correct for a recipient
3035 verification with use_sender or use_postmaster set, because in that case
3036 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3037 case of MAIL FROM:<> rejection from other early rejections (e.g.
3038 rejection of HELO). When verifying a recipient using a non-null MAIL
3039 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3040 Whatever the result of the callout, the value of the domain cache is
3041 left unchanged (for any other kind of callout, getting as far as trying
3042 RCPT means that the domain itself is ok).
3044 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3047 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3048 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3049 session, the daemon ignores ECONNECT errors and logs others; it now
3050 ignores EPIPE as well.
3052 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3053 (quoted-printable decoding).
3055 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3056 later the small subsequent patch to fix an introduced bug.
3058 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3060 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3062 PH/08 An error is now given if message_size_limit is specified negative.
3064 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3065 to be given (somewhat) arbitrary names.
3067 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3070 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3071 miscellaneous code fixes
3073 PH/10 Added the log_reject_target ACL modifier to specify where to log
3076 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3077 hostname. This is wrong, because it relates to the incoming message (and
3078 probably the interface on which it is arriving) and not to the outgoing
3079 callout (which could be using a different interface). This has been
3080 changed to use the value of the helo_data option from the smtp transport
3081 instead - this is what is used when a message is actually being sent. If
3082 there is no remote transport (possible with a router that sets up host
3083 addresses), $smtp_active_hostname is used.
3085 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3086 tweaks were necessary in order to get it to work (see also 21 below):
3087 (a) The code assumed that strncpy() returns a negative number on buffer
3088 overflow, which isn't the case. Replaced with Exim's string_format()
3090 (b) There were several signed/unsigned issues. I just did the minimum
3091 hacking in of casts. There is scope for a larger refactoring.
3092 (c) The code used strcasecmp() which is not a standard C function.
3093 Replaced with Exim's strcmpic() function.
3094 (d) The code set only $1; it now sets $auth1 as well.
3095 (e) A simple test gave the error "authentication client didn't specify
3096 service in request". It would seem that Dovecot has changed its
3097 interface. Fortunately there's a specification; I followed it and
3098 changed what the client sends and it appears to be working now.
3100 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3103 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3104 address A is aliased to B and C, where B exists and C does not. Without
3105 -v the output is "A verified" because verification stops after a
3106 successful redirection if more than one address is generated. However,
3107 with -v the child addresses are also verified. Exim was outputting "A
3108 failed to verify" and then showing the successful verification for C,
3109 with its parentage. It now outputs "B failed to verify", showing B's
3110 parentage before showing the successful verification of C.
3112 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3113 look up a TXT record in a specific list after matching in a combined
3116 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3117 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3118 they consult the DNS. I had assumed they would set it the way they
3119 wanted; and indeed my experiments on Linux seem to show that in some
3120 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3121 To be on the safe side, however, I have now made the interface to
3122 host_find_byname() similar to host_find_bydns(), with an argument
3123 containing the DNS resolver options. The host_find_byname() function now
3124 sets these options at its start, just as host_find_bydns() does. The smtp
3125 transport options dns_qualify_single and dns_search_parents are passed to
3126 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3127 of host_find_byname() use the default settings of RES_DEFNAMES
3128 (qualify_single) but not RES_DNSRCH (search_parents).
3130 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3131 spool_read_header() do less string testing, by means of a preliminary
3132 switch on the second character of optional "-foo" lines. (This is
3133 overdue, caused by the large number of possibilities that now exist.
3134 Originally there were few.) While I was there, I also converted the
3135 str(n)cmp tests so they don't re-test the leading "-" and the first
3136 character, in the hope this might squeeze out yet more improvement.
3138 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3139 flag allowing group syntax was set by the header_syntax check but not
3140 turned off, possible causing trouble later; (2) The flag was not being
3141 set at all for the header_verify test, causing "group"-style headers to
3142 be rejected. I have now set it in this case, and also caused header_
3143 verify to ignore an empty address taken from a group. While doing this, I
3144 came across some other cases where the code for allowing group syntax
3145 while scanning a header line wasn't quite right (mostly, not resetting
3146 the flag correctly in the right place). These bugs could have caused
3147 trouble for malformed header lines. I hope it is now all correct.
3149 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3150 with the "reply" argument non-NULL. The code, however (which originally
3151 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3152 but it didn't always do it. This confused somebody who was copying the
3153 code for some other use. I have removed all the tests.
3155 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3156 feature that was used to support insecure browsers during the U.S. crypto
3157 embargo. It requires special client support, and Exim is probably the
3158 only MTA that supported it -- and would never use it because real RSA is
3159 always available. This code has been removed, because it had the bad
3160 effect of slowing Exim down by computing (never used) parameters for the
3161 RSA_EXPORT functionality.
3163 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3164 authenticator to fail if there's a tab character in the incoming data
3165 (there should never be unless someone is messing about, as it's supposed
3166 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3167 if the connection is using TLS or if the remote IP is the same as the
3168 local IP, and the "valid-client-cert option" if a client certificate has
3171 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3172 authenticators. This can be used for authorization after authentication
3173 succeeds. (In the case of plaintext, it servers for both authentication
3176 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3177 if any retry times were supplied.
3179 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3180 connection, where there is no client IP address to check. In this
3181 situation, the verify now always succeeds.
3183 PH/25 Applied John Jetmore's -Mset patch.
3185 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3187 PH/27 In a string expansion for a processed (not raw) header when multiple
3188 headers of the same name were present, leading whitespace was being
3189 removed from all of them, but trailing whitespace was being removed only
3190 from the last one. Now trailing whitespace is removed from each header
3191 before concatenation. Completely empty headers in a concatenation (as
3192 before) are ignored.
3194 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3195 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3197 PH/29 [Removed. This was a change that I later backed out, and forgot to
3198 correct the ChangeLog entry (that I had efficiently created) before
3199 committing the later change.]
3201 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3202 address errors (4xx response to RCPT) over the same connection as other
3203 messages routed to the same hosts. Such deliveries are always "forced",
3204 so retry times are not inspected. This resulted in far too many retries
3205 for the affected addresses. The effect occurred only when there were more
3206 hosts than the hosts_max_try setting in the smtp transport when it had
3207 the 4xx errors. Those hosts that it had tried were not added to the list
3208 of hosts for which the message was waiting, so if all were tried, there
3209 was no problem. Two fixes have been applied:
3211 (i) If there are any address or message errors in an SMTP delivery, none
3212 of the hosts (tried or untried) are now added to the list of hosts
3213 for which the message is waiting, so the message should not be a
3214 candidate for sending over the same connection that was used for a
3215 successful delivery of some other message. This seems entirely
3216 reasonable: after all the message is NOT "waiting for some host".
3217 This is so "obvious" that I'm not sure why it wasn't done
3218 previously. Hope I haven't missed anything, but it can't do any
3219 harm, as the worst effect is to miss an optimization.
3221 (ii) If, despite (i), such a delivery is accidentally attempted, the
3222 routing retry time is respected, so at least it doesn't keep
3223 hammering the server.
3225 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3226 in ${readsocket because some servers need this prod.
3228 PH/32 Added some extra debug output when updating a wait-xxx database.
3230 PH/33 The hint "could be header name not terminated by colon", which has been
3231 given for certain expansion errors for a long time, was not being given
3232 for the ${if def:h_colon_omitted{... case.
3234 PH/34 The spec says: "With one important exception, whenever a domain list is
3235 being scanned, $domain contains the subject domain." There was at least
3236 one case where this was not true.
3238 PH/35 The error "getsockname() failed: connection reset by peer" was being
3239 written to the panic log as well as the main log, but it isn't really
3240 panic-worthy as it just means the connection died rather early on. I have
3241 removed the panic log writing for the ECONNRESET error when getsockname()
3244 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3245 runs only) independently of the message's sender address. This meant
3246 that, if the 4xx error was in fact related to the sender, a different
3247 message to the same recipient with a different sender could confuse
3248 things. In particular, this can happen when sending to a greylisting
3249 server, but other circumstances could also provoke similar problems.
3250 I have changed the default so that the retry time for these errors is now
3251 based a combination of the sender and recipient addresses. This change
3252 can be overridden by setting address_retry_include_sender=false in the
3255 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3256 remote server are returned as part of bounce messages. This was not
3257 happening for LMTP over a pipe (the lmtp transport), but now it is the
3258 same for both kinds of LMTP.
3260 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3261 in header lines that were in fact CNAMEs. This is no longer the case.
3263 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3264 and queue runs started by the daemon processed all messages. This has
3265 been fixed so that -R and -S can now usefully be given with -q<time>.
3267 PH/40 Import PCRE release 6.7 (fixes some bugs).
3269 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3271 PH/42 Give an error if -q is specified more than once.
3273 PH/43 Renamed the variables $interface_address and $interface_port as
3274 $received_ip_address and $received_port, to make it clear that these
3275 values apply to message reception, and not to the outgoing interface when
3276 a message is delivered. (The old names remain recognized, of course.)
3278 PH/44 There was no timeout on the connect() call when using a Unix domain
3279 socket in the ${readsocket expansion. There now is.
3281 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3282 be meaningful with "accept".
3284 SC/01 Eximstats V1.43
3285 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3287 SC/02 Eximstats V1.44
3288 Use a glob alias rather than an array ref in the generated
3289 parser. This improves both readability and performance.
3291 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3292 Collect SpamAssassin and rejection statistics.
3293 Don't display local sender or destination tables unless
3294 there is data to show.
3295 Added average volumes into the top table text output.
3297 SC/04 Eximstats V1.46
3298 Collect data on the number of addresses (recipients)
3299 as well as the number of messages.
3301 SC/05 Eximstats V1.47
3302 Added 'Message too big' to the list of mail rejection
3303 reasons (thanks to Marco Gaiarin).
3305 SC/06 Eximstats V1.48
3306 Mainlog lines which have GMT offsets and are too short to
3307 have a flag are now skipped.
3309 SC/07 Eximstats V1.49 (Alain Williams)
3310 Added the -emptyok flag.
3312 SC/08 Eximstats V1.50
3313 Fixes for obtaining the IP address from reject messages.
3315 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3316 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3317 whitespace changes from 4.64-PH/27
3319 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3322 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3323 are found, allow negative numbers in numeric criteria)
3325 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3327 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3328 to match changes made in 4.64-PH/43
3330 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3332 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3333 "domain" is actually an IP address, and force a failure. This locks out
3334 those revolvers/nameservers that support "A-for-A" lookups, in
3335 contravention of the specifications.
3337 PH/48 When a host name was looked up from an IP address, and the subsequent
3338 forward lookup of the name timed out, the host name was left in
3339 $sender_host_name, contrary to the specification.
3341 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3342 restricted to single-key lookups, Exim was not diagnosing an error if
3343 * or *@ was used with a query-style lookup.
3345 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3347 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3348 long ago, but noone interested enough thought of it. Let's just say that
3349 the "1.1" means that there are some new functions that weren't there at
3350 some point in the past.
3352 PH/51 Error processing for expansion failure of helo_data from an smtp
3353 transport during callout processing was broken.
3355 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3356 tested/used via the -bh/-bhc/-bs options.
3358 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3359 bug, fixed in subsequent PCRE releases).
3361 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3362 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3364 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3370 SC/01 Use a glob alias rather than an array ref in eximstats generated
3371 parser. This improves both readability and performance.
3373 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3374 Don't display local sender or destination tables in eximstats unless
3375 there is data to show.
3376 Added average volumes into the eximstats top table text output.
3378 SC/03 Collect data on the number of addresses (recipients) as well
3379 as the number of messages in eximstats.
3381 TF/01 Correct an error in the documentation for the redirect router. Exim
3382 does (usually) call initgroups() when daemonizing.
3384 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3385 with consistent privilege compared to when running as a daemon.
3387 TF/03 Note in the spec that $authenticated_id is not set for local
3388 submissions from trusted users.
3390 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3391 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3393 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3394 by adding some example configuration directives to the default
3395 configuration file. A little bit of work is required to uncomment the
3396 directives and define how usernames and passwords are checked, but
3397 there is now a framework to start from.
3399 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3400 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3401 without this. I don't know how relevant this is to other LDAP libraries.
3403 PH/02 Add the verb name to the "unknown ACL verb" error.
3405 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3407 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3409 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3410 directory not expanded when it should be if an expanded home directory
3411 was set for the address (which is overridden by the transport).
3413 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3416 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3417 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3418 because it is too late at that time, and has no effect.
3420 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3421 security issue with \' (bugzilla #107). I could not use the
3422 PQescapeStringConn() function, because it needs a PGconn value as one of
3425 PH/08 When testing addresses using -bt, indicate those final addresses that
3426 are duplicates that would not cause an additional delivery. At least one
3427 person was confused, thinking that -bt output corresponded to deliveries.
3428 (Suppressing duplicates isn't a good idea as you lose the information
3429 about possibly different redirections that led to the duplicates.)
3431 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3432 systems where poll() doesn't work, in particular OS X.
3434 PH/10 Added more information to debugging output for retry time not reached.
3436 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3437 operations in malware.c.
3439 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3442 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3443 syslog_duplication set false, log lines that would normally be written
3444 both the the main log and to the reject log were not written to syslog at
3447 PH/14 In the default configuration, change the use of "message" in ACL warn
3448 statements to "add_header".
3450 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3451 not followed by a command (e.g. "seen endif").
3453 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3454 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3457 PH/17 Added extra conditions to the default value of delay_warning_condition
3461 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3462 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3463 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3466 The Auto-Submitted: and various List- headers are standardised, whereas I
3467 don't think Precedence: ever was.
3469 PH/18 Refactored debugging code in route_finduser() to show more information,
3470 in particular, the error code if getpwnam() issues one.
3472 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3473 This is apparently needed in addition to the PH/07 change above to avoid
3474 any possible encoding problems.
3476 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3477 but not after initializing Perl.
3479 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3480 output them only if debugging. By default they are written stderr,
3481 apparently, which is not desirable.
3483 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3486 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3489 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3491 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3492 authenticated or an ident call has been made. Suppress the default
3493 values for $authenticated_id and $authenticated_sender (but permit -oMai
3494 and -oMas) when testing with -bh.
3496 PH/24 Re-jigged the order of the tests in the default configuration so that the
3497 tests for valid domains and recipients precede the DNS black list and CSA
3498 tests, on the grounds that those ones are more expensive.
3500 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3501 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3502 This bug exists in every version of Exim that I still have, right back to
3505 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3506 However, an attempt to turn on case-sensitivity in a regex key by
3507 including (?-i) didn't work because the subject string was already
3508 lowercased, and the effects were non-intuitive. It turns out that a
3509 one-line patch can be used to allow (?-i) to work as expected.
3515 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3516 other effects) broke the use of negated acl sub-conditions.
3518 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3521 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3522 "Deny" causes Exim to reject the incoming connection with a 554 error.
3523 Unfortunately, if there is a major crisis, such as a disk failure,
3524 tcp-wrappers gives "deny", whereas what one would like would be some
3525 kind of temporary error. A kludge has been added to help with this.
3526 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3527 554 error is used if errno is still zero or contains ENOENT (which occurs
3528 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3531 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3533 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3534 errors. Otherwise a message that provokes a temporary error (when other
3535 messages do not) can cause a whole host to time out.
3537 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3538 addresses were routed directly to files or pipes from a redirect router.
3539 File deliveries just didn't batch; pipe deliveries might have suffered
3542 PH/06 A failure to get a lock for a hints database would erroneously always say
3543 "Failed to get write lock", even when it was really a read lock.
3545 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3546 of 0600. This has been changed to use the value of the lockfile_mode
3547 option (which defaults to 0600).
3549 PH/08 Applied small patch from the Sieve maintainer.
3551 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3552 folder from quota calculations, a direct delivery into this folder messed
3553 up the contents of the maildirsize file. This was because the regex was
3554 used only to exclude .Trash (or whatever) when the size of the mailbox
3555 was calculated. There was no check that a delivery was happening into an
3556 excluded directory. This bug has been fixed by ignoring all quota
3557 processing for deliveries into excluded directories.
3559 PH/10 Added the maildirfolder_create_regex option to appendfile.
3565 PH/01 The code for finding all the local interface addresses on a FreeBSD
3566 system running IPv6 was broken. This may well have applied to all BSD
3567 systems, as well as to others that have similar system calls. The broken
3568 code found IPv4 interfaces correctly, but gave incorrect values for the
3569 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3570 that it would not match correctly against @[] and not recognize the IPv6
3573 PH/02 The ipliteral router was not recognizing addresses of the form user@
3574 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3576 PH/03 Added disable_ipv6.
3578 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3579 decoded form, because it is most often used to construct To: headers
3580 lines in autoreplies, and the decoded form may well be syntactically
3581 invalid. However, $reply_address has leading white space removed, and all
3582 newlines turned into spaces so that the autoreply transport does not
3585 PH/05 If group was specified without a user on a router, and no group or user
3586 was specified on a transport, the group from the router was ignored.
3588 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3589 for visible compile-time settings that can be used to change these
3590 numbers, for those that want even more. Backwards compatibility with old
3591 spool files has been maintained. However, going back to a previous Exim
3592 release will lost any variables that are in spool files.
3594 PH/07 Two small changes when running in the test harness: increase delay when
3595 passing a TCP/IP connection to a new process, in case the original
3596 process has to generate a bounce, and remove special handling of
3597 127.0.0.2 (sic), which is no longer necessary.
3599 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3600 be the same on different OS.
3602 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3605 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3606 whether --show-vars was specified or not
3608 JJ/02 exipick: Added support for new ACL variable spool format introduced
3611 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3612 syntactically invalid From: or Reply-to: line, and a filter used this to
3613 generate an autoreply, and therefore failed to obtain an address for the
3614 autoreply, Exim could try to deliver to a non-existent relative file
3615 name, causing unrelated and misleading errors. What now happens is that
3616 it logs this as a hard delivery error, but does not attempt to create a
3619 PH/11 The exinext utility has a -C option for testing purposes, but although
3620 the given file was scanned by exinext itself; it wasn't being passed on
3621 when Exim was called.
3623 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3624 an end-of-file indication when reading a command response.
3626 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3627 compiled. In many other places in Exim, IPv6 addresses are always
3628 recognized, so I have changed this. It also means that IPv4 domain
3629 literals of the form [IPV4:n.n.n.n] are now always recognized.
