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-
127 JH/01 Remove code calling the customisable local_scan function, unless a new
128 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
130 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
131 non-signal-safe functions being used.
133 JH/03 Bug 2269: When presented with a received message having a stupidly large
134 number of DKIM-Signature headers, disable DKIM verification to avoid
135 a resource-consumption attack. The limit is set at twenty.
137 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
138 report of oldest_pass in ${authres } in consequence, and separate out
139 some descriptions of reasons for verification fail.
141 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
142 files in the spool were present and unlocked. A queue-runner could spot
143 them, resulting in a duplicate delivery. Fix that by doing the unlock
144 after the unlink. Investigation by Tim Stewart. Take the opportunity to
145 add more error-checking on spoolfile handling while that code is being
148 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
149 No known attacks, no CVE, this is defensive hardening.
151 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
152 a queue-runner could start a delivery while other operations were ongoing.
153 Cutthrough delivery was a common victim, resulting in duplicate delivery.
154 Found and investigated by Tim Stewart. Fix by using the open message data
155 file handle rather than opening another, and not locally closing it (which
156 releases a lock) for that case, while creating the temporary .eml format
157 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
159 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
160 $sender_verify_failure/$recipient_verify_failure to "random".
162 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
165 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
166 Previously this would segfault.
168 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
171 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
172 like zero, since the resolver should be doing this for us, But we need one
173 as a CNAME but no MX presence gets the CNAME returned; we need to check
174 that doesn't point to an MX to declare it "no MX returned" rather than
175 "error, loop". A new main option is added so the older capability of
176 following some limited number of chain links is maintained.
178 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
180 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
181 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
182 now-deprecated earlier definitions used only specified the range up to TLS
183 1.2 (in the older-version library docs).
185 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
187 JH/15 Rework TLS client-side context management. Stop using a global, and
188 explicitly pass a context around. This enables future use of TLS for
189 connections to service-daemons (eg. malware scanning) while a client smtp
190 connection is using TLS; with cutthrough connections this is quite likely.
192 JH/16 Fix ARC verification to do AS checks in reverse order.
194 JH/17 Support a "tls" option on the ${readsocket } expansion item.
196 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
197 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
198 Previously the "utf8" would be re-prepended for every additional message.
200 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
201 Previously thery were accepted, resulting in issues when attempting to
202 forward messages to a non-supporting MTA.
204 PP/02 Let -n work with printing macros too, not just options.
206 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
207 one parent address was copied, and bogus data was used at delivery-logging
208 time. Either a crash (after delivery) or bogus log data could result.
209 Discovery and analysis by Tim Stewart.
211 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
212 Previously if the string ended mid-character, we did not insert the
213 promised '?' replacement.
215 PP/04 Documentation: current string operators work on bytes, not codepoints.
217 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
218 should pack well giving a smaller memory footprint so better caching and
219 therefore performance. Group the declarations where this can't be done so
220 that the byte-sized flag variables are not interspersed among pointer
221 variables, giving a better chance of good packing by the compiler.
223 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
224 non-null, to avoid issues with sites running BATV. Previously reports were
225 sent with an empty envelope sender so looked like bounces.
227 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
228 The ignore_error flag wasn't being returned from the filter subprocess so
229 was not set for later routers. Investigation and fix by Matthias Kurz.
231 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
232 and a msg:complete for the whole, when a message is manually removed using
233 -Mrm. Developement by Matthias Kurz, hacked on by JH.
235 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
236 a "Gnu special" function, asprintf() in the DB utility binary builds; I
237 hope that is portable enough.
239 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
240 requiring a known-CA anchor certificate; make it now rely entirely on the
241 TLSA as an anchor. Checking the name on the leaf cert against the name
242 on the A-record for the host is still done for TA (but not for EE mode).
244 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
245 included in delivery lines for non-proxied connections, when compiled with
246 SUPPORT_SOCKS and running with proxy logging enabled.
248 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
249 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
250 move the existing event to fire before the normal logging of message
251 failure so that custom logging is bracketed by normal logging.
253 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
254 msg:fail:internal event. Developement by Matthias Kurz.
256 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
257 far too small for todays use of crypto signatures stored there. Go all
258 the way to the max DNS message size of 64kB, even though this might be
259 overmuch for IOT constrained device use.
261 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
262 copy a string over itself. The library routine is documented as not
263 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
265 JH/32 For main options check_spool_space and check_inode_space, where the
266 platform supports 64b integers, support more than the previous 2^31 kB
267 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
268 the previous G, M, k.
270 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
271 $authenticated_fail_id variable on authentication failure. Previously
274 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
275 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
276 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
277 GNUTLS_SEC_PARAM_MEDIUM.
279 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
280 side. Previously we would continue as if no SNI had been received.