3631 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3632 used if the router is not running as root, for example, when verifying at
3633 ACL time, or when using -bh. The debugging output from this situation was
3634 non-existent - all you got was a failure to exec. I have made two
3637 (a) Failures to set uid/gid, the current directory, or a process leader
3638 in a subprocess such as that created by queryprogram now generate
3639 suitable debugging output when -d is set.
3641 (b) The queryprogram router detects when it is not running as root,
3642 outputs suitable debugging information if -d is set, and then runs
3643 the subprocess without attempting to change uid/gid.
3645 PH/15 Minor change to Makefile for building test_host (undocumented testing
3648 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3649 additional section of a DNS packet that returns MX or SRV records.
3650 Instead, it always explicitly searches for A/AAAA records. This avoids
3651 major problems that occur when a DNS server includes only records of one
3652 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3653 fixed another bug: if SRV records were looked up and the corresponding
3654 address records were *not* found in the additional section, the port
3655 values from the SRV records were lost.
3657 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3658 using the correct key (the original address) when searching the retry
3659 rules in order to find which one to use for generating the retry hint.
3661 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3662 adding the default one. Similarly, if it contains a Reply-To: header, the
3663 errors_reply_to option, if set, is not used.
3665 PH/19 When calculating a retry time, Exim used to measure the "time since
3666 failure" by looking at the "first failed" field in the retry record. Now
3667 it does not use this if it is later than than the arrival time of the
3668 message. Instead it uses the arrival time. This makes for better
3669 behaviour in cases where some deliveries succeed, thus re-setting the
3670 "first failed" field. An example is a quota failure for a huge message
3671 when small messages continue to be delivered. Without this change, the
3672 "time since failure" will always be short, possible causing more frequent
3673 delivery attempts for the huge message than are intended.
3674 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3676 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3677 $1, $2, $3) because the numerical variables can be reset during some
3678 expansion items (e.g. "match"), thereby losing the authentication data.
3680 PH/21 Make -bV show the size of off_t variables so that the test suite can
3681 decide whether to run tests for quotas > 2G.
3683 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3684 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3685 filecount value is greater than 2G or if a quota value is greater than 2G
3686 on a system where the size of off_t is not greater than 4, a panic error
3689 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3690 never match. The debug and -bh output now contains an explicit error
3691 message indicating a malformed IPv4 address or mask.
3693 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3694 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3695 PH/23 above applies.
3697 PH/25 Do not write to syslog when running in the test harness. The only
3698 occasion when this arises is a failure to open the main or panic logs
3699 (for which there is an explicit test).
3701 PH/26 Added the /no_tell option to "control=freeze".
3703 PH/27 If a host name lookup failed very early in a connection, for example, if
3704 the IP address matched host_lookup and the reverse lookup yielded a name
3705 that did not have a forward lookup, an error message of the form "no IP
3706 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3707 could be logged. Now it outputs the IP address instead of "NULL".
3709 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3710 allows the sender and the authenticated sender to be set when
3711 submitting a message from within Exim. Since child_open_exim() is
3712 documented for local_scan(), the new function should be too.
3714 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3715 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3716 results in an empty string is now treated as unset.
3718 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3720 PH/31 Added sender_verify_fail logging option.
3722 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3723 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3724 tidied the source and removed it altogether.
3726 PH/33 When a queue run was abandoned because the load average was too high, a
3727 log line was always written; now it is written only if the queue_run log
3728 selector is set. In addition, the log line for abandonment now contains
3729 information about the queue run such as the pid. This is always present
3730 in "start" and "stop" lines but was omitted from the "abandon" line.
3732 PH/34 Omit spaces between a header name and the colon in the error message that
3733 is given when verify = headers_syntax fails (if there are lots of them,
3734 the message gets confusing).
3736 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3737 names, as there are now some PTR records that contain slashes. This check
3738 is only to protect against broken name servers that fall over on strange
3739 characters, so the fact that it applies to all lookups doesn't matter.
3741 PH/36 Now that the new test suite is complete, we can remove some of the
3742 special code in Exim that was needed for the old test suite. For example,
3743 sorting DNS records because real resolvers return them in an arbitrary
3744 order. The new test suite's fake resolver always returns records in the
3747 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3748 bounces) except when queue_only is set, to avoid logging races between
3749 the different processes.
3751 PH/38 Panic-die if .include specifies a non-absolute path.
3753 PH/39 A tweak to the "H" retry rule from its user.
3755 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3756 a label. They prevented compilation on older perls.
3758 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3759 a warning to be raised on newish perls.
3761 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3762 on queue. Changes to match documented behaviour of showing count of
3763 messages matching specified criteria.
3765 PH/40 Changed the default ident timeout from 30s to 5s.
3767 PH/41 Added support for the use of login_cap features, on those BSD systems
3768 that have them, for controlling the resources used by pipe deliveries.
3770 PH/42 The content-scanning code uses fopen() to create files in which to put
3771 message data. Previously it was not paying any attention to the mode of
3772 the files. Exim runs with umask(0) because the rest of the code creates
3773 files with open(), and sets the required mode explicitly. Thus, these
3774 files were ending up world-writeable. This was not a big issue, because,
3775 being within the spool directory, they were not world-accessible. I have
3776 created a function called modefopen, which takes an additional mode
3777 argument. It sets umask(777), creates the file, chmods it to the required
3778 mode, then resets the umask. All the relevant calls to fopen() in the
3779 content scanning code have been changed to use this function.
3781 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3782 to 24 hours. This avoids potential overflow problems when processing G
3783 and H retry rules. I suspect nobody ever tinkers with this value.
3785 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3787 PH/45 When the plaintext authenticator is running as a client, the server's
3788 challenges are checked to ensure they are valid base64 strings. By
3789 default, the authentication attempt is cancelled if an invalid string is
3790 received. Setting client_ignore_invalid_base64 true ignores these errors.
3791 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3792 they are received. Thus, the responses can be made to depend on the
3793 challenges. If an invalid string is ignored, an empty string is placed in
3796 PH/46 Messages that are created by the autoreply transport now contains a
3797 References: header, in accordance with RFCs 2822 and 3834.
3799 PH/47 Added authenticated_sender_force to the smtp transport.
3801 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3803 PH/49 Installed latest patch from the Sieve maintainer.
3805 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3806 was also set, the appendfile transport was unnecessarily scanning a
3807 directory of message files (e.g. for maildir delivery) to find the count
3808 of files (along with the size), even though it did not need this
3809 information. It now does the scan only if it needs to find either the
3810 size of the count of files.
3812 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3814 PH/52 Two bugs concerned with error handling when the smtp transport is
3817 (i) Exim was not creating retry information for temporary errors given
3818 for individual recipients after the DATA command when the smtp transport
3819 was used in LMTP mode. This meant that they could be retried too
3820 frequently, and not timed out correctly.
3822 (ii) Exim was setting the flag that allows error details to be returned
3823 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3824 recipients that were returned after the DATA command.
3826 PH/53 This is related to PH/52, but is more general: for any failing address,
3827 when detailed error information was permitted to be returned to the
3828 sender, but the error was temporary, then after the final timeout, only
3829 "retry timeout exceeded" was returned. Now it returns the full error as
3830 well as "retry timeout exceeded".
3832 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3833 do this, and (what is worse) MTAs that accept it.
3835 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3836 will now be deprecated.
3838 PH/56 New os.c-cygwin from the Cygwin maintainer.
3840 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3841 formats (previously only available in exim formats via -bpr, -bpru,
3842 and -bpra. Now also available in native and exiqgrep formats)
3844 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3845 with very large, slow to parse queues
3847 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3849 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3851 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3852 responses to authentication challenges, though it was showing the
3853 challenges; (ii) I've removed the CR characters from the debug output for
3856 PH/58 Allow for the insertion of a newline as well as a space when a string
3857 is turned into more than one encoded-word during RFC 2047 encoding. The
3858 Sieve code now uses this.
3860 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3861 data_4xx, lost_connection, tls_required.
3863 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3864 message was being sent as an SMTP response.
3866 PH/61 Add -l and -k options to exicyclog.
3868 PH/62 When verifying, if an address was redirected to one new address, so that
3869 verification continued, and the new address failed or deferred after
3870 having set something in $address_data, the value of $address_data was not
3871 passed back to the ACL. This was different to the case when no
3872 redirection occurred. The value is now passed back in both cases.
3874 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3875 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3876 use login_cap.h, so on its own it isn't the distinguishing feature. The
3877 new name refers directly to the setclassresources() function.
3879 PH/65 Added configuration files for NetBSD3.
3881 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3883 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3884 is preferred over IPv4.
3886 PH/68 The bounce_return_message and bounce_return_body options were not being
3887 honoured for bounces generated during the reception of non-SMTP messages.
3888 In particular, this applied to messages rejected by the ACL. This bug has
3889 been fixed. However, if bounce_return_message is true and bounce_return_
3890 body is false, the headers that are returned for a non-SMTP message
3891 include only those that have been read before the error was detected.
3892 (In the case of an ACL rejection, they have all been read.)
3894 PH/69 The HTML version of the specification is now built in a directory called
3895 spec_html instead of spec.html, because the latter looks like a path with
3896 a MIME-type, and this confuses some software.
3898 PH/70 Catch two compiler warnings in sieve.c.
3900 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3901 function verify_get_ident() calls ip_connect() to connect a socket, but
3902 if the "connect()" function timed out, ip_connect() used to close the
3903 socket. However, verify_get_ident() also closes the socket later, and in
3904 between Exim writes to the log, which may get opened at this point. When
3905 the socket was closed in ip_connect(), the log could get the same file
3906 descriptor number as the socket. This naturally causes chaos. The fix is
3907 not to close the socket in ip_connect(); the socket should be closed by
3908 the function that creates it. There was only one place in the code where
3909 this was missing, in the iplookup router, which I don't think anybody now
3910 uses, but I've fixed it anyway.
3912 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3913 well as to direct DNS lookups. Otherwise the handling of names in host
3914 lists is inconsistent and therefore confusing.
3920 PH/01 Two changes to the default runtime configuration:
3922 (1) Move the checks for relay_from_hosts and authenticated clients from
3923 after to before the (commented out) DNS black list checks.
3925 (2) Add control=submission to the relay_from_hosts and authenticated
3926 clients checks, on the grounds that messages accepted by these
3927 statements are most likely to be submissions.
3929 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3931 (1) Generate an error if the third argument for the ${prvs expansion is
3934 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3937 (3) Reset the variables that are obtained from the first argument of
3938 ${prvscheck and used in the second argument before leaving the code,
3939 because their memory is reclaimed, so using them afterwards may do
3942 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3943 one (it's much easier than Tom thought :-).
3945 (5) Because of (4), we can now allow for the use of $prvscheck_result
3946 inside the third argument.
3948 PH/03 For some reason, the default setting of PATH when running a command from
3949 a pipe transport was just "/usr/bin". I have changed it to
3952 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3953 anything to be listed in the output from -bV.
3955 PH/05 When a filter generated an autoreply, the entire To: header line was
3956 quoted in the delivery log line, like this:
3958 => >A.N.Other <ano@some.domain> <original@ddress> ...
3960 This has been changed so that it extracts the operative address. There
3961 may be more than one such address. If so, they are comma-separated, like
3964 => >ano@some.domain,ona@other.domain <original@ddress> ...
3966 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3967 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3968 not being looked up in the rDNS to get a host name, Exim was showing the
3969 IP address twice in Received: lines, even though the IP addresses were
3970 identical. For example:
3972 Received: from [1.2.3.4] (helo=[1.2.3.4])
3974 However, if the real host name was known, it was omitting the HELO data
3975 if it matched the actual IP address. This has been tidied up so that it
3976 doesn't show the same IP address twice.
3978 PH/07 When both +timestamp and +memory debugging was on, the value given by
3979 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3980 called by the debug printing, thereby overwriting the timestamp buffer.
3981 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3983 PH/08 When the original message was included in an autoreply transport, it
3984 always said "this is a copy of the message, including all the headers",
3985 even if body_only or headers_only was set. It now gives an appropriate
3988 PH/09 Applied a patch from the Sieve maintainer which:
3990 o fixes some comments
3991 o adds the (disabled) notify extension core
3992 o adds some debug output for the result of if/elsif tests
3993 o points to the current vacation draft in the documentation
3994 and documents the missing references header update
3998 o fixes a bug in processing the envelope test (when testing
3999 multiple envelope elements, the last element determined the
4002 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4003 Electronic Mail") by including:
4005 Auto-submitted: auto-generated
4007 in the messages that it generates (bounce messages and others, such as
4008 warnings). In the case of bounce messages for non-SMTP messages, there was
4009 also a typo: it was using "Auto_submitted" (underscore instead of
4010 hyphen). Since every message generated by Exim is necessarily in response
4011 to another message, thes have all been changed to:
4013 Auto-Submitted: auto-replied
4015 in accordance with these statements in the RFC:
4017 The auto-replied keyword:
4019 - SHOULD be used on messages sent in direct response to another
4020 message by an automatic process,
4022 - MUST NOT be used on manually-generated messages,
4024 - MAY be used on Delivery Status Notifications (DSNs) and Message
4025 Disposition Notifications (MDNs),
4027 - MUST NOT be used on messages generated by automatic or periodic
4028 processes, except for messages which are automatic responses to
4031 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4032 to the default Received: header definition.
4034 PH/12 Added log selector acl_warn_skipped (default on).
4036 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4037 variables because (a) they are in the wrong storage pool and (b) even if
4038 they were copied, it wouldn't work properly because of the caching.
4040 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4041 checking when decoding. Apparently there are clients that generate
4042 overlong encoded strings. Why am I not surprised?
4044 PH/15 If the first argument of "${if match_address" was not empty, but did not
4045 contain an "@" character, Exim crashed. Now it writes a panic log message
4046 and treats the condition as false.
4048 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4050 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4051 extension "envelope-auth". The code is finished and in agreement with
4052 other implementations, but there is no documentation so far and in fact,
4053 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4054 not changing the active code.
4056 Print executed "if" and "elsif" statements when debugging is used. This
4057 helps a great deal to understand what a filter does.
4059 Document more things not specified clearly in RFC3028. I had all this
4060 sorted out, when out of a sudden new issues came to my mind. Oops."
4062 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4065 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4066 canonical form (as documented). However, after a host name lookup from
4067 the IP address, check_host() was doing a simple string comparison with
4068 addresses acquired from the DNS when checking that the found name did
4069 have the original IP as one of its addresses. Since any found IPv6
4070 addresses are likely to be in abbreviated form, the comparison could
4071 fail. Luckily, there already exists a function for doing the comparison
4072 by converting both addresses to binary, so now that is used instead of
4073 the text comparison.
4075 PH/20 There was another similar case to PH/19, when a complete host name was
4076 given in a host list; looking up its IP address could give an abbreviated
4077 form, whereas the current host's name might or might not be abbreviated.
4078 The same fix has been applied.
4084 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4085 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4088 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4089 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4091 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4093 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4094 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4095 Also, for "Auto-Submitted", treat anything other than "no" as
4096 non-personal, in accordance with RFC 3834. (Previously it treated
4097 anything starting "auto-" as non-personal.)
4099 TF/01 The control=submission/name=... option had a problem with syntax
4100 errors if the name included a slash character. The /name= option
4101 now slurps the rest of the string, so it can include any characters
4102 but it must come last in the list of options (after /sender_retain
4105 PH/05 Some modifications to the interface to the fake nameserver for the new
4113 TK/01 Added the "success_on_redirect" address verification option. See
4114 NewStuff for rationale and an example.
4116 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4118 PH/02 Patch to exigrep to allow it to work on syslog lines.
4120 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4121 fread() to skip over the body file's header line, because in Cygwin the
4122 header line is locked and is inaccessible.
4124 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4125 co-exist for some time) to make it clear that it is the Exim ID that is
4126 referenced, not the Message-ID: header line.
4128 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4129 string_format() function, because snprintf() does not exist on all
4132 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4133 ${stat: expansion item.
4135 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4136 protocol synchronization error", to keep the pedants happy.
4138 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4139 well as for IRIX systems, when gcc is being used. See the host.c source
4142 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4144 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4147 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4148 transport and to the smtp transport in LMTP mode.
4150 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4152 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4153 run a filter in a subprocess. This could lead to confusion in subsequent
4154 lookups in the parent process. There should also be a search_tidyup() at
4155 the end of the subprocess.
4157 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4158 only if the host matched helo_try_verify_hosts, which caused the
4159 verification to occur when the EHLO/HELO command was issued. The ACL just
4160 tested the remembered result. Now, if a previous verification attempt has
4161 not happened, "verify = helo" does it there and then.
4163 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4165 TK/03 Fix log output including CR from clamd.
4167 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4168 exist provoked a memory error which could cause a segfault.
4170 PH/15 Installed PCRE 6.2
4172 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4174 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4175 of the problem. Specifically, suggested +O2 rather than +O1 for the
4178 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4180 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4181 record, the port was not correctly logged when the outgoing_port log
4182 selector was set (it logged the transort's default port).
4184 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4185 fallback_hosts, and "hosts" in the smtp transport.
4187 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4188 host errors such as "Connection refused".
4190 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4191 authentication with radiusclient 0.4.9:
4193 - Error returned from rc_read_config was caught wrongly
4194 - Username/password not passed on to radius server due to wrong length.
4196 The presumption is that some radiusclient API changes for 4.51/PH/17
4197 were not taken care of correctly. The code is still untested by me (my
4198 Linux distribution still has 0.3.2 of radiusclient), but it was
4199 contributed by a Radius user.
4201 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4202 expanding the "port" option of the smtp transport.
4204 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4205 while reading a MIME header. Thanks to Tom Hughes for a patch.
4207 PH/24 Include config.h inside local_scan.h so that configuration settings are
4210 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4211 that in an ACL for RCPT (for example), you can examine exactly what was
4214 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4215 commands, but it was not correctly comparing the address with the actual
4216 client host address. Thus, it would show the EHLO address in Received:
4217 header lines when this was not necessary.
4219 PH/27 Added the % operator to ${eval:}.
4221 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4222 it should be ignoring failures (because with -C, for example, it has lost
4223 privilege). It wasn't ignoring creation failures other than "already
4226 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4229 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4230 because an input line was too long, either on its own, or by virtue of
4231 too many continuations, the temporary file was not being removed, and the
4232 return code was incorrect.
4234 PH/31 Missing "BOOL" in function definition in filtertest.c.
4236 PH/32 Applied Sieve patches from the maintainer.
4238 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4240 PH/33 Added "verify = not_blind".
4242 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4243 Local/Makefile (with some defaults set). These are used in built scripts
4244 such as exicyclog, but they have never been used in the exim_install
4245 script (though there are many overriding facilities there). I have
4246 arranged that the exim_install script now takes note of these two
4249 PH/35 Installed configuration files for Dragonfly.
4251 PH/36 When a locally submitted message by a trusted user did not contain a
4252 From: header, and the sender address was obtained from -f or from an SMTP
4253 MAIL command, and the trusted user did not use -F to supply a sender
4254 name, $originator_name was incorrectly used when constructing a From:
4255 header. Furthermore, $originator_name was used for submission mode
4256 messages from external hosts without From: headers in a similar way,
4257 which is clearly wrong.
4259 PH/37 Added control=suppress_local_fixups.
4261 PH/38 When log_selector = +received_sender was set, and the addition of the
4262 sender made the log line's construction buffer exactly full, or one byte
4263 less than full, an overflow happened when the terminating "\n" was
4266 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4267 when the result of a list match is failure because a DNS lookup failed.
4269 PH/40 RM_COMMAND is now used in the building process.
4271 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4272 the "build-* directories that it finds.
4274 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4275 address in a domain literal was a prefix of an interface address.
4277 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4278 when verifying a sender address, unless rewrite_headers is false.
4280 PH/44 Wrote a long comment about why errors_to addresses are verified as
4281 recipients, not senders.
4283 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4284 the ratelimit ACL was added.
4286 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4288 PH/46 Added extra information about PostgreSQL errors to the error string.
4290 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4291 avoiding the need to install special zones in a real server. This is
4292 backwards compatible; if it can't find the fake resolver, it drops back.
4293 Thus, both old and new test suites can be run.
4295 TF/02 Added util/ratelimit.pl
4297 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4298 clock is set back in time.
4300 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4301 Candler <B.Candler@pobox.com>.
4303 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4304 OK for addresses that are the result of redirections.
4306 PH/48 A number of further additions for the benefit of the new test suite,
4307 including a fake gethostbyname() that interfaces to the fake DNS resolver
4310 TF/06 The fix for widen_domains has also been applied to qualify_single and
4311 search_parents which are the other dnslookup options that can cause
4314 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4317 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4319 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4320 service" if it encountered an MX record with an empty target hostname.
4321 The message is now "an MX or SRV record indicated no SMTP service".
4323 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4324 if the DNS of the sending site is misconfigured. This is quite a
4325 common situation. This change restores the behaviour of treating a
4326 helo verification defer as a failure.
4328 PH/51 If self=fail was set on a router, the bounce message did not include the
4329 actual error message.
4335 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4337 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4338 command itself ended in error, the underlying message about the transport
4339 filter timeout was being overwritten with the pipe command error. Now the
4340 underlying error message should be appended to the second error message.
4342 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4344 PH/02 Reduce the amount of output that "make" produces by default. Full output
4345 can still be requested.
4347 PH/03 The warning log line about a condition test deferring for a "warn" verb
4348 was being output only once per connection, rather than after each
4349 occurrence (because it was using the same function as for successful
4350 "warn" verbs). This seems wrong, so I have changed it.
4352 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4353 it should not have, which might have caused a crash in the right
4354 circumstances, but probably never did.
4356 PH/04 Installed a modified version of Tony Finch's patch to make submission
4357 mode fix the return path as well as the Sender: header line, and to
4358 add a /name= option so that you can make the user's friendly name appear
4361 TF/03 Added the control = fakedefer ACL modifier.
4363 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4364 Mark Lowes for thorough testing.
4366 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4368 TK/03 Merged latest SRS patch from Miles Wilton.
4370 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4371 with the definition in sysexits.h (which is #included earlier).
4372 Fortunately, Exim does not actually use EX_OK. The code used to try to
4373 preserve the sysexits.h value, by assuming that macro definitions were
4374 scanned for macro replacements. I have been disabused of this notion,
4375 so now the code just undefines EX_OK before #including unistd.h.
4377 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4378 in the smtp transport. When a block could not be written in a single
4379 write() function, the timeout was being re-applied to each part-write.
4380 This seems wrong - if the receiver was accepting one byte at a time it
4381 would take for ever. The timeout is now adjusted when this happens. It
4382 doesn't have to be particularly precise.
4384 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4385 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4387 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4388 without a domain if the check to <postmaster@domain> fails.
4390 SC/01 Eximstats: added -xls and the ability to specify output files
4391 (patch written by Frank Heydlauf).
4393 SC/02 Eximstats: use FileHandles for outputting results.
4395 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4397 SC/04 Eximstats: fixed display of large numbers with -nvr option
4399 SC/05 Eximstats: fixed merging of reports with empty tables.
4401 SC/06 Eximstats: added the -include_original_destination flag
4403 SC/07 Eximstats: removed tabs and trailing whitespace.
4405 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4407 TK/06 MBOX spool code: Add real "From " MBOX separator line
4408 so the .eml file is really in mbox format (even though
4409 most programs do not really care). Patch from Alex Miller.
4411 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4412 The latter is generated from $received_to and is only set if the
4413 message has one envelope recipient. SA can use these headers,
4414 obviously out-of-the-box. Patch from Alex Miller.
4416 PH/08 The ${def test on a variable was returning false if the variable's
4417 value was "0", contrary to what the specification has always said!
4418 The result should be true unless the variable is empty.
4420 PH/09 The syntax error of a character other than { following "${if
4421 def:variable_name" (after optional whitespace) was not being diagnosed.
4422 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4423 accidental colon was present, for example, could give incorrect results.
4425 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4426 result is used (not including appendfile, where other changes are about
4429 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4430 This involved changing a lot of size variables from int to off_t. It
4431 should work with maildirs and everything.
4433 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4434 spamd dying while we are connected to it.
4436 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4439 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4440 function for BDB 4.3.
4442 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4444 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4445 the sources for PCRE have been re-arranged and more files are now
4448 PH/15 The code I had for printing potentially long long variables in PH/11
4449 above was not the best (it lost precision). The length of off_t variables
4450 is now inspected at build time, and an appropriate printing format (%ld
4451 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4452 to be "long long int" or "long int". This is needed for the internal
4453 formatting function string_vformat().
4455 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4456 the configuration file to be ":syslog", then the script "guesses" where
4457 the logs files are, rather than using the compiled in default. In our
4458 case the guess is not the same as the compiled default, so the script
4459 suddenly stopped working when I started to use syslog. The patch checks
4460 to see if log_file_path is "". If so, it attempts to read it from exim
4461 with no configuration file to get the compiled in version, before it
4462 falls back to the previous guessing code."
4464 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4465 implementing BATV in an Exim configuration. See NewStuff for the gory
4468 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4469 Makefile that are specific to HP-UX.
4471 PH/18 If the "use_postmaster" option was set for a recipient callout together
4472 with the "random" option, the postmaster address was used as the MAIL
4473 FROM address for the random test, but not for the subsequent recipient
4474 test. It is now used for both.
4476 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4477 patch removes a few documentation additions to RFC 3028, because the
4478 latest draft now contains them. It adds the new en;ascii-case comparator
4479 and a new error check for 8bit text in MIME parts. Comparator and
4480 require names are now matched exactly. I enabled the subaddress
4481 extension, but it is not well tested yet (read: it works for me)."
4483 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4484 rework some of the code of TK/09 above to avoid the hardwired use of
4485 "%lld" and "long long". Replaced the call to snprintf() with a call to
4488 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4489 records point to non-existent hosts", "retry timeout exceeded", and
4490 "retry time not reached for any host after a long failure period".
4492 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4493 experimental DomainKeys support:
4495 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4496 (2) On an error such as an illegally used "control", the wrong name for
4497 the control was given.
4499 These problems did NOT occur unless DomainKeys support was compiled.
4501 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4503 PH/24 Added ${if match_ip condition.
4505 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4506 databases so that it will be absolutely obvious if a crash occurs in the
4507 DB library. This is a regular occurrence (often caused by mis-matched
4510 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4511 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4512 picked up on a user's system that detects such things. There doesn't seem
4513 to be a gcc warning option for this - only an attribute that has to be
4514 put on the function's prototype. It seems that in Fedora Core 4 they have
4515 set this on a number of new functions. No doubt there will be more in due
4518 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4519 specify a transport. However, if an address that was verified by such a
4520 router was the subject of a callout, Exim crashed because it tried to
4521 read the rcpt_include_affixes from the non-existent transport. Now it
4522 just assumes that the setting of that option is false. This bug was
4523 introduced by 4.51/PH/31.
4525 PH/28 Changed -d+all to exclude +memory, because that information is very
4526 rarely of interest, but it makes the output a lot bigger. People tend to
4527 do -d+all out of habit.
4529 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4530 code in os-type was giving problems when libc.so lives in lib64, like on
4533 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4534 aren't the modern standard, and it seems that some systems' include files
4535 don't always have them. Exim was already checking for some of the newer
4536 ones like T_AAAA, and defining it itself. I've added checks for all the
4537 record types that Exim uses.
4539 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4540 not automatically generating a new one, as it is supposed to. This
4541 prevented TLS from working. If the file did exist, but contained invalid
4542 data, a new version was generated, as expected. It was only the case of a
4543 non-existent file that was broken.
4545 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4546 with a change in libdomainkeys > 0.64.
4548 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4549 from DNS. If the selector record carries the flag, it now has
4550 precedence over the domain-wide flag.
4552 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4554 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4555 the use of a port name that isn't defined in /etc/services) occurred, the
4556 message was deferred as in a normal delivery, and thus remained on the
4557 spool, instead of being failed because of the mua_wrapper setting. This
4558 is now fixed, and I tidied up some of the mua_wrapper messages at the
4561 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4562 the messages in a hash of arrays rather than using individual hashes.
4563 This is a bit cleaner and results in dramatic memory savings, albeit
4564 at a slight CPU cost.
4566 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4567 as requested by Marc Sherman.
4569 SC/10 Eximstats: added histograms for user specified patterns as requested
4572 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4574 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4575 fopen() in the content-scanning modules that did not already have it.
4581 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4582 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4584 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4586 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4588 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4589 to test Sieve filters that use "vacation".
4591 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4592 that changes the way the GnuTLS parameters are stored in the cache file.
4593 The new format can be generated externally. For backward compatibility,
4594 if the data in the cache doesn't make sense, Exim assumes it has read an
4595 old-format file, and it generates new data and writes a new file. This
4596 means that you can't go back to an older release without removing the
4599 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4600 work if there are any delivery delays because "one_time" forces the
4601 parent to be marked "delivered", so its unseen clone is never tried
4602 again. For this reason, Exim now forbids the simultaneous setting of
4605 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4606 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4607 entry if you want to know the details. The fix, however, neglected to
4608 consider the case when local delivery batching is involved. The test for
4609 "previously delivered" was not happening when checking to see if an
4610 address could be batched with a previous (undelivered) one; under
4611 certain circumstances this could lead to multiple deliveries to the same
4614 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4615 in its include files, and this causes problems building Exim.
4617 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4618 header_syntax) but Exim was just ignoring anything given after a slash.
4619 In particular, this caused confusion with an attempt to use "verify =
4620 reverse_host_lookup/defer_ok". An error is now given when options are
4621 supplied for verify items that do not have them. (Maybe reverse_host_
4622 lookup should have a defer_ok option, but that's a different point.)
4624 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4625 defined by RFC 821) to 2048, because there were problems with some AUTH
4626 commands, and RFC 1869 says the size should be increased for extended
4627 SMTP commands that take arguments.
4629 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4632 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4633 "unknown" error; now it says that the functionality isn't in the binary.
4635 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4636 an address' error message when a string expansion fails (syntax or
4637 whatever). Otherwise the password may appear in the log. Following change
4638 PH/42 below, there is no longer a chance of it appearing in a bounce
4641 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4643 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4644 of its addresses was ever tried. (Bugzilla bug #2.)
4646 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4647 the result incorrectly in the debug output. (It correctly added a newline
4648 to what was transported.)
4650 TF/01 Added $received_time.
4652 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4653 commented out examples of how to interface to a virus scanner and to
4654 SpamAssassin. Also added commented examples of av_scanner and
4655 spamd_address settings.
4657 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4658 and controls are allowed in which ACLs. There were a couple of minor
4659 errors. Some of the entries in the conditions table (which is a table of
4660 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4661 negation of where the condition IS allowed.
4663 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4665 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4666 header file does not have a version number, so I've had to invent a new
4667 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4668 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4669 radiusclient), but it was contributed by a Radius user.
4671 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4672 files or directories, for OpenSSL.
4674 PH/19 When an Exim process that is running as root has to create an Exim log
4675 file, it does so in a subprocess that runs as exim:exim so as to get the
4676 ownership right at creation (otherwise, other Exim processes might see
4677 the file with the wrong ownership). There was no test for failure of this
4678 fork() call, which would lead to the process getting stuck as it waited
4679 for a non-existent subprocess. Forks do occasionally fail when resources
4680 run out. I reviewed all the other calls to fork(); they all seem to check
4683 PH/20 When checking for unexpected SMTP input at connect time (before writing
4684 the banner), Exim was not dealing correctly with a non-positive return
4685 from the read() function. If the client had disconnected by this time,
4686 the result was a log entry for a synchronization error with an empty
4687 string after "input=" when read() returned zero. If read() returned -1
4688 (an event I could not check), uninitialized data bytes were printed.
4689 There were reports of junk text (parts of files, etc) appearing after
4692 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4694 PH/22 Added support for macro redefinition, and (re)definition in between
4695 driver and ACL definitions.
4697 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4698 forgetting to use the resulting value; it was using the unexpanded value.
4700 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4701 hadn't been configured. The fix is from Juergen Kreileder, who
4702 understands it better than I do:
4704 "Here's what I see happening with three configured cyrus_sasl
4705 authenticators configured (plain, login, cram-md5):
4707 On startup auth_cyrus_sasl_init() gets called for each of these.
4708 This means three calls to sasl_listmech() without a specified mech_list.
4709 => SASL tests which mechs of all available mechs actually work
4710 => three warnings about OTP not working
4711 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4713 With the patch, sasl_listmech() also gets called three times. But now
4714 SASL's mech_list option is set to the server_mech specified in the the
4715 authenticator. Or in other words, the answer from sasl_listmech()
4716 gets limited to just the mech you're testing for (which is different
4718 => the return list contains just 'plain' or 'login', 'cram-md5' or
4719 nothing depending on the value of ob->server_mech.
4721 I've just tested the patch: Authentication still works fine,
4722 unavailable mechs specified in the exim configuration are still
4723 caught, and the auth.log warnings about OTP are gone."
4725 PH/25 When debugging is enabled, the contents of the command line are added
4726 to the debugging output, even when log_selector=+arguments is not
4729 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4730 answer is "GNU", and only if the return is "GNU/something" is the answer
4733 PH/27 $acl_verify_message is now set immediately after the failure of a
4734 verification in an ACL, and so is available in subsequent modifiers. In
4735 particular, the message can be preserved by coding like this:
4737 warn !verify = sender
4738 set acl_m0 = $acl_verify_message
4740 Previously, $acl_verify_message was set only while expanding "message"
4741 and "log_message" when a very denied access.
4743 PH/28 Modified OS/os.c-Linux with
4745 -#ifndef OS_LOAD_AVERAGE
4746 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4748 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4749 nomenclature these days.)
4751 PH/29 Installed patch from the Sieve maintainer that adds the options
4752 sieve_useraddress and sieve_subaddress to the redirect router.
4754 PH/30 In these circumstances:
4755 . Two addresses routed to the same list of hosts;
4756 . First host does not offer TLS;
4757 . First host accepts first address;
4758 . First host gives temporary error to second address;
4759 . Second host offers TLS and a TLS session is established;
4760 . Second host accepts second address.
4761 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4762 and peerdn, if requested) that were in fact used only for the second
4765 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4766 attention to any local part prefix or suffix that was matched by the
4767 router that accepted the address. It now behaves in the same way as it
4768 does for delivery: the affixes are removed from the local part unless
4769 rcpt_include_affixes is set on the transport.
4771 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4772 timeout during the DATA phase of an incoming message.
4774 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4775 applied a patch sanctioned by the Sieve maintainer.
4777 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4778 the uid or gid is negative. A case of a negative gid caused this to be
4779 noticed. The fix allows for either to be negative.
4781 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4782 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4785 PH/36 The change PH/12 above was broken. Fixed it.
4787 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4788 the grounds that routing the same address twice would always produce the
4789 same answer. This might have been true once, but it is certainly no
4790 longer true now. Routing a child address may depend on the previous
4791 routing that produced that child. Some complicated redirection strategies
4792 went wrong when messages had multiple recipients, and made Exim's
4793 behaviour dependent on the order in which the addresses were given.
4795 I have moved the duplicate checking until after the routing is complete.
4796 Exim scans the addresses that are assigned to local and remote
4797 transports, and removes any duplicates. This means that more work will be
4798 done, as duplicates will always all be routed, but duplicates are
4799 presumably rare, so I don't expect this is of any significance.
4801 For deliveries to pipes, files, and autoreplies, the duplicate checking
4802 still happens during the routing process, since they are not going to be
4805 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4806 It corrects a timeout issue with spamd. This is Ian's comment: "The
4807 background is that sometimes spamd either never reads data from a
4808 connection it has accepted, or it never writes response data. The exiscan
4809 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4810 blindly assumes that writes won't block so it may never time out."
4812 PH/39 Allow G after quota size as well as K and M.
4814 PH/40 The value set for $authenticated_id in an authenticator may not contain
4815 binary zeroes or newlines because the value is written to log lines and
4816 to spool files. There was no check on this. Now the value is run through
4817 the string_printing() function so that such characters are converted to
4818 printable escape sequences.
4820 PH/41 $message_linecount is a new variable that contains the total number of
4821 lines in the message. Compare $body_linecount, which is the count for the
4824 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4825 bounce and delay warning messages, except in certain special cases, which
4828 (a) An SMTP error message from a remote host;
4829 (b) A message specified in a :fail: redirection;
4830 (c) A message specified in a "fail" command in a system filter;
4831 (d) A message specified in a FAIL return from the queryprogram router;
4832 (e) A message specified by the cannot_route_message router option.