282 JH/36 Harden the handling of string-lists. When a list consisted of a sole
283 "<" character, which should be a list-separator specification, we walked
284 off past the nul-terimation.
286 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
287 causes) even when the retry time is not yet met. Previously they were
288 not, meaning that when (say) an account was over-quota and temp-rejecting,
289 and multiple senders' messages were queued, only one sender would get
290 notified on each configured delay_warning cycle.
292 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
294 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
295 cipher-suites, an error can be left on the stack even for a succeeding
296 accept; this results in impossible error messages when a later operation
299 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
300 return error codes indicating retry. Under TLS1.3 this becomes required.
302 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
303 it only wrote the new authenticators, resulting in a lack of tracking of
304 peer changes of ESMTP extensions until the next cache flush.
306 JH/41 Fix the loop reading a message header line to check for integer overflow,
307 and more-often against header_maxsize. Previously a crafted message could
308 induce a crash of the recive process; now the message is cleanly rejected.
310 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
311 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
317 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
318 When redis_servers is set to a list of > 1 element, and the Redis servers
319 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
320 case of MOVED into a DEFER case instead, thus moving the query onto the
321 next server in the list. For a cluster of N elements, all N servers must
322 be defined in redis_servers.
324 GF/02 Catch and remove uninitialized value warning in exiqsumm
325 Check for existence of @ARGV before looking at $ARGV[0]
327 JH/01 Replace the store_release() internal interface with store_newblock(),
328 which internalises the check required to safely use the old one, plus
329 the allocate and data copy operations duplicated in both (!) of the
330 extant use locations.
332 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
333 modifier. This matches the restriction on the commandline.
335 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
336 Previously only the last row was returned.
338 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
339 we assumed that tags in the header were well-formed, and parsed the
340 element content after inspecting only the first char of the tag.
341 Assumptions at that stage could crash the receive process on malformed
344 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
345 While running the DKIM ACL we operate on the Permanent memory pool so that
346 variables created with "set" persist to the DATA ACL. Also (at any time)
347 DNS lookups that fail create cache records using the Permanent pool. But
348 expansions release any allocations made on the current pool - so a dnsdb
349 lookup expansion done in the DKIM ACL releases the memory used for the
350 DNS negative-cache, and bad things result. Solution is to switch to the
351 Main pool for expansions.
352 While we're in that code, add checks on the DNS cache during store_reset,
353 active in the testsuite.
354 Problem spotted, and debugging aided, by Wolfgang Breyha.
356 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
357 When none of the hosts presented to a transport match an already-open
358 connection, close it and proceed with the list. Previously we would
359 queue the message. Spotted by Lena with Yahoo, probably involving
362 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
363 Previously a spurious "250 OK id=" response was appended to the proper
366 JH/08 The "support for" informational output now, which built with Content
367 Scanning support, has a line for the malware scanner interfaces compiled
368 in. Interface can be individually included or not at build time.
370 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
371 by the template makefile "src/EDITME". The "STREAM" support for an older
372 ClamAV interface method is removed.
374 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
375 rows affected is given instead).
377 JH/11 The runtime Berkeley DB library version is now additionally output by
378 "exim -d -bV". Previously only the compile-time version was shown.
380 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
381 SMTP connection. Previously, when one had more receipients than the
382 first, an abortive onward connection was made. Move to full support for
383 multiple onward connections in sequence, handling cutthrough connection
384 for all multi-message initiating connections.
386 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
387 routers. Previously, a multi-recipient message would fail to match the
388 onward-connection opened for the first recipient, and cause its closure.
390 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
391 a timeout on read on a GnuTLS initiating connection, resulting in the
392 initiating connection being dropped. This mattered most when the callout
393 was marked defer_ok. Fix to keep the two timeout-detection methods
396 JH/15 Relax results from ACL control request to enable cutthrough, in
397 unsupported situations, from error to silently (except under debug)
398 ignoring. This covers use with PRDR, frozen messages, queue-only and
401 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
403 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
404 metadata, resulting in a crash in free().
406 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
407 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
408 Broken also in d185889f4, with init system revamp.
410 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
411 Previously we abruptly closed the connection after reading a malware-
412 found indication; now we go on to read the "scan ok" response line,
415 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
416 ACL. Previously, a crash would result.
418 JH/19 Speed up macro lookups during configuration file read, by skipping non-
419 macro text after a replacement (previously it was only once per line) and
420 by skipping builtin macros when searching for an uppercase lead character.
422 JH/20 DANE support moved from Experimental to mainline. The Makefile control
423 for the build is renamed.
425 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
426 was allocated for every new TLS startup, meaning one per message. Fix
427 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
429 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
430 reported the original. Fix to report (as far as possible) the ACL
431 result replacing the original.
433 JH/23 Fix memory leak during multi-message connections using STARTTLS under
434 OpenSSL. Certificate information is loaded for every new TLS startup,
435 and the resources needed to be freed.
437 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
439 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
442 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
443 DATA response info to the (existing) per-recipient response info for
444 the "C=" log element. It can have useful tracking info from the
445 destination system. Patch from Simon Arlott.
447 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
448 length value. Previously this would segfault.
450 HS/02 Support Avast multiline protoocol, this allows passing flags to
451 newer versions of the scanner.
453 JH/28 Ensure that variables possibly set during message acceptance are marked
454 dead before release of memory in the daemon loop. This stops complaints
455 about them when the debug_store option is enabled. Discovered specifically
456 for sender_rate_period, but applies to a whole set of variables.
457 Do the same for the queue-runner and queue-list loops, for variables set
458 from spool message files. Do the same for the SMTP per-message loop, for
459 certain variables indirectly set in ACL operations.
461 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
462 as a multi-recipient message from a mailinglist manager). The coding had
463 an arbitrary cutoff number of characters while checking for more input;
464 enforced by writing a NUL into the buffer. This corrupted long / fast
465 input. The problem was exposed more widely when more pipelineing of SMTP
466 responses was introduced, and one Exim system was feeding another.
467 The symptom is log complaints of SMTP syntax error (NUL chars) on the
468 receiving system, and refused recipients seen by the sending system
469 (propating to people being dropped from mailing lists).
470 Discovered and pinpointed by David Carter.
472 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
473 replaced by the ${authresults } expansion.
475 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
477 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
478 allows proper process termination in container environments.
480 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
481 Previously the "final dot" had a newline after it; ensure it is CR,LF.
483 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
484 and "err_perm", deprecated since 4.83 when the RFC-defined words
485 "temperror" and "permerror" were introduced.
487 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
488 transport-filters or DKIM-signing. The restriction was lost in the
489 consolidation of verify-callout and delivery SMTP handling.
490 Extend the restriction to also cover ARC-signing.
492 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
493 in defer=pass mode supply a 450 to the initiator. Previously the message
496 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
497 tls_require_ciphers is used as before.
499 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
500 "pass_unscanned". Only tmpfails from the scanner are written to
501 the paniclog, as they may require admin intervention (permission
502 denied, license issues). Other scanner errors (like decompression
503 bombs) do not cause a paniclog entry.
505 JH/36 Fix reinitialisation of DKIM logging variable between messages.
506 Previously it was possible to log spurious information in receive log
509 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
510 triggered odd behaviour from Outlook Express clients.
512 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
515 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
516 since the IETF WG has not yet settled on that versus the original
517 "bare" representation.
519 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
520 Previously the millisecond value corrupted the output.
521 Fix also for syslog_pid=no and log_selector +pid, for which the pid
522 corrupted the output.
528 JH/01 Rework error string handling in TLS interface so that the caller in
529 more cases is responsible for logging. This permits library-sourced
530 string to be attached to addresses during delivery, and collapses
531 pairs of long lines into single ones.
533 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
534 during configuration. Wildcards are allowed and expanded.
536 JH/02 Rework error string handling in DKIM to pass more info back to callers.
537 This permits better logging.
539 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
540 do not close it down and have the child transport start it up again on
541 the passed-on TCP connection. Instead, proxy the child (and any
542 subsequent ones) for TLS via a unix-domain socket channel. Logging is
543 affected: the continued delivery log lines do not have any DNSSEC, TLS
544 Certificate or OCSP information. TLS cipher information is still logged.
546 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
547 identical IP addresses on different listening ports. Will also affect
550 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
551 add noisy ifdef guards to special-case this sillyness.
552 Patch from Bernd Kuhls.
554 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
555 than 255 are no longer allowed.
557 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
558 Disable the session-cache too, which might reduce our load. Since we
559 currrectly use a new context for every connection, both as server and
560 client, there is no benefit for these.
561 GnuTLS appears to not support tickets server-side by default (we don't
562 call gnutls_session_ticket_enable_server()) but client side is enabled
563 by default on recent versions (3.1.3 +) unless the PFS priority string
566 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
567 <https://reproducible-builds.org/specs/source-date-epoch/>.
569 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
570 the check for any unsuccessful recipients did not notice the limit, and
571 erroneously found still-pending ones.
573 JH/08 Pipeline CHUNKING command and data together, on kernels that support
574 MSG_MORE. Only in