4834 In these cases only, Exim does include the error details in bounce and
4835 warning messages. There are also a few cases where bland messages such
4836 as "unrouteable address" or "local delivery error" are given.
4838 PH/43 $value is now also set for the "else" part of a ${run expansion.
4840 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4841 being worked on, but at least Exim now implements the latest version to
4844 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4845 process to complete was treated as a delivery failure, a timeout while
4846 writing the message to the pipe was logged, but erroneously treated as a
4847 successful delivery. Such timeouts include transport filter timeouts. For
4848 consistency with the overall process timeout, these timeouts are now
4849 treated as errors, giving rise to delivery failures by default. However,
4850 there is now a new Boolean option for the pipe transport called
4851 timeout_defer, which, if set TRUE, converts the failures into defers for
4852 both kinds of timeout. A transport filter timeout is now identified in
4855 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4856 systems where "make" and "gmake" are different, calling "gmake" at top
4857 level broke things. I've arranged for the value of $(MAKE) to be passed
4858 from the Makefile to this script so that it can call the same version of
4862 A note about Exim versions 4.44 and 4.50
4863 ----------------------------------------
4865 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4866 changes of various kinds. As a consequence, a big documentation update was
4867 needed. This delayed the release for rather longer than seemed good, especially
4868 in the light of a couple of (minor) security issues. Therefore, the changes
4869 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4870 release. So 4.44 and 4.50 are in effect two different branches that both start
4873 I have left the 4.50 change log unchanged; it contains all the changes since
4874 4.43. The change log for 4.44 is below; many of its items are identical to
4875 those for 4.50. This seems to be the most sensible way to preserve the
4876 historical information.
4882 1. Minor wording change to the doc/README.SIEVE file.
4884 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4885 computation of the current number of files was incorrect.
4887 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4888 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4891 4. Give more explanation in the error message when the command for a transport
4892 filter fails to execute.
4894 5. There are several places where Exim runs a non-Exim command in a
4895 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4896 was being done only for the command run by the queryprogram router. It is
4897 now done for all such subprocesses. The other cases are: ${run, transport
4898 filters, and the commands run by the lmtp and pipe transports.
4900 6. Added CONFIGURE_GROUP build-time option.
4902 7. Some older OS have a limit of 256 on the maximum number of file
4903 descriptors. Exim was using setrlimit() to set 1000 as a large value
4904 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4905 systems. I've change it so that if it can't get 1000, it tries for 256.
4907 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4908 was an oversight, and furthermore, ever since the addition of extra
4909 controls (e.g. 4.43/32), the checks on when to allow different forms of
4910 "control" were broken. There should now be diagnostics for all cases when a
4911 control that does not make sense is encountered.
4913 9. Added the /retain_sender option to "control=submission".
4915 10. $recipients is now available in the predata ACL (oversight).
4917 11. Tidy the search cache before the fork to do a delivery from a message
4918 received from the command line. Otherwise the child will trigger a lookup
4919 failure and thereby defer the delivery if it tries to use (for example) a
4920 cached ldap connection that the parent has called unbind on.
4922 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4923 of $address_data from the recipient verification was clobbered by the
4924 sender verification.
4926 13. The value of address_data from a sender verification is now available in
4927 $sender_address_data in subsequent conditions in the ACL statement.
4929 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4931 15. Added a new option "connect=<time>" to callout options, to set a different
4934 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4935 was its contents. (It was OK if the option was not defined at all.)
4937 17. A "Completed" log line is now written for messages that are removed from
4938 the spool by the -Mrm option.
4940 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4941 information about exactly what failed.
4943 19. Added -dd to debug only the daemon process.
4945 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4946 handles timeouts, both on the server side and network timeouts. Renamed the
4947 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4949 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4950 It is now set to "smtps".
4952 22. $host_address is now set to the target address during the checking of
4953 ignore_target_hosts.
4955 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4956 being passed; this would have caused $sender_host_name to have been used if
4957 matching the list had actually called for a host name (not very likely,
4958 since this list is usually IP addresses). A host name is now passed as
4961 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4962 code that specifies a non-restarting handler (typically sigaction() in
4963 modern systems) in an attempt to fix a rare and obscure crash bug.
4965 25. Narrowed the window for a race in the daemon that could cause it to ignore
4966 SIGCHLD signals. This is not a major problem, because they are used only to
4967 wake it up if nothing else does.
4969 26. A malformed maildirsize file could cause Exim to calculate negative values
4970 for the mailbox size or file count. Odd effects could occur as a result.
4971 The maildirsize information is now recalculated if the size or filecount
4974 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4975 support for a long time. Removed HAVE_SYS_VFS_H.
4977 28. Installed the latest version of exipick from John Jetmore.
4979 29. In an address list, if the pattern was not a regular expression, an empty
4980 subject address (from a bounce message) matched only if the pattern was an
4981 empty string. Non-empty patterns were not even tested. This was the wrong
4982 because it is perfectly reasonable to use an empty address as part of a
4983 database query. An empty address is now tested by patterns that are
4984 lookups. However, all the other forms of pattern expect the subject to
4985 contain a local part and a domain, and therefore, for them, an empty
4986 address still always fails if the pattern is not itself empty.
4988 30. Exim went into a mad DNS loop when attempting to do a callout where the
4989 host was specified on an smtp transport, and looking it up yielded more
4990 than one IP address.
4992 31. Re-factored the code for checking spool and log partition space into a
4993 function that finds that data and another that does the check. The former
4994 is then used to implement four new variables: $spool_space, $log_space,
4995 $spool_inodes, and $log_inodes.
4997 32. The RFC2047 encoding function was originally intended for short strings
4998 such as real names; it was not keeping to the 75-character limit for
4999 encoded words that the RFC imposes. It now respects the limit, and
5000 generates multiple encoded words if necessary. To be on the safe side, I
5001 have increased the buffer size for the ${rfc2047: expansion operator from
5004 33. It is now permitted to omit both strings after an "if" condition; if the
5005 condition is true, the result is "true". As before, when the second string
5006 is omitted, a false condition yields an empty string. This makes it less
5007 cumbersome to write custom ACL and router conditions.
5009 34. Failure to deliver a bounce message always caused it to be frozen, even if
5010 there was an errors_to setting on the router. The errors_to setting is now
5013 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5014 canonical form (fully expanded) before being placed in
5015 $sender_host_address.
5017 36. The table in the code that translates DNS record types into text (T_A to
5018 "A" for instance) was missing entries for NS and CNAME. It is just possible
5019 that this could have caused confusion if both these types were looked up
5020 for the same domain, because the text type is used as part of Exim's
5021 per-process caching. But the chance of anyone hitting this buglet seems
5024 37. The dnsdb lookup has been extended in a number of ways.
5026 (1) There is a new type, "zns", which walks up the domain tree until it
5027 finds some nameserver records. It should be used with care.
5029 (2) There is a new type, "mxh", which is like "mx" except that it returns
5030 just the host names, not the priorities.
5032 (3) It is now possible to give a list of domains (or IP addresses) to be
5033 looked up. The behaviour when one of the lookups defers can be
5034 controlled by a keyword.
5036 (4) It is now possible to specify the separator character for use when
5037 multiple records are returned.
5039 38. The dnslists ACL condition has been extended: it is now possible to supply
5040 a list of IP addresses and/or domains to be looked up in a particular DNS
5043 39. Added log_selector=+queue_time_overall.
5045 40. When running the queue in the test harness, wait just a tad after forking a
5046 delivery process, to get repeatability of debugging output.
5048 41. Include certificate and key file names in error message when GnuTLS fails
5049 to set them up, because the GnuTLS error message doesn't include the name
5050 of the failing file when there is a problem reading it.
5052 42. Allow both -bf and -bF in the same test run.
5054 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5056 44. The "Exiscan patch" is now merged into the mainline Exim source.
5058 45. Sometimes the final signoff response after QUIT could fail to get
5059 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5060 < 0 before doing a fflush(). This bug looks as though it goes back to the
5061 introduction of TLS in release 3.20, but "sometimes" must have been rare
5062 because the tests only now provoked it.
5064 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5065 (this can affect the format of dates).
5067 47. exim_tidydb, when checking for the continued existence of a message for
5068 which it has found a message-specific retry record, was not finding
5069 messages that were in split spool directories. Consequently, it was
5070 deleting retry records that should have stayed in existence.
5072 48. Steve fixed some bugs in eximstats.
5074 49. The SPA authentication driver was not abandoning authentication and moving
5075 on to the next authenticator when an expansion was forced to fail,
5076 contradicting the general specification for all authenticators. Instead it
5077 was generating a temporary error. It now behaves as specified.
5079 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5080 (the order specifies the preference for clients). The order is now AES256,
5081 AES128, 3DES, ARCFOUR128.
5083 51. Small patch to Sieve code - explicitly set From: when generating an
5086 52. Exim crashed if a remote delivery caused a very long error message to be
5087 recorded - for instance if somebody sent an entire SpamAssassin report back
5088 as a large number of 550 error lines. This bug was coincidentally fixed by
5089 increasing the size of one of Exim's internal buffers (big_buffer) that
5090 happened as part of the Exiscan merge. However, to be on the safe side, I
5091 have made the code more robust (and fixed the comments that describe what
5094 53. Now that there can be additional text after "Completed" in log lines (if
5095 the queue_time_overall log selector is set), a one-byte patch to exigrep
5096 was needed to allow it to recognize "Completed" as not the last thing in
5099 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5100 patch that reportedly fixes this has been added. I am not expert enough to
5101 create a test for it. This is what the patch creator wrote:
5103 "I found a little strange behaviour of ldap code when working with
5104 Windows 2003 AD Domain, where users was placed in more than one
5105 Organization Units. When I tried to give exim partial DN, the exit code
5106 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5107 But simultaneously result of request was absolutely normal ldap result,
5108 so I produce this patch..."
5110 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5111 REFERENCE, so I have modified the code to exclude the patch when that macro
5114 55. Some experimental protocols are using DNS PTR records for new purposes. The
5115 keys for these records are domain names, not reversed IP addresses. The
5116 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5117 leaves it alone. Component reversal etc. now happens only for IP addresses.
5120 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5122 57. Double the size of the debug message buffer (to 2048) so that more of very
5123 long debug lines gets shown.
5125 58. The exicyclog utility now does better if the number of log files to keep
5126 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5128 59. Two changes related to the smtp_active_hostname option:
5130 (1) $smtp_active_hostname is now available as a variable.
5131 (2) The default for smtp_banner uses $smtp_active_hostname instead
5132 of $primary_hostname.
5134 60. The host_aton() function is supposed to be passed a string that is known
5135 to be a valid IP address. However, in the case of IPv6 addresses, it was
5136 not checking this. This is a hostage to fortune. Exim now panics and dies
5137 if the condition is not met. A case was found where this could be provoked
5138 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5139 components; fortuitously, this particular loophole had already been fixed
5140 by change 4.50/55 above.
5142 If there are any other similar loopholes, the new check in host_aton()
5143 itself should stop them being exploited. The report I received stated that
5144 data on the command line could provoke the exploit when Exim was running as
5145 exim, but did not say which command line option was involved. All I could
5146 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5147 running as the user.
5150 61. There was a buffer overflow vulnerability in the SPA authentication code
5151 (which came originally from the Samba project). I have added a test to the
5152 spa_base64_to_bits() function which I hope fixes it.
5155 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5156 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5158 63. The daemon start-up calls getloadavg() while still root for those OS that
5159 need the first call to be done as root, but it missed one case: when
5160 deliver_queue_load_max is set with deliver_drop_privilege. This is
5161 necessary for the benefit of the queue runner, because there is no re-exec
5162 when deliver_drop_privilege is set.
5164 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5165 This has been fixed.
5167 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5168 was in use, was not putting the data itself into the right store pool;
5169 consequently, it could be overwritten for a subsequent message in the same
5170 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5173 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5175 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5176 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5177 the function were treating the return as a boolean value, which happened to
5178 work because 0=false and not-0=true, but is not correct code.
5180 68. The host_aton() function was not handling scoped IPv6 addresses (those
5181 with, for example, "%eth0" on the end) correctly.
5183 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5184 negated items (that is, ~something) in unsigned ints. Some compilers
5185 apparently mutter when there is no cast.
5187 70. If an address verification called from an ACL failed, and did not produce a
5188 user-specific message (i.e. there was only a "system" message), nothing was
5189 put in $acl_verify_message. In this situation, it now puts the system
5192 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5193 session; change 4.31/43 added the unwanted input to the log line - except
5194 that it did not do this in the start of session case. It now does.
5196 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5197 This is wrong and can cause the other end to generate a synchronization
5198 error if it is another Exim or anything else that does the synchronization
5199 check. A QUIT command is no longer sent after a timeout.
5201 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5202 during host lookups.
5204 74. The defer_ok option of callout verification was not working if it was used
5205 when verifying addresses in header lines, that is, for this case:
5207 verify = header_sender/callout=defer_ok
5209 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5210 those file descriptors could be used for SMTP connections. If anything
5211 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5212 could be sent to the SMTP client, causing chaos. The daemon now opens
5213 stdin, stdout, and stderr to /dev/null when it puts itself into the
5216 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5217 log by default. The user can override this with suitable Perl magic.
5219 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5220 the log message when discard triggers, was not working for the DATA ACL or
5221 for the non-SMTP ACL.
5223 78. Error message wording change in sieve.c.
5225 79. If smtp_accept_max_per_host was set, the number of connections could be
5226 restricted to fewer than expected, because the daemon was trying to set up
5227 a new connection before checking whether the processes handling previous
5228 connections had finished. The check for completed processes is now done
5229 earlier. On busy systems, this bug wouldn't be noticed because something
5230 else would have woken the daemon, and it would have reaped the completed
5233 80. If a message was submitted locally by a user whose login name contained one
5234 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5235 It caused a spool format error. I have fixed the spool reading code. A
5236 related problem was that the "from" clause in the Received: line became
5237 illegal because of the space(s). It is now covered by ${quote_local_part.
5239 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5242 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5244 83. Patch from Sieve maintainer to fix unterminated string problem in
5245 "vacation" handling.
5247 84. Some minor changes to the Linux configuration files to help with other
5248 OS variants using glibc.
5250 85. One more patch for Sieve to update vacation handling to latest spec.
5253 ----------------------------------------------------
5254 See the note above about the 4.44 and 4.50 releases.
5255 ----------------------------------------------------
5261 1. Change 4.43/35 introduced a bug that caused file counts to be
5262 incorrectly computed when quota_filecount was set in an appendfile
5265 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5266 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5269 3. Give more explanation in the error message when the command for a transport
5270 filter fails to execute.
5272 4. There are several places where Exim runs a non-Exim command in a
5273 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5274 was being done only for the command run by the queryprogram router. It is
5275 now done for all such subprocesses. The other cases are: ${run, transport
5276 filters, and the commands run by the lmtp and pipe transports.
5278 5. Some older OS have a limit of 256 on the maximum number of file
5279 descriptors. Exim was using setrlimit() to set 1000 as a large value
5280 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5281 systems. I've change it so that if it can't get 1000, it tries for 256.
5283 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5284 was an oversight, and furthermore, ever since the addition of extra
5285 controls (e.g. 4.43/32), the checks on when to allow different forms of
5286 "control" were broken. There should now be diagnostics for all cases when a
5287 control that does not make sense is encountered.
5289 7. $recipients is now available in the predata ACL (oversight).
5291 8. Tidy the search cache before the fork to do a delivery from a message
5292 received from the command line. Otherwise the child will trigger a lookup
5293 failure and thereby defer the delivery if it tries to use (for example) a
5294 cached ldap connection that the parent has called unbind on.
5296 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5297 of $address_data from the recipient verification was clobbered by the
5298 sender verification.
5300 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5301 was its contents. (It was OK if the option was not defined at all.)
5303 11. A "Completed" log line is now written for messages that are removed from
5304 the spool by the -Mrm option.
5306 12. $host_address is now set to the target address during the checking of
5307 ignore_target_hosts.
5309 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5310 being passed; this would have caused $sender_host_name to have been used if
5311 matching the list had actually called for a host name (not very likely,
5312 since this list is usually IP addresses). A host name is now passed as
5315 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5316 code that specifies a non-restarting handler (typically sigaction() in
5317 modern systems) in an attempt to fix a rare and obscure crash bug.
5319 15. Narrowed the window for a race in the daemon that could cause it to ignore
5320 SIGCHLD signals. This is not a major problem, because they are used only to
5321 wake it up if nothing else does.
5323 16. A malformed maildirsize file could cause Exim to calculate negative values
5324 for the mailbox size or file count. Odd effects could occur as a result.
5325 The maildirsize information is now recalculated if the size or filecount
5328 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5329 support for a long time. Removed HAVE_SYS_VFS_H.
5331 18. Updated exipick to current release from John Jetmore.
5333 19. Allow an empty sender to be matched against a lookup in an address list.
5334 Previously the only cases considered were a regular expression, or an
5337 20. Exim went into a mad DNS lookup loop when doing a callout where the
5338 host was specified on the transport, if the DNS lookup yielded more than
5341 21. The RFC2047 encoding function was originally intended for short strings
5342 such as real names; it was not keeping to the 75-character limit for
5343 encoded words that the RFC imposes. It now respects the limit, and
5344 generates multiple encoded words if necessary. To be on the safe side, I
5345 have increased the buffer size for the ${rfc2047: expansion operator from
5348 22. Failure to deliver a bounce message always caused it to be frozen, even if
5349 there was an errors_to setting on the router. The errors_to setting is now
5352 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5353 canonical form (fully expanded) before being placed in
5354 $sender_host_address.
5356 24. Updated eximstats to version 1.33
5358 25. Include certificate and key file names in error message when GnuTLS fails
5359 to set them up, because the GnuTLS error message doesn't include the name
5360 of the failing file when there is a problem reading it.
5362 26. Expand error message when OpenSSL has problems setting up cert/key files.
5365 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5366 (this can affect the format of dates).
5368 28. exim_tidydb, when checking for the continued existence of a message for
5369 which it has found a message-specific retry record, was not finding
5370 messages that were in split spool directories. Consequently, it was
5371 deleting retry records that should have stayed in existence.
5373 29. eximstats updated to version 1.35
5374 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5375 1.35 - bugfix such that pie charts by volume are generated correctly
5377 30. The SPA authentication driver was not abandoning authentication and moving
5378 on to the next authenticator when an expansion was forced to fail,
5379 contradicting the general specification for all authenticators. Instead it
5380 was generating a temporary error. It now behaves as specified.
5382 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5383 (the order specifies the preference for clients). The order is now AES256,
5384 AES128, 3DES, ARCFOUR128.
5386 31. Small patch to Sieve code - explicitly set From: when generating an
5389 32. Exim crashed if a remote delivery caused a very long error message to be
5390 recorded - for instance if somebody sent an entire SpamAssassin report back
5391 as a large number of 550 error lines. This bug was coincidentally fixed by
5392 increasing the size of one of Exim's internal buffers (big_buffer) that
5393 happened as part of the Exiscan merge. However, to be on the safe side, I
5394 have made the code more robust (and fixed the comments that describe what
5397 33. Some experimental protocols are using DNS PTR records for new purposes. The
5398 keys for these records are domain names, not reversed IP addresses. The
5399 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5400 leaves it alone. Component reversal etc. now happens only for IP addresses.
5403 34. The host_aton() function is supposed to be passed a string that is known
5404 to be a valid IP address. However, in the case of IPv6 addresses, it was
5405 not checking this. This is a hostage to fortune. Exim now panics and dies
5406 if the condition is not met. A case was found where this could be provoked
5407 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5408 components; fortuitously, this particular loophole had already been fixed
5409 by change 4.50/55 or 4.44/33 above.
5411 If there are any other similar loopholes, the new check in host_aton()
5412 itself should stop them being exploited. The report I received stated that
5413 data on the command line could provoke the exploit when Exim was running as
5414 exim, but did not say which command line option was involved. All I could
5415 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5416 running as the user.
5419 35. There was a buffer overflow vulnerability in the SPA authentication code
5420 (which came originally from the Samba project). I have added a test to the
5421 spa_base64_to_bits() function which I hope fixes it.
5424 36. The daemon start-up calls getloadavg() while still root for those OS that
5425 need the first call to be done as root, but it missed one case: when
5426 deliver_queue_load_max is set with deliver_drop_privilege. This is
5427 necessary for the benefit of the queue runner, because there is no re-exec
5428 when deliver_drop_privilege is set.
5430 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5431 was in use, was not putting the data itself into the right store pool;
5432 consequently, it could be overwritten for a subsequent message in the same
5433 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5436 38. Sometimes the final signoff response after QUIT could fail to get
5437 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5438 < 0 before doing a fflush(). This bug looks as though it goes back to the
5439 introduction of TLS in release 3.20, but "sometimes" must have been rare
5440 because the tests only now provoked it.
5446 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5447 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5448 Now it returns an integer. A number of calls were still expecting a T/F
5449 return. Fortuitously, in all cases, the tests worked in OK situations,
5450 which is the norm. However, things would have gone wrong on any write
5451 failures on the smtp file descriptor. This function is used when sending
5452 messages over SMTP and also when doing verify callouts.
5454 2. When Exim is called to do synchronous delivery of a locally submitted
5455 message (the -odf or -odi options), it no longer closes stderr before doing
5458 3. Implemented the mua_wrapper option.
5460 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5462 5. Implemented the functions header_remove(), header_testname(),
5463 header_add_at_position(), and receive_remove_recipient(), and exported them
5466 6. If an ACL "warn" statement specified the addition of headers, Exim already
5467 inserted X-ACL-Warn: at the start if there was no header name. However, it
5468 was not making this test for the second and subsequent header lines if
5469 there were newlines in the string. This meant that an invalid header could
5470 be inserted if Exim was badly configured.
5472 7. Allow an ACL "warn" statement to add header lines at the start or after all
5473 the Received: headers, as well as at the end.
5475 8. Added the rcpt_4xx retry error code.
5477 9. Added postmaster_mailfrom=xxx to callout verification option.
5479 10. Added mailfrom=xxxx to the callout verification option, for verify=
5482 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5483 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5485 12. Inserted some casts to stop certain compilers warning when using pointer
5486 differences as field lengths or precisions in printf-type calls (mostly
5487 affecting debugging statements).
5489 13. Added optional readline() support for -be (dynamically loaded).
5491 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5492 same clock tick as a message's arrival, so that its received time was the
5493 same as the "first fail" time on the retry record, and that message
5494 remained on the queue past the ultimate address timeout, every queue runner
5495 would try a delivery (because it was past the ultimate address timeout) but
5496 after another failure, the ultimate address timeout, which should have then
5497 bounced the address, did not kick in. This was a "< instead of <=" error;
5498 in most cases the first failure would have been in the next clock tick
5499 after the received time, and all would be well.
5501 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5502 being recognized when the domain list was tested by the match_domain
5503 condition in an expansion string.
5505 16. Added the ${str2b64: operator.
5507 17. Exim was always calling setrlimit() to set a large limit for the number of
5508 processes, without checking whether the existing limit was already
5509 adequate. (It did check for the limit on file descriptors.) Furthermore,
5510 errors from getrlimit() and setrlimit() were being ignored. Now they are
5511 logged to the main and panic logs, but Exim does carry on, to try to do its
5512 job under whatever limits there are.
5514 18. Imported PCRE 5.0.
5516 19. Trivial typo in log message " temporarily refused connection" (the leading
5519 20. If the log selector return_path_on_delivery was set and an address was
5520 redirected to /dev/null, the delivery process crashed because it assumed
5521 that a return path would always be set for a "successful" delivery. In this
5522 case, the whole delivery is bypassed as an optimization, and therefore no
5525 21. Internal re-arrangement: the function for sending a challenge and reading
5526 a response while authentication was assuming a zero-terminated challenge
5527 string. It's now changed to take a pointer and a length, to allow for
5528 binary data in such strings.
5530 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5532 23. Exim was not respecting finduser_retries when seeking the login of the
5533 uid under which it was called; it was always trying 10 times. (The default
5534 setting of finduser_retries is zero.) Also, it was sleeping after the final
5535 failure, which is pointless.
5537 24. Implemented tls_on_connect_ports.
5539 25. Implemented acl_smtp_predata.
5541 26. If the domain in control=submission is set empty, Exim assumes that the
5542 authenticated id is a complete email address when it generates From: or
5543 Sender: header lines.
5545 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5546 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5547 chown and chgrp in /bin and hostname in /usr/bin.
5549 28. Exim was keeping the "process log" file open after each use, just as it
5550 does for the main log. This opens the possibility of it remaining open for
5551 long periods when the USR1 signal hits a daemon. Occasional processlog
5552 errors were reported, that could have been caused by this. Anyway, it seems
5553 much more sensible not to leave this file open at all, so that is what now
5556 29. The long-running daemon process does not normally write to the log once it
5557 has entered its main loop, and it closes the log before doing so. This is
5558 so that log files can straightforwardly be renamed and moved. However,
5559 there are a couple of unusual error situations where the daemon does write
5560 log entries, and I had neglected to close the log afterwards.
5562 30. The text of an SMTP error response that was received during a remote
5563 delivery was being truncated at 512 bytes. This is too short for some of
5564 the long messages that one sometimes sees. I've increased the limit to
5567 31. It is now possible to make retry rules that apply only when a message has a
5568 specific sender, in particular, an empty sender.
5570 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5571 it possible to be selective about when SMTP synchronization is enforced.
5573 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5575 32. Implemented hosts_connection_nolog.
5577 33. Added an ACL for QUIT.
5579 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5582 35. Added mailbox_size and mailbox_filecount to appendfile.
5584 36. Added control = no_multiline_responses to ACLs.
5586 37. There was a bug in the logic of the code that waits for the clock to tick
5587 in the case where the clock went backwards by a substantial amount such
5588 that the microsecond fraction of "now" was more than the microsecond
5589 fraction of "then" (but the whole seconds number was less).
5591 38. Added support for the libradius Radius client library this is found on
5592 FreeBSD (previously only the radiusclient library was supported).
5598 1. When certain lookups returned multiple values in the form name=value, the
5599 quoting of the values was not always being done properly. Specifically:
5600 (a) If the value started with a double quote, but contained no whitespace,
5602 (b) If the value contained whitespace other than a space character (i.e.
5603 tabs or newlines or carriage returns) it was not quoted.
5604 This fix has been applied to the mysql and pgsql lookups by writing a
5605 separate quoting function and calling it from the lookup code. The fix
5606 should probably also be applied to nisplus, ibase and oracle lookups, but
5607 since I cannot test any of those, I have not disturbed their existing code.
5609 2. A hit in the callout cache for a specific address caused a log line with no
5610 reason for rejecting RCPT. Now it says "Previous (cached) callout
5611 verification failure".
5613 3. There was an off-by-one bug in the queryprogram router. An over-long
5614 return line was truncated at 256 instead of 255 characters, thereby
5615 overflowing its buffer with the terminating zero. As well as fixing this, I
5616 have increased the buffer size to 1024 (and made a note to document this).
5618 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5619 when Exim is waiting for an SMTP response from a remote server, Exim
5620 restarts its select() call on the socket, thereby resetting its timeout.
5621 This is not a problem when such interrupts are rare. Somebody set up a cron
5622 job to run exiwhat every 2 minutes, which is less than the normal select()
5623 timeout (5 or 10 minutes). This meant that the select() timeout never
5624 kicked in because it was always reset. I have fixed this by comparing the
5625 time when an interrupt arrives with the time at the start of the first call
5626 to select(). If more time than the timeout has elapsed, the interrupt is
5627 treated as a timeout.
5629 5. Some internal re-factoring in preparation for the addition of Sieve
5630 extensions (by MH). In particular, the "personal" test is moved to a
5631 separate function, and given an option for scanning Cc: and Bcc: (which is
5632 not set for Exim filters).
5634 6. When Exim created an email address using the login of the caller as the
5635 local part (e.g. when creating a From: or Sender: header line), it was not
5636 quoting the local part when it contained special characters such as @.
5638 7. Installed new OpenBSD configuration files.
5640 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5641 try to make them clearer.
5643 9. Callout options, other than the timeout value, were being ignored when
5644 verifying sender addresses in header lines. For example, when using
5646 verify = header_sender/callout=no_cache
5648 the cache was (incorrectly) being used.
5650 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5651 only the Cygwin environment.
5653 11. When return_path_on_delivery was set as a log selector, if different remote
5654 addresses in the same message used different return paths and parallel
5655 remote delivery occurred, the wrong values would sometimes be logged.
5656 (Whenever a remote delivery process finished, the return path value from
5657 the most recently started remote delivery process was logged.)
5659 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5660 lines when AUTH and/or TLS are in use. This is the "received protocol"
5661 field. Exim used to use "asmtp" for authenticated SMTP, without any
5662 indication (in the protocol name) for TLS use. Now it follows the RFC and
5663 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5664 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5665 names appear in log lines as well as in Received: header lines.
5667 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5668 extensions, and comparison tests, and to fix some bugs.
5670 14. Changes to the "personal" filter test:
5672 (1) The test was buggy in that it was just doing the equivalent of
5673 "contains" tests on header lines. For example, if a user's address was
5674 anne@some.where, the "personal" test would incorrectly be true for
5676 To: susanne@some.where
5678 This test is now done by extracting each address from the header in turn,
5679 and checking the entire address. Other tests that are part of "personal"
5680 are now done using regular expressions (for example, to check local parts
5681 of addresses in From: header lines).
5683 (2) The list of non-personal local parts in From: addresses has been
5684 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5685 taken from the Sieve specification recommendations.
5687 (3) If the message contains any header line starting with "List-" it is
5688 treated as non-personal.
5690 (4) The test for "circular" in the Subject: header line has been removed
5691 because it now seems ill-conceived.
5693 15. Minor typos in src/EDITME comments corrected.
5695 16. Installed latest exipick from John Jetmore.
5697 17. If headers_add on a router specified a text string that was too long for
5698 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5699 of string_sprintf() is now avoided.
5701 18. $message_body_size was not set (it was always zero) when running the DATA
5702 ACL and the local_scan() function.
5704 19. For the "mail" command in an Exim filter, no default was being set for
5705 the once_repeat time, causing a random time value to be used if "once" was
5706 specified. (If the value happened to be <= 0, no repeat happened.) The
5707 default is now 0s, meaning "never repeat". The "vacation" command was OK
5708 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5709 (I found it when inspecting the code).
5711 20. There is now an overall timeout for performing a callout verification. It
5712 defaults to 4 times the callout timeout, which applies to individual SMTP
5713 commands during the callout. The overall timeout applies when there is more
5714 than one host that can be tried. The timeout is checked before trying the
5715 next host. This prevents very long delays if there are a large number of
5716 hosts and all are timing out (e.g. when the network connections are timing
5717 out). The value of the overall timeout can be changed by specifying an
5718 additional sub-option for "callout", called "maxwait". For example:
5720 verify = sender/callout=5s,maxwait=20s
5722 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5723 to the end before writing, but this should make it even safer).
5725 22. Exim was forgetting that it had advertised PIPELINING for the second and
5726 subsequent messages on an SMTP connection. It was also not resetting its
5727 memory on STARTTLS and an internal HELO.
5729 23. When Exim logs an SMTP synchronization error within a session, it now
5730 records whether PIPELINING has been advertised or not.
5732 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5733 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5734 rather than long int.
5736 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5738 26. Added the never_mail option to autoreply.
5744 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5745 crash if the getsockname() call failed; this can happen if a connection is
5746 closed very soon after it is established. The problem was simply in the
5747 order in which certain operations were done, causing Exim to try to write
5748 to the SMTP stream before it had set up the file descriptor. The bug has
5749 been fixed by making things happen in the correct order.
5755 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5756 before the connection was closed, thus losing the rejection response.
5758 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5759 some early Solaris releases, but causes trouble in current releases where
5760 socklen_t is defined.
5762 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5765 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5768 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5769 that releasing the top of it at the end releases what was used for sub-
5770 expansions (unless the block got too big). However, discard this block if
5771 the first thing is a variable or header, so that we can use its block when
5772 it is dynamic (useful for very large $message_headers, for example).
5774 6. Lookups now cache *every* query, not just the most recent. A new, separate
5775 store pool is used for this. It can be recovered when all lookup caches are
5776 flushed. Lookups now release memory at the end of their result strings.
5777 This has involved some general refactoring of the lookup sources.
5779 7. Some code has been added to the store_xxx() functions to reduce the amount
5780 of flapping under certain conditions.
5782 8. log_incoming_interface used to affect only the <= reception log lines. Now
5783 it causes the local interface and port to be added to several more SMTP log
5784 lines, for example "SMTP connection from", and rejection lines.
5786 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5788 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5790 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5791 could be overwritten at the end of the current message (or the start of a
5792 new message if it was set in a HELO ACL). The value is now preserved for
5793 the duration of the SMTP connection.
5795 12. If a transport had a headers_rewrite setting, and a matching header line
5796 contained an unqualified address, that address was qualified, even if it
5797 did not match any rewriting rules. The underlying bug was that the values
5798 of the flags that permit the existence of unqualified sender and recipient
5799 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5800 non-local messages, and by -bnq for local messages) were not being
5801 preserved with the message after it was received.
5803 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5804 "next input=" as part of the text comprising the host identity instead of
5805 the correct text. The code was using the same buffer for two different
5806 strings. However, depending on which order the printing function evaluated
5807 its arguments, the bug did not always show up. Under Linux, for example, my
5808 test suite worked just fine.
5810 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5811 doesn't work with some older versions of Perl. It has been changed to "my",
5812 which in any case is probably the better facility to use.
5814 15. A really picky compiler found some instances of statements for creating
5815 error messages that either had too many or two few arguments for the format
5818 16. The size of the buffer for calls to the DNS resolver has been increased
5819 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5820 for addresses that have a lot of PTR records. This alleviates a problem; it
5821 does not fully solve it.
5823 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5824 buffer now truncates the list and logs the incident, which is the same
5825 action as happens when Exim is looking up a host name and its aliases.
5826 Previously in this situation something unpredictable would happen;
5827 sometimes it was "internal error: store_reset failed".
5829 18. If a server dropped the connection unexpectedly when an Exim client was
5830 using GnuTLS and trying to read a response, the client delivery process
5831 crashed while trying to generate an error log message.
5833 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5834 string, for example:
5836 warn message = H1: something\nH2: something
5838 the text was added as a single header line from Exim's point of view
5839 though it ended up OK in the delivered message. However, searching for the
5840 second and subsequent header lines using $h_h2: did not work. This has been
5841 fixed. Similarly, if a system filter added multiple headers in this way,
5842 the routers could not see them.
5844 20. Expanded the error message when iplsearch is called with an invalid key to
5845 suggest using net-iplsearch in a host list.
5847 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5848 ACLs are no longer actually imposed (and a message to that effect is
5851 22. If a "gecos" field in a passwd entry contained escaped characters, in
5852 particular, if it contained a \" sequence, Exim got it wrong when building
5853 a From: or a Sender: header from that name. A second bug also caused
5854 incorrect handling when an unquoted " was present following a character
5855 that needed quoting.
5857 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5858 was not being matched caselessly.
5860 24. Arranged for all hyphens in the exim.8 source to be escaped with
5863 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5864 properly. Recipient callout cache records were still being keyed to include
5865 the sender, even when use_sender was set false. This led to far more
5866 callouts that were necessary. The sender is no longer included in the key
5867 when use_sender is false.
5869 26. Added "control = submission" modifier to ACLs.
5871 27. Added the ${base62d: operator to decode base 62 numbers.
5873 28. dnsdb lookups can now access SRV records.
5875 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5876 the configuration file.
5878 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5879 (-v) mode. This makes the output for a verbose queue run more intelligible.
5881 31. Added a use_postmaster feature to recipient callouts.
5883 32. Added the $body_zerocount variable, containing the number of binary zero
5884 bytes in the message body.
5886 33. The time of last modification of the "new" subdirectory is now used as the
5887 "mailbox time last read" when there is a quota error for a maildir
5890 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5892 35. Added +ignore_unknown as a special item in host lists.
5894 36. Code for decoding IPv6 addresses in host lists is now included, even if
5895 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5896 address was recognized as an IP address, but was then not correctly decoded
5897 into binary, causing unexpected and incorrect effects when compared with
5904 1. Very minor rewording of debugging text in manualroute to say "list of
5905 hosts" instead of "hostlist".
5907 2. If verify=header_syntax was set, and a header line with an unqualified
5908 address (no domain) and a large number of spaces between the end of the
5909 name and the colon was received, the reception process suffered a buffer
5910 overflow, and (when I tested it) crashed. This was caused by some obsolete
5911 code that should have been removed. The fix is to remove it!
5913 3. When running in the test harness, delay a bit after writing a bounce
5914 message to get a bit more predictability in the log output.
5916 4. Added a call to search_tidyup() just before forking a reception process. In
5917 theory, someone could use a lookup in the expansion of smtp_accept_max_
5918 per_host which, without the tidyup, could leave open a database connection.
5920 5. Added the variables $recipient_data and $sender_data which get set from a
5921 lookup success in an ACL "recipients" or "senders" condition, or a router
5922 "senders" option, similar to $domain_data and $local_part_data.
5924 6. Moved the writing of debug_print from before to after the "senders" test
5927 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5928 problems for message scanning, either using a data ACL, or using
5929 local_scan() because the Received: header was not generated till after they
5930 were called (in order to set the time as the time of reception completion).
5931 I have revised the way this works. The header is now generated after the
5932 body is received, but before the ACL or local_scan() are called. After they
5933 are run, the timestamp in the header is updated.
5939 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5940 before starting a queue runner without re-exec. This happened only when
5941 deliver_drop_privilege was set or when the Exim user was set to root. The
5942 effect of the bug was that timeouts during subsequent deliveries caused
5943 crashes instead of being properly handled. The handler is now left at its
5944 default (and expected) setting.
5946 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5947 message, again when deliver_drop_privilege is set or Exim is run as root.
5948 The bug described in (1) was not present in this case, but the tidying up
5949 of the other signals was missing. I have made the two cases consistent.
5951 3. The ignore_target_hosts setting on a manualroute router was being ignored
5952 for hosts that were looked up using the /MX notation.
5954 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5957 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5958 operated on the sender address. After changing the $sender_address to <>
5959 for the sender address verify, Exim was re-instated it as the original
5960 (before rewriting) address, but remembering that it had rewritten it, so it
5961 wasn't rewriting it again. This bug also had the effect of breaking the
5962 sender address verification caching when the sender address was rewritten.
5964 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5965 This has been changed so that if the ip literal address matches
5966 ignore_target_hosts, the router declines.
5968 7. Added expansion conditions match_domain, match_address, and match_local_
5969 part (NOT match_host).
5971 8. The placeholder for the Received: header didn't have a length field set.
5973 9. Added code to Exim itself and to exim_lock to test for a specific race
5974 condition that could lead to file corruption when using MBX delivery. The
5975 issue is with the lockfile that is created in /tmp. If this file is removed
5976 after a process has opened it but before that process has acquired a lock,
5977 there is the potential for a second process to recreate the file and also
5978 acquire a lock. This could lead to two Exim processes writing to the file
5979 at the same time. The added code performs the same test as UW imapd; it
5980 checks after acquiring the lock that its file descriptor still refers to
5981 the same named file.
5983 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5984 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5987 11. Added the smtp_active_hostname option. If used, this will typically be made
5988 to depend on the incoming interface address. Because $interface_address is
5989 not set up until the daemon has forked a reception process, error responses
5990 that can happen earlier (such as "too many connections") no longer contain
5993 12. If an expansion in a condition on a "warn" statement fails because a lookup
5994 defers, the "warn" statement is abandoned, and the next ACL statement is
5995 processed. Previously this caused the whole ACL to be aborted.
5997 13. Added the iplsearch lookup type.
5999 14. Added ident_timeout as a log selector.
6001 15. Added tls_certificate_verified as a log selector.
6003 16. Added a global option tls_require_ciphers (compare the smtp transport
6004 option of the same name). This controls incoming TLS connections.
6006 17. I finally figured out how to make tls_require_ciphers do a similar thing
6007 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6008 before starting the TLS session.
6010 18. Tabs are now shown as \t in -bP output.
6012 19. If the log selector return_path_on_delivery was set, Exim crashed when
6013 bouncing a message because it had too many Received: header lines.
6015 20. If two routers both had headers_remove settings, and the first one included
6016 a superfluous trailing colon, the final name in the first list and the
6017 first name in the second list were incorrectly joined into one item (with a
6018 colon in the middle).
6024 1. Added -C and -D options to the exinext utility, mainly to make it easier
6025 to include in the automated testing, but these could be helpful when
6026 multiple configurations are in use.
6028 2. The exinext utility was not formatting the output nicely when there was
6029 an alternate port involved in the retry record key, nor when there was a
6030 message id as well (for retries that were specific to a specific message
6031 and a specific host). It was also confused by IPv6 addresses, because of
6032 the additional colons they contain. I have fixed the IPv4 problem, and
6033 patched it up to do a reasonable job for IPv6.
6035 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6036 delivery, the log line now contains "pipelined" if PIPELINING was used.
6038 4. An SMTP transport process used to panic and die if the bind() call to set
6039 an explicit outgoing interface failed. This has been changed; it is now
6040 treated in the same way as a connect() failure.
6042 5. A reference to $sender_host_name in the part of a conditional expansion
6043 that was being skipped was still causing a DNS lookup. This no longer
6046 6. The def: expansion condition was not recognizing references to header lines
6047 that used bh_ and bheader_.
6049 7. Added the _cache feature to named lists.
6051 8. The code for checking quota_filecount in the appendfile transport was
6052 allowing one more file than it should have been.
6054 9. For compatibility with Sendmail, the command line option
6062 and sets the incoming protocol and host name (for trusted callers). The
6063 host name and its colon can be omitted when only the protocol is to be set.
6064 Note the Exim already has two private options, -pd and -ps, that refer to
6065 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6066 "s", but I don't think that's a major issue.
6068 10. A number of refactoring changes to the code, none of which should affect
6071 (a) The number of logging options was getting close to filling up the
6072 32-bit word that was used as a bit map. I have split them into two classes:
6073 those that are passed in the argument to log_write(), and those that are
6074 only ever tested independently outside of that function. These are now in
6075 separate 32-bit words, so there is plenty of room for expansion again.
6076 There is no change in the user interface or the logging behaviour.
6078 (b) When building, for example, log lines, the code previously used a
6079 macro that called string_cat() twice, in order to add two strings. This is
6080 not really sufficiently general. Furthermore, there was one instance where
6081 it was actually wrong because one of the argument was used twice, and in
6082 one call a function was used. (As it happened, calling the function twice
6083 did not affect the overall behaviour.) The macro has been replaced by a
6084 function that can join an arbitrary number of extra strings onto a growing
6087 (c) The code for expansion conditions now uses a table and a binary chop
6088 instead of a serial search (which was left over from when there were very
6089 few conditions). Also, it now recognizes conditions like "pam" even when
6090 the relevant support is not compiled in: a suitably worded error message is
6091 given if an attempt is made to use such a condition.
6093 11. Added ${time_interval:xxxxx}.
6095 12. A bug was causing one of the ddress fields not to be passed back correctly
6096 from remote delivery subprocesses. The field in question was not being
6097 subsequently used, so this caused to problems in practice.
6099 13. Added new log selectors queue_time and deliver_time.
6101 14. Might have fixed a bug in maildirsizefile handling that threw up
6102 "unexpected character" debug warnings, and recalculated the data
6103 unnecessarily. In any case, I expanded the warning message to give more
6106 15. Added the message "Restricted characters in address" to the statements in
6107 the default ACL that block characters like @ and % in local parts.
6109 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6110 Three changes have been made:
6112 (a) There was a serious bug; a negative response to MAIL caused the whole
6113 recipient domain to be cached as invalid, thereby blocking all messages
6114 to all local parts at the same domain, from all senders. This bug has
6115 been fixed. The domain is no longer cached after a negative response to
6116 MAIL if the sender used is not empty.
6118 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6121 (c) A new callout option, "use_sender" has been added for people who want
6122 the modified behaviour.
6128 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6131 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6132 indeed breaks things for older releases.
6134 3. Added additional logging to the case where there is a problem reading data
6135 from a filter that is running in a subprocess using a pipe, in order to
6136 try to track down a specific problem.
6138 4. Testing facility fudge: when running in the test harness and attempting
6139 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6140 getting "No route to host". Convert this to a timeout.
6142 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6145 6. Some OS don't have socklen_t but use size_t instead. This affects the
6146 fifth argument of getsockopt() amongst other things. This is now
6147 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6148 can be set for individual OS. I have set it for SunOS5, OSF1, and
6149 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6150 some earlier ones do not.
6152 7. Change 4.30/15 was not doing the test caselessly.
6154 8. The standard form for an IPv6 address literal was being rejected by address
6155 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6156 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6157 this, as well as the form without the "IPv6" on the front (but only when
6158 address literals are enabled, of course).
6160 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6162 10. Exim crashed if a message with an empty sender address specified by -f
6163 encountered a router with an errors_to setting. This could be provoked only
6164 by a command such as
6168 where an empty string was supplied; "<>" did not hit this bug.
6170 11. Installed PCRE release 4.5.
6172 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6173 remained set. It is now erased.
6175 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6176 times from message ids (which are base 36 rather than the normal 62).
6178 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6179 were being counted as actual protocol errors, and logged if the log
6180 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6181 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6182 and DATA following a set of rejected RCPTs do not count as protocol errors.
6183 In other words, Exim assumes they were pipelined, though this may not
6184 actually be the case. Of course, in all cases the client gets an
6185 appropriate error code.
6187 15. If a lookup fails in an ACL condition, a message about the failure may
6188 be available; it is used if testing the ACL cannot continue, because most
6189 such messages specify what the cause of the deferral is. However, some
6190 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6191 that caused an old message to be retained and used if a later statement
6192 caused a defer, replacing the real cause of the deferral.
6194 16. If an IP address had so many PTR records that the DNS lookup buffer
6195 was not large enough to hold them, Exim could crash while trying to process
6196 the truncated data. It now detects and logs this case.
6198 17. Further to 4.21/58, another change has been made: if (and only if) the
6199 first line of a message (the first header line) ends with CRLF, a bare LF
6200 in a subsequent header line has a space inserted after it, so as not to
6201 terminate the header.
6203 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6204 unnecessarily, used atoi() instead of strtol(), and didn't check the
6205 termination when getting file sizes from file names by regex.
6207 19. Completely re-implemented the support for maildirsize files, in the light
6208 of a number of problems with the previous contributed implementation
6209 (4.30/29). In particular:
6211 . If the quota is zero, the maildirsize file is maintained, but no quota is
6214 . If the maildir directory does not exist, it is created before any attempt
6215 to write a maildirsize file.
6217 . The quota value in the file is just a cache; if the quota is changed in
6218 the transport, the new value overrides.
6220 . A regular expression is available for excluding directories from the
6223 20. The autoreply transport checks the characters in options that define the
6224 message's headers; it allows continued headers, but it was checking with
6225 isspace() after an embedded newline instead of explicitly looking for a
6228 21. If all the "regular" hosts to which an address was routed had passed their
6229 expiry times, and had not reached their retry times, the address was
6230 bounced, even if fallback hosts were defined. Now Exim should go on to try
6233 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6234 equivalent code in the SMTP transport. Some hosts send humungous responses
6235 to HELO/EHLO, more than 1024 it seems.
6237 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6238 gives compiler warnings in some environments. I've now done it "properly",
6241 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6242 (because of problems with the built-in one) was declared to return uschar *
6243 instead of char *, causing compiler failure.
6245 25. Fixed a file descriptor leak when processing alias/forward files.
6247 26. Fixed a minor format string issue in dbfn.c.
6249 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6251 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6252 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6255 29. When a sender address is verified, it is cached, to save repeating the test
6256 when there is more than one recipient in a message. However, when the
6257 verification involves a callout, it is possible for different callout
6258 options to be set for different recipients. It is too complicated to keep
6259 track of this in the cache, so now Exim always runs a verification when a
6260 callout is required, relying on the callout cache for the optimization.
6261 The overhead is duplication of the address routing, but this should not be
6264 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6265 to be verified with callout=postmaster, and the main callout worked but the
6266 postmaster check failed, the verification correctly failed. However, if a
6267 subsequent RCPT command asked for sender verification *without* the
6268 postmaster check, incorrect caching caused this verification also to fail,
6271 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6272 it was not caching the DNS options (qualify_single, search_parents) that
6273 were used when the lookup failed. A subsequent lookup with different
6274 options therefore always gave the same answer, though there were cases
6275 where it should not have. (Example: a "domains = !$mx_any" option on a
6276 dnslookup router: the "domains" option is always processed without any
6277 widening, but the router might have qualify_single set.) Now Exim uses the
6278 cached value only when the same options are set.
6280 32. Added John Jetmore's "exipick" utility to the distribution.
6282 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6283 than a timeout (e.g. a certificate is required, and is not provided), an
6284 Exim server now closes the connection immediately. Previously it waited for
6285 the client to close - but if the client is SSL, it seems that they each
6286 wait for each other, leading to a delay before one of them times out.
6288 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6289 maintained 0.8.x compatibility because I don't think many are using it, and
6290 it is clearly obsolete.
6292 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6295 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6296 client certificate was expired. A simple patch fixes this, though I don't
6297 understand the full logic of why the verify callback is called multiple
6300 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6301 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6302 which causes problems with some clients (such as the Certicom SSL Plus
6303 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6304 disables the coutermeasure allowing Eudora to connect."
6306 38. Exim was not checking that a write() to a log file succeeded. This could
6307 lead to Bad Things if a log got too big, in particular if it hit a file
6308 size limit. Exim now panics and dies if it cannot write to a log file, just
6309 as it does if it cannot open a log file.
6311 39. Modified OS/Makefile-Linux so that it now contains
6313 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6315 The two -D definitions ensure that Exim is compiled with large file
6316 support, which makes it possible to handle log files that are bigger than
6319 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6320 instance) a domain was checked against a named list that involved a lookup,
6321 causing $domain_data to be set, then another domain was checked against the
6322 same list, then the first domain was re-checked, the value of $domain_data
6323 after the final check could be wrong. In particular, if the second check
6324 failed, it could be set empty. This bug probably also applied to
6327 41. The strip_trailing_dot option was not being applied to the address given
6328 with the -f command-line option.
6330 42. The code for reading a message's header from the spool was incrementing
6331 $received_count, but never initializing it. This meant that the value was
6332 incorrect (doubled) while delivering a message in the same process in which
6333 it was received. In the most common configuration of Exim, this never
6334 happens - a fresh exec is done - but it can happen when
6335 deliver_drop_privilege is set.
6337 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6338 it now includes up to 150 characters of the unexpected data in the log
6341 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6342 and building data strings. The size of both of these buffers was 10 000
6343 bytes - far larger than anybody would *ever* want, thought I. Needless to
6344 say, somebody hit the limit. I have increased the maximum line length to
6345 20 000 and the maximum data length of concatenated lines to 100 000. I have
6346 also fixed two bugs, because there was no checking on these buffers. Tsk,
6347 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6348 buffer is too small.
6350 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6351 lsearch lookups. Now it does.
6353 46. When parsing a route_list item in a manualroute router, a fixed-length
6354 buffer was used for the list of hosts. I made this 1024 bytes long,
6355 thinking that nobody would ever have a list of hosts that long. Wrong.
6356 Somebody had a whole pile of complicated expansion conditions, and the
6357 string was silently truncated, leading to an expansion error. It turns out
6358 that it is easier to change to an unlimited length (owing to other changes
6359 that have happened since this code was originally written) than to build
6360 structure for giving a limitation error. The length of the item that
6361 expands into the list of hosts is now unlimited.
6363 47. The lsearch lookup could not handle data where the length of text line was
6364 more than 4095 characters. Such lines were truncated, leading to shortened
6365 data being returned. It should now handle lines of any length.
6367 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6368 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6371 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6372 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6373 no longer happens in comment lines. A list of replacements is now placed
6374 at the head of all of the source files, except those whose only change is
6375 to replace PERL_COMMAND in the very first #! line.
6377 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6378 messages on the queue, with a bottom-up merge sort, using code contributed
6379 by Michael Haardt. This should make operations like -bp somewhat faster on
6380 large queues. It won't affect queue runners, except when queue_run_in_order
6383 51. Installed eximstats 1.31 in the distribution.
6385 52. Added support for SRV lookups to the dnslookup router.
6387 53. If an ACL referred to $message_body or $message_body_end, the value was not
6388 reset for any messages that followed in the same SMTP session.
6390 54. The store-handling optimization for building very long strings was not
6391 differentiating between the different store pools. I don't think this
6392 actually made any difference in practice, but I've tidied it.
6394 55. While running the routers to verify a sender address, $sender_address
6395 was still set to the sender address. This is wrong, because when routing to
6396 send a bounce to the sender, it would be empty. Therefore, I have changed
6397 it so that, while verifying a sender address, $sender_address is set to <>.
6398 (There is no change to what happens when verifying a recipient address.)
6400 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6401 target A or AAAA records (if not already returned) without resetting the
6402 qualify_single or search_parents options of the DNS resolver. These are
6403 inappropriate in this case because the targets of MX and SRV records must
6404 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6405 target that, when qualified, matched something in the local domain. These
6406 two options are now turned off when doing these lookups.
6408 57. It seems that at least some releases of Reiserfs (which does not have the
6409 concept of a fixed number of inodes) returns zero and not -1 for the
6410 number of available inodes. This interacted badly with check_spool_inodes,
6411 which assumed that -1 was the "no such thing" setting. What I have done is
6412 to check that the total number of inodes is greater than zero before doing
6413 the test of how many are available.
6415 58. When a "warn" ACL statement has a log_message modifier, the message is
6416 remembered, and not repeated. This is to avoid a lot of repetition when a
6417 message has many recipients that cause the same warning to be written.
6418 However, Exim was preserving the list of already written lines for an
6419 entire SMTP session, which doesn't seem right. The memory is now reset if a
6420 new message is started.
6422 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6423 debugging output unless log_rewrite was also set.
6425 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6426 of (int)(handle) when we know that handle contains (void *)(-1).
6428 61. The Exim daemon panic-logs an error return when it closes the incoming
6429 connection. However "connection reset by peer" seems to be common, and
6430 isn't really an error worthy of noting specially, so that particular error
6433 62. When Exim is trying to find all the local interfaces, it used to panic and
6434 die if the ioctl to get the interface flags failed. However, it seems that
6435 on at least one OS (Solaris 9) it is possible to have an interface that is
6436 included in the list of interfaces, but for which you get a failure error
6437 for this call. This happens when the interface is not "plumbed" into a
6438 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6439 failure of the "get flags" call assumes that the interface is down.
6441 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6442 makes life easier for people who are doing arithmetic on fields extracted
6443 from dates, where you often get leading zeros that should not be
6444 interpreted as octal.
6446 64. Added qualify_domain to the redirect router, to override the global
6449 65. If a pathologically long header line contained very many addresses (the
6450 report of this problem mentioned 10 000) and each of them was rewritten,
6451 Exim could use up a very large amount of memory. (It kept on making new
6452 copies of the header line as it rewrote, and never released the old ones.)
6453 At the expense of a bit more processing, the header rewriting function has
6454 been changed so that it no longer eats memory in this way.
6456 66. The generation of the Received: header has been moved from the time that a
6457 message starts to be received, to the time that it finishes. The timestamp
6458 in the Received: header should now be very close to that of the <= log
6459 line. There are two side-effects of this change:
6461 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6462 logged header lines no longer include the local Received: line, because
6463 it has not yet been created. The same applies to a copy of the message
6464 that is returned to a non-SMTP sender when a message is rejected.
6466 (b) When a filter file is tested using -bf, no additional Received: header
6467 is added to the test message. After some thought, I decided that this
6470 This change does not affect the value of $received_for. It is still set
6471 after address rewriting, but before local_scan() is called.
6473 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6475 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6476 gave an unhelpful panic error message, and a defer error. I have managed to
6477 change this behaviour so that it now rejects any supplied certificate,
6478 which seems right, as the list of acceptable certificates is empty.
6480 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6481 gave an unhelpful defer error. I have not managed to make this reject any
6482 supplied certificates, but the error message it gives is "no certificate
6483 supplied", which is not helpful.
6485 70. exigrep's output now also includes lines that are not associated with any
6486 message, but which match the given pattern. Implemented by a patch from
6487 Martin Sluka, which also tidied up the Perl a bit.
6489 71. Recipient callout verification, like sender verification, was using <> in
6490 the MAIL FROM command. This isn't really the right thing, since the actual
6491 sender may affect whether the remote host accepts the recipient or not. I
6492 have changed it to use the actual sender in the callout; this means that
6493 the cache record is now keyed on a recipient/sender pair, not just the
6494 recipient address. There doesn't seem to be a real danger of callout loops,
6495 since a callout by the remote host to check the sender would use <>.
6496 [SEE ABOVE: changed after hitting problems.]
6498 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6499 temporary errors. However, in the case of such a code being given after
6500 the end of a data transmission (i.e. after ".") Exim was failing to write
6501 a retry record for the message. (Yes, there was some broken host that was
6502 actually sending 8xx at this point.)
6504 73. An unknown lookup type in a host list could cause Exim to panic-die when
6505 the list was checked. (An example that provoked this was putting <; in the
6506 middle of a list instead of at the start.) If this happened during a DATA
6507 ACL check, a -D file could be left lying around. This kind of configuration
6508 error no longer causes Exim to die; instead it causes a defer error. The
6509 incident is still logged to the main and panic logs.
6511 74. Buglet left over from Exim 3 conversion. The message "too many messages
6512 in one connection" was written to the rejectlog but not the mainlog, except
6513 when address rewriting (yes!) was being logged.
6515 75. Added write_rejectlog option.
6517 76. When a system filter was run not as root (that is, when system_filter_user
6518 was set), the values of the $n variables were not being returned to the
6519 main process; thus, they were not subsequently available in the $sn
6522 77. Added +return_path_on_delivery log selector.
6524 78. A connection timeout was being treated differently from recipients deferred
6525 when testing hosts_max_try with a message that was older than the host's
6526 retry timeout. (The host should not be counted, thus allowing all hosts to
6527 be tried at least once before bouncing.) This may have been the cause of an
6528 occasionally reported bug whereby a message would remain on the queue
6529 longer than the retry timeout, but would be bounced if a delivery was
6530 forced. I say "may" because I never totally pinned down the problem;
6531 setting up timeout/retry tests is difficult. See also the next item.
6533 79. The ultimate address timeout was not being applied to errors that involved
6534 a combination of host plus message (for example, a timeout on a MAIL
6535 command). When an address resolved to a number of possible hosts, and they
6536 were not all tried for each delivery (e.g. because of hosts_max_try), a
6537 message could remain on the queue longer than the retry timeout.
6539 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6542 81. Fixed an obscure SMTP outgoing bug which required at least the following
6543 conditions: (a) there was another message waiting for the same server;
6544 (b) the server returned 5xx to all RCPT commands in the first message so
6545 that the message was not completed; (c) the server dropped the connection
6546 or gave a negative response to the RSET that Exim sends to abort the
6547 transaction. The observed case was a dropped connection after DATA that had
6548 been sent in pipelining mode. That is, the server had advertised PIPELINING
6549 but was not implementing it correctly. The effect of the bug was incorrect
6550 behaviour, such as trying another host, and this could lead to a crash.
6556 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6557 and daemon.c were passed as pointers to ints; they should have been
6558 pointers to socklen_t variables (which are typically unsigned ints).
6560 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6563 3. Fixed a really odd bug that affected only the testing scheme; patching a
6564 certain fixed string in the binary changed the value of another string that
6565 happened to be identical to the end of the original first string.
6567 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6568 name", it returns that address as the IP address. On some operating
6569 systems (e.g. Solaris), it also passes back the IP address string as the
6570 "host name". However, on others (e.g. Linux), it passes back an empty
6571 string. Exim wasn't checking for this, and was changing the host name to an
6572 empty string, assuming it had been canonicalized.
6574 5. Although rare, it is permitted to have more than one PTR record for a given
6575 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6576 all the names associated with an address, because they do in Solaris.
6577 However, it seems that they do not in Linux for data that comes from the
6578 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6579 I found this out when I moved to a new Linux workstation and tried to run
6580 the Exim test suite.
6582 To get round this problem I have changed the code so that it now does its
6583 own call to the DNS to look up PTR records when searching for a host name.
6584 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6585 addresses that are only in /etc/hosts are still found.
6587 This behaviour is, however, controlled by an option called host_lookup_
6588 order, which defaults to "bydns:byaddr". If people want to use the other
6589 order, or indeed, just use one or the other means of lookup, they can
6590 specify it in this variable.
6592 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6593 some operating systems, this comes back from gethostbyaddr() as an empty
6594 string, and this is what Exim used to test for. However, it seems that in
6595 other systems, "." is yielded. Exim now tests for this case too.
6597 7. The values of check_spool_space and check_log_space are now held internally
6598 as a number of kilobytes instead of an absolute number of bytes. If a
6599 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6600 kilobyte. This means that much larger values can be stored.
6602 8. Exim monitor: an attempt to get the action menu when not actually pointing
6603 at a message produces an empty menu entitled "No message selected". This
6604 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6605 no entries in it ("Shell widget menu has zero width and/or height"). So I
6606 have added a single, blank menu entry in this case.
6608 9. Added ${quote_local_part.
6610 10. MIME decoding is now applied to the contents of Subject: header lines when
6613 11. Now that a reference to $sender_host_address automatically causes a reverse
6614 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6615 host lookup before query-style lookups in lists that might use this
6616 variable. This has therefore been abolished, and the "net-" prefix is no
6617 longer necessary for query-style lookups.
6619 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6620 have been a typo for LFLAGS, so it has been changed.
6622 13. The install script calls Exim with "-C /dev/null" in order to find the
6623 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6624 to be output. However, since Exim outputs its version number before the
6625 error, it didn't break the script. It just looked ugly. I fixed this by
6626 always allowing "-C /dev/null" if the caller is root.
6628 14. Ignore overlarge ACL variable number when reading spool file - insurance
6629 against a later release with more variables having written the file.
6631 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6632 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6633 this, as well as the form without the "IPv6" on the front.
6635 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6636 OS/Makefile-Darwin file.
6638 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6639 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6641 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6642 variable. However, some OS use ints for inodes. I've added cast to long int
6643 to get rid of the compiler warning.
6645 19. I had forgotten to lock out "/../" in configuration file names when
6646 ALT_CONFIG_PREFIX was set.
6648 20. Routers used for verification do not need to specify transports. However,
6649 if such a router generated a host list, and callout was configured, Exim
6650 crashed, because it could not find a port number from the (non-existent)
6651 transport. It now assumes port 25 in this circumstance.
6653 21. Added the -t option to exigrep.
6655 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6656 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6657 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6658 them is set, LOOKUP_LSEARCH is forced.
6660 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6662 24. Added sender and host information to the "rejected by local_scan()" log
6663 line; previously there was no indication of these.
6665 25. Added .include_if_exists.
6667 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6668 Linux. It turns out that not all file systems support this. Apparently some
6669 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6670 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6671 not supported on the file descriptor, is now ignored when Exim is trying to
6672 sync a directory. This applies only to Linux.
6674 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6676 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6677 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6680 29. Added support for maildirsize files from supplied patch (modified a bit).
6682 30. The use of :fail: followed by an empty string could lead Exim to respond to
6683 sender verification failures with (e.g.):
6685 550 Verification failed for <xxx>
6686 550 Sender verify failed
6688 where the first response line was missing the '-' that indicates it is not
6689 the final line of the response.
6691 31. The loop for finding the name of the user that called Exim had a hardwired
6692 limit of 10; it now uses the value of finduser_retries, which is used for
6693 all other user lookups.
6695 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6698 33. Exim was neglecting to zero errno before one call of strtol() when
6699 expanding a string and expecting an integer value. On some systems this
6700 resulted in spurious "integer overflow" errors. Also, it was casting the
6701 result into an int without checking.
6703 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6704 did not work. The code looks as if it has *never* worked, though it appears
6705 to have been documented since at least release 1.62. I have made it work.
6707 35. The "timeout_DNS" error in retry rules, also documented since at least
6708 1.62, also never worked. As it isn't clear exactly what this means, and
6709 clearly it isn't a major issue, I have abolished the feature by treating it
6710 as "timeout", and writing a warning to the main and panic logs.
6712 36. The display of retry rules for -brt wasn't always showing the error code
6715 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6716 timeout_connect_A, timeout_connect_MX.
6718 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6719 to the empty sender.
6721 39. The daemon was not analysing the content of -oX till after it had closed
6722 stderr and disconnected from the controlling terminal. This meant that any
6723 syntax errors were only noted on the panic log, and the return code from
6724 the command was 0. By re-arranging the code a little, I've made the
6725 decoding happen first, so such errors now appear on stderr, and the return
6726 code is 1. However, the actual setting up of the sockets still happens in
6727 the disconnected process, so errors there are still only recorded on the
6730 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6731 connections (as happens on some IP stacks) was logged at start up time as
6732 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6733 it from "IPv6 and IPv4", which means that two separate sockets are being
6736 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6737 says whether AF_INET or AF_INET6 was passed as an argument.
6739 42. Exiwhat output was messed up when time zones were included in log
6742 43. Exiwhat now gives more information about the daemon's listening ports,
6743 and whether -tls-on-connect was used.
6745 44. The "port" option of the smtp transport is now expanded.
6747 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6748 silently ignored. Now an error message is written to the main and panic
6751 46. There's a new ACL modifier called "logwrite" which writes to a log file
6752 as soon as it is encountered.
6754 47. Added $local_user_uid and $local_user_gid at routing time.
6756 48. Exim crashed when trying to verify a sender address that was being
6759 49. Exim was recognizing only a space character after ".include". It now also
6760 recognizes a tab character.
6762 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6763 extracting the relevant information from the specification. The man page no
6764 longer contains scrambled data for the -d option, and I've added a section
6765 at the front about calling Exim under different names.
6767 51. Added "extra_headers" argument to the "mail" command in filter files.
6769 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6772 53. Installed eximstats 1.29.
6774 54. Added transport_filter_timeout as a generic transport option.
6776 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6777 Cc: header lines. This was required by RFC 822, but it not required by RFC
6780 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6781 incoming messages that did not have them. Now it does so only if the
6782 message originates locally, that is, if there is no associated remote host
6783 address. When Resent- header lines are present, this applies to the Resent-
6784 lines rather than the non-Resent- lines.
6786 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6787 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6789 58. Messages for configuration errors now include the name of the main
6790 configuration file - useful now that there may be more than one file in a
6791 list (.included file names were always shown).
6793 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6794 for those rare installations that do not start the daemon as root or run it
6795 setuid root. I've cut out the call to initgroups() if the daemon is not
6798 60. The Exim user and group can now be bound into the binary as text strings
6799 that are looked up at the start of Exim's processing.
6801 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6803 62. Added $mailstore_basename variable.
6805 63. Installed patch to sieve.c from Michael Haardt.
6807 64. When Exim failed to open the panic log after failing to open the main log,
6808 the original message it was trying to log was written to stderr and debug
6809 output, but if they were not available (the usual case in production), it
6810 was lost. Now it is written to syslog before the two lines that record the
6811 failures to open the logs.
6813 65. Users' Exim filters run in subprocesses under the user's uid. It is
6814 possible for a "deliver" command or an alias in a "personal" command to
6815 provoke an address rewrite. If logging of address rewriting is configured,
6816 this fails because the process is not running as root or exim. There may be
6817 a better way of dealing with this, but for the moment (because 4.30 needs
6818 to be released), I have disabled address rewrite logging when running a
6819 filter in a non-root, non-exim process.
6825 1. The buildconfig auxiliary program wasn't quoting the value set for
6826 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6827 not defined. This bug was masked in 4.22 by the effect that was fixed in
6830 2. Some messages that were rejected after a message id was allocated were
6831 shown as "incomplete" by exigrep. It no longer does this for messages that
6832 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6834 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6835 have allow_domain_literals set, the ID did not get logged in the <= line.
6836 Domain literals are now always recognized in Message-ID: header lines.
6838 4. The first argument for a ${extract expansion item is the key name or field
6839 number. Leading and trailing spaces in this item were not being ignored,
6840 causing some misleading effects.
6842 5. When deliver_drop_privilege was set, single queue runner processes started
6843 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6844 same command in the process it spins off) were not dropping privilege.
6846 6. When the daemon running as "exim" started a queue runner, it always
6847 re-executed Exim in the spun-off process. This is a waste of effort when
6848 deliver_drop_privilege is set. The new process now just calls the
6849 queue-runner function directly.
6855 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6858 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6859 ignored. Though the use of -oP was forcing the writing of a pid file, it
6860 was always written to the default place.
6862 3. If the message "no IP address found for host xxxx" is generated during
6863 incoming verification, it is now followed by identification of the incoming
6864 connection (so you can more easily find what provoked it).
6866 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6868 5. Added some features to "harden" Exim a bit more against certain attacks:
6870 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6871 be put in Local/Makefile. This is like the never_users runtime option,
6872 but it cannot be overridden. The default setting is "root".
6874 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6875 prefix string with which any file named in a -C command line option
6878 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6879 is retained for -C and -D only if the caller of Exim is root. Without
6880 it, the exim user may also use -C and -D and retain privilege.
6882 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6883 command line option is disabled.
6885 6. Macro names set by the -D option must start with an upper case letter, just
6886 like macro names defined in the configuration file.
6888 7. Added "dereference=" facility to LDAP.
6890 8. Two instances of the typo "uknown" in the source files are fixed.
6892 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6893 the Configure-Makefile script screwed up while processing it.
6895 10. Incorporated PCRE 4.4.
6897 11. The SMTP synchronization check was not operating right at the start of an
6898 SMTP session. For example, it could not catch a HELO sent before the client
6899 waited for the greeting. There is now a check for outstanding input at the
6900 point when the greeting is written. Because of the duplex, asynchronous
6901 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6902 way, but not yet received, when the check is performed.
6904 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6905 on TCP/IP sockets, because this apparently causes some broken clients to
6908 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6909 unchanged) from the Cygwin maintainer.
6911 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6912 was supported instead of testing for mbx. Effectively a typo.
6914 15. The spa authenticator server code was not checking that the input it
6915 received was valid base64.
6917 16. The debug output line for the "set" modifier in ACLs was not showing the
6918 name of the variable that was being set.
6920 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6922 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6923 lookup on its own. Something else was needed to trigger the lookup. For
6924 example, a match in host_lookup or the need for a host name in a host list.
6925 Now, if $sender_host_name is referenced and the host name has not yet been
6926 looked up, a lookup is performed. If the lookup fails, the variable remains
6927 empty, and $host_lookup_failed is set to "1".
6929 19. Added "eqi" as a case-independent comparison operator.
6931 20. The saslauthd authentication condition could segfault if neither service
6932 nor realm was specified.
6934 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6935 error message that was logged was misleading, and incoming SMTP
6936 connections were dropped. The message is now more accurate, and temporary
6937 errors are given to SMTP connections.
6939 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6940 (except RSET) with a 421 error, until QUIT is received. However, it was
6941 failing to send a response to QUIT.
6943 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6944 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6945 ACL is now run after the helo_try_verify_hosts code.
6947 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6948 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6949 case-independent, so other case variants are also recognized). Apparently
6950 some systems use these upper case variants.
6952 25. If more than two messages were waiting for the same host, and a transport
6953 filter was specified for the transport, Exim sent two messages over the
6954 same TCP/IP connection, and then failed with "socket operation on non-
6955 socket" when it tried to send the third.
6957 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6959 27. The extern definition of crypt16() in expand.c was not being excluded when
6960 the OS had its own crypt16() function.
6962 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6963 as a preferred synonym for return_size_limit, both as an option and as an
6966 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6968 30. Changed the default configuration ACL to relax the local part checking rule
6969 for addresses that are not in any local domains. For these addresses,
6970 slashes and pipe symbols are allowed within local parts, but the sequence
6971 /../ is explicitly forbidden.
6973 31. SPA server authentication was not clearing the challenge buffer before
6976 32. log_message in a "warn" ACL statement was writing to the reject log as
6977 well as to the main log, which contradicts the documentation and doesn't
6978 seem right (because no rejection is happening). So I have stopped it.
6980 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6981 However, I am unable to do any testing of this.
6983 34. Fixed an infelicity in the appendfile transport. When checking directories
6984 for a mailbox, to see if any needed to be created, it was accidentally
6985 using path names with one or more superfluous leading slashes; tracing
6986 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6988 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6989 contents are added to the log line that is written for every discarded
6990 recipient. (Previously a log_message setting was ignored.)
6992 36. The ${quote: operator now quotes the string if it is empty.
6994 37. The install script runs exim in order to find its version number. If for
6995 some reason other than non-existence or emptiness, which it checks, it
6996 could not run './exim', it was installing it with an empty version number,
6997 i.e. as "exim-". This error state is now caught, and the installation is
7000 38. An argument was missing from the function that creates an error message
7001 when Exim fails to connect to the socket for saslauthd authentication.
7002 This could cause Exim to crash, or give a corrupted message.
7004 39. Added isip, isip4, and isip6 to ${if conditions.
7006 40. The ACL variables $acl_xx are now saved with the message, and can be
7007 accessed later in routers, transports, and filters.
7009 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7010 strings in the file are not string-expanded.
7012 42. If a MAIL command specified a SIZE value that was too large to fit into an
7013 int variable, the check against message_size_limit failed. Such values are
7014 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7015 day this will have to be increased, but I don't think I want to be around
7016 when emails are that large.
7023 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7024 iconv() is not standard in FreeBSD.
7026 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7027 IPv6 enabled. The observed symptom was a segmentation fault on return from
7028 the function os_common_find_running_interfaces() in src/os.c.
7030 3. In the check_special_case() function in daemon.c I had used "errno" as an
7031 argument name, which causes warnings on some systems. This was basically a
7032 typo, since it was named "eno" in the comments!
7034 4. The code that waits for the clock to tick (at a resolution of some fraction
7035 of a second) so as to ensure message-id uniqueness was always waiting for
7036 at least one whole tick, when it could have waited for less. [This is
7037 almost certainly not relevant at current processor speeds, where it is
7038 unlikely to ever wait at all. But we try to future-proof.]
7040 5. The function that sleeps for a time interval that includes fractions of a
7041 second contained a race. It did not block SIGALRM between setting the
7042 timer, and suspending (a couple of lines later). If the interval was short
7043 and the sigsuspend() was delayed until after it had expired, the suspension
7044 never ended. On busy systems this could lead to processes getting stuck for
7047 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7048 process, before it forks any delivery processes. The open lookup caching
7049 mechanism meant that the open file or database connection was passed into
7050 the delivery process. The problem was that delivery processes always tidy
7051 up cached lookup data. This could cause a problem for the next delivery
7052 process started by the queue runner, because the external queue runner
7053 process does not know about the closure. So the next delivery process
7054 still has data in the lookup cache. In the case of a file lookup, there was
7055 no problem because closing a file descriptor in a subprocess doesn't affect
7056 the parent. However, if the lookup was caching a connection to a database,
7057 the connection was closed, and the second delivery process was likely to
7058 see errors such as "PGSQL: query failed: server closed the connection
7059 unexpectedly". The problem has been fixed by closing all cached lookups
7060 in a queue runner before running a delivery process.
7062 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7063 seem to have the "const" qualifier which it has on other OS. I've
7066 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7067 *of the same type* (client or server) with the same public name that an
7068 error should be diagnosed.
7070 9. When Exim looked up a host name for an IP address, but failed to find the
7071 original IP address when looking up the host name (a safety check), it
7072 output the message "<ip address> does not match any IP for NULL", which was
7073 confusing, to say the least. The bug was that the host name should have
7074 appeared instead of "NULL".
7076 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7077 user that is built into the binary, and the -C or -D options is used, root
7078 privilege is dropped before the configuration file is read. In addition,
7079 logging is switched to stderr instead of the normal log files. If the
7080 configuration then re-defines the Exim user, the unprivileged environment
7081 is probably not what is expected, so Exim logs a panic warning message (but
7084 However, if deliver_drop_privilege is set, the unprivileged state may well
7085 be exactly what is intended, so the warning has been cut out in that case,
7086 and Exim is allowed to try to write to its normal log files.
7092 1. smtp_return_error_details was not giving details for temporary sender
7093 or receiver verification errors.
7095 2. Diagnose a configuration error if two authenticators have the same public
7098 3. Exim used not to create the message log file for a message until the first
7099 delivery attempt. This could be confusing when incoming messages were held
7100 for policy or load reasons. The message log file is now created at the time
7101 the message is received, and an initial "Received" line is written to it.
7103 4. The automatically generated man page for command line options had a minor
7104 bug that caused no ill effects; however, a more serious problem was that
7105 the procedure for building the man page automatically didn't always
7106 operate. Consequently, release 4.20 contains an out-of-date version. This
7107 shouldn't happen again.
7109 5. When building Exim with embedded Perl support, the script that builds the
7110 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7111 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7113 6. The freeze_tell option was not being used for messages that were frozen on
7114 arrival, either by an ACL or by local_scan().
7116 7. Added the smtp_incomplete_transaction log selector.
7118 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7119 was accepting AUTH without a new EHLO.
7121 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7122 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7125 10. Logging of TCP/IP connections (when configured) now happens in the main
7126 daemon process instead of the child process, so that the TCP/IP connection
7127 count is more accurate (but it can never be perfect).
7129 11. The use of "drop" in a nested ACL was not being handled correctly in the
7130 outer ACL. Now, if condition failure induced by the nested "drop" causes
7131 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7132 or "require"), the connection is dropped.
7134 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7135 that yield "discard" can now be used with an "accept" or a "discard" verb,
7136 but an error is generated for any others (because I can't see a useful way
7137 to define what should happen).
7139 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7140 are now processed in the same way as lines in the Exim configuration file.
7141 In particular, continuation lines are supported.
7143 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7145 15. Added -ti meaning -t -i.
7147 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7148 domains, and warn by logging if others are found.
7150 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7151 returned from GIFCONF when Exim is trying to find the list of interfaces on
7152 a host. The code in os.c has been modified to copy each ifreq to an aligned
7153 structure in all cases.
7155 Also, in some cases, the returned ifreq's were being copied to a 'struct
7156 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7157 means the last couple of bytes of an IPv6 address could be chopped if the
7158 ifreq contained only a normal sockaddr (14 bytes storage).
7160 18. Named domain lists were not supported in the hosts_treat_as_local option.
7161 An entry such as +xxxx was not recognized, and was treated as a literal
7164 19. Ensure that header lines added by a DATA ACL are included in the reject log
7165 if the ACL subsequently rejects the message.
7167 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7168 MD5 (which is deprecated).
7170 21. When testing a filter file using -bf, Exim was writing a message when it
7171 took the sender from a "From " line in the message, but it was not doing so
7172 when it took $return_path from a Return-Path: header line. It now does.
7174 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7175 with a valid header line field name (a series of printing characters
7176 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7178 23. Changed "disc" in the source to "disk" to conform to the documentation and
7179 the book and for uniformity.
7181 24. Ignore Sendmail's -Ooption=value command line item.
7183 25. When execve() failed while trying to run a command in a pipe transport,
7184 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7185 could be confused with a return value of 69 from the command itself. This
7186 has been changed to 127, the value the shell returns if it is asked to run
7187 a non-existent command. The wording for the related log line suggests a
7188 non-existent command as the problem.
7190 26. If received_header_text expands to an empty string, do not add a Received:
7191 header line to the message. (Well, it adds a token one on the spool, but
7192 marks it "old" so that it doesn't get used or transmitted.)
7194 27. Installed eximstats 1.28 (addition of -nt option).
7196 28. There was no check for failure on the call to getsockname() in the daemon
7197 code. This can fail if there is a shortage of resources on the system, with
7198 ENOMEM, for example. A temporary error is now given on failure.
7200 29. Contrary to the C standard, it seems that in some environments, the
7201 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7202 program. Exim now does this explicitly; it affects the formatting of
7203 timestamps using strftime().
7205 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7206 complaints. I've now initialized all the variables, to avoid this.
7208 32. Header lines added by a system filter were not being "seen" during
7209 transport-time rewrites.
7211 33. The info_callback() function passed to OpenSSL is set up with type void
7212 (*)(SSL *, int, int), as described somewhere. However, when calling the
7213 function (actually a macro) that sets it up, the type void(*)() is
7214 expected. I've put in a cast to prevent warnings from picky compilers.
7216 34. If a DNS black list lookup found a CNAME record, but there were no A
7217 records associated with the domain it pointed at, Exim crashed.
7219 35. If a DNS black list lookup returned more than one A record, Exim ignored
7220 all but the first. It now scans all returned addresses if a particular IP
7221 value is being sought. In this situation, the contents of the
7222 $dnslist_value variable are a list of all the addresses, separated by a
7225 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7226 to accept a host name and all its aliases if the forward lookup for any of
7227 them yielded the IP address of the incoming connection. Now it accepts only
7228 those names whose forward lookup yields the correct IP address. Any other
7229 names are discarded. This closes a loophole whereby a rogue DNS
7230 administrator could create reverse DNS records to break through a
7231 wildcarded host restriction in an ACL.
7233 37. If a user filter or a system filter that ran in a subprocess used any of
7234 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7235 the wrong values were passed to the pipe command ($thisaddress had the
7236 value of $0, $0 had the value of $1, etc). This bug was introduced by
7237 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7239 38. Improved the line breaking for long SMTP error messages from ACLs.
7240 Previously, if there was no break point between 40 and 75 characters, Exim
7241 left the rest of the message alone. Two changes have been made: (a) I've
7242 reduced the minimum length to 35 characters; (b) if it can't find a break
7243 point between 35 and 75 characters, it looks ahead and uses the first one
7244 that it finds. This may give the occasional overlong line, but at least the
7245 remaining text gets split now.
7247 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7248 file descriptors might be low, and that setting 1000 would always raise it.
7249 It turns out that in some environments, the limit is already over 1000 and
7250 that lowering it causes trouble. So now Exim takes care not to decrease it.
7252 40. When delivering a message, the value of $return_path is set to $sender_
7253 address at the start of routing (routers may change the value). By an
7254 oversight, this default was not being set up when an address was tested by
7255 -bt or -bv, which affected the outcome if any router or filter referred to
7258 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7259 log only when "message" or "log_message" are set. However, if one of the
7260 conditions was an address verification, or a call to a nested ACL, the
7261 messages generated by the underlying test were being passed through. This
7262 no longer happens. The underlying message is available in $acl_verify_
7263 message for both "message" and "log_message" expansions, so it can be
7264 passed through if needed.
7266 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7267 new expansion $bh_ to give the encoded byte string without charset
7268 translation. Translation happens only if iconv() is available; HAVE_ICONV
7269 indicates this at build time. HEADERS_CHARSET gives the charset to
7270 translate to; headers_charset can change it in the configuration, and
7271 "headers charset" can change it in an individual filter file.
7273 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7274 that creates RFC 2047 encoded "words" labels them as that charset instead
7275 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7276 expansion operator; (ii) when Exim creates a From: line for a local
7277 message; (iii) when a header line is rewritten to include a "phrase" part.
7279 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7280 buggy, causing it to skip the first lines of messages whose message ID
7281 ended in 'D'. This would not have bitten before Exim release 4.14, because
7282 message IDs were unlikely to end in 'D' before then. The effect was to have
7283 incorrect size information for certain domains.
7285 45. #include "config.h" was missing at the start of the crypt16.c module. This
7286 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7289 46. If there was a timeout during a "random" callout check, Exim treated it as
7290 a failure of the random address, and carried on sending RSET and the real
7291 address. If the delay was just some slowness somewhere, the response to the
7292 original RCPT would be taken as a response to RSET and so on, causing
7293 mayhem of various kinds.
7295 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7296 when I implemented it. It didn't allow for the fact that some option values
7297 may legitimately be negative (e.g. size_addition), and it didn't even do
7298 the right test for positive values.
7300 48. Domain names in DNS records are case-independent. Exim always looks them up
7301 in lower case. Some resolvers return domain names in exactly the case they
7302 appear in the zone file, that is, they may contain uppercase letters. Not
7303 all resolvers do this - some return always lower case. Exim was treating a
7304 change of case by a resolver as a change of domain, similar to a widening
7305 of a domain abbreviation. This triggered its re-routing code and so it was
7306 trying to route what was effectively the same domain again. This normally
7307 caused routing to fail (because the router wouldn't handle the domain
7308 twice). Now Exim checks for this case specially, and just changes the
7309 casing of the domain that it ultimately uses when it transmits the message
7312 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7315 50. If a filter generated a file delivery with a non-absolute name (possible if
7316 no home directory exists for the router), the forbid_file option was not
7319 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7320 the existing equality matching.
7322 52. Exim was using ints instead of ino_t variables in some places where it was
7323 dealing with inode numbers.
7325 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7326 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7327 called TMPDIR, and if it finds it is different, it changes its value.
7329 54. The smtp_printf() function is now made available to local_scan() so
7330 additional output lines can be written before returning. There is also an
7331 smtp_fflush() function to enable the detection of a dropped connection.
7332 The variables smtp_input and smtp_batched_input are exported to
7335 55. Changed the default runtime configuration: the message "Unknown user"
7336 has been removed from the ACL, and instead placed on the localuser router,
7337 using the cannot_route_message feature. This means that any verification
7338 failures that generate their own messages won't get overridden. Similarly,
7339 the "Unrouteable address" message that was in the ACL for unverifiable
7340 relay addresses has also been removed.
7342 56. Added hosts_avoid_esmtp to the smtp transport.
7344 57. The exicyclog script was not checking for the esoteric option
7345 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7346 will work only if exicyclog is run under the appropriate euid.
7348 58. Following a discussion on the list, the rules by which Exim recognises line
7349 endings on incoming messages have been changed. The -dropcr and drop_cr
7350 options are now no-ops, retained only for backwards compatibility. The
7351 following line terminators are recognized: LF CRLF CR. However, special
7352 processing applies to CR:
7354 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7355 nor a local message in the state where . is a terminator.
7357 (ii) If a bare CR is encountered in a header line, an extra space is added
7358 after the line terminator so as not to end the header. The reasoning
7359 behind this is that bare CRs in header lines are most likely either
7360 to be mistakes, or people trying to play silly games.
7362 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7363 was being incorrectly given as 18 bytes larger than it should have been.
7364 This is a VOB (very old bug).
7366 60. This may never have affected anything current, but just in case it has:
7367 When the local host is found other than at the start of a list of hosts,
7368 the local host, those with the same MX, and any that follow, are discarded.
7369 When the list in question was part of a longer list of hosts, the following
7370 hosts (not currently being processed) were also being discarded. This no
7371 longer happens. I'm not sure if this situation could ever has previously
7374 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7377 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7378 header. This is something that is recommended in a new Internet Draft, and
7379 is something that is documented as being done by Sendmail. There are two
7380 possible values. For messages generated by the autoreply transport, Exim
7383 Auto-Submitted: auto-replied
7385 whereas for all other generated messages (e.g. bounces) it adds
7387 Auto-Submitted: auto-generated
7389 63. The "personal" condition in filters now includes a test for the
7390 Auto-Submitted: header. If it contains the string "auto-" the message it
7391 not considered personal.
7393 64. Added rcpt_include_affixes as a generic transport option.
7395 65. Added queue_only_override (default true).
7397 66. Added the syslog_duplication option.
7399 67. If what should have been the first header line of a message consisted of
7400 a space followed by a colon, Exim was mis-interpreting it as a header line.
7401 It isn't of course - it is syntactically invalid and should therefore be
7402 treated as the start of the message body. The misbehaviour could have
7403 caused a number of strange effects, including loss of data in subsequent
7404 header lines, and spool format errors.
7406 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7407 client host had authenticated. This control can now be exercised by an ACL
7408 for more flexibility.
7410 69. By default, callouts do not happen when testing with -bh. There is now a
7411 variant, -bhc, which does actually run the callout code, including
7412 consulting and updating the callout cache.
7414 70. Added support for saslauthd authentication, courtesy of Alexander
7417 71. If statvfs() failed on the spool or log directories while checking their
7418 size for availability, Exim confusingly gave the error "space shortage".
7419 Furthermore, in debugging mode it crashed with a floating point exception.
7420 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7421 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7422 serious problem, Exim now writes to the main and panic logs when this
7423 happens, with details of the failure. It then refuses to accept the
7424 incoming message, giving the message "spool directory problem" or "log
7425 directory problem" with a 421 code for SMTP messages.
7427 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7428 0, 1, and 2 exist, because some OS complain for execs without them (see
7429 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7430 descriptors. However, the code omitted to check that the open succeeded,
7431 causing mysterious errors if for some reason the permissions on /dev/null
7432 got screwed. Now Exim writes a message to the main and panic logs, and
7433 bombs out if it can't open /dev/null.
7435 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7436 interact so that it is all more flexible. It is supposed to remain
7437 backwards compatible. Also added extra_local_interfaces.
7439 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7440 to bomb out with an assertion failure - to the client this appears as a
7441 connection drop. This problem occurs in the part of the code that was taken
7442 from the Samba project. Fortunately, the assertion is in a very simple
7443 function, so I have fixed this by reproducing the function inline in the
7444 one place where it is called, and arranging for authentication to fail
7445 instead of killing the process with assert().
7447 75. The SPA client code was not working when the server requested OEM rather
7448 than Unicode encoding.
7450 76. Added code to make require_files with a specific uid setting more usable in
7451 the case where statting the file as root fails - usually a non-root-mounted
7452 NFS file system. When this happens and the failure is EACCES, Exim now
7453 forks a subprocess and does the per-uid checking as the relevant uid.
7455 77. Added process_log_path.
7457 78. If log_file_path was not explicitly set, a setting of check_log_space or
7458 check_log_inodes was ignored.
7460 79. If a space check for the spool or log partitions fails, the incident is now
7461 logged. Of course, in the latter case the data may get lost...
7463 80. Added the %p formatting code to string_format() so that it can be used to
7464 print addresses in debug_print(). Adjusted all the address printing in the
7465 debugging in store.c to use %p rather than %d.
7467 81. There was a concern that a line of code in smtp_in.c could overflow a
7468 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7469 initially expressed, the concern was not well-founded, because trailing
7470 spaces are removed early. However, if the trailing spaces were followed by
7471 a NULL, they did not get removed, so the overflow was possible. Two fixes
7474 (a) I re-wrote the offending code in a cleaner fashion.
7475 (b) If an incoming SMTP command contains a NULL character, it is rejected
7478 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7479 runs initgroups(), so that if the Exim user is in any additional groups,
7480 they will be used during message reception.
7486 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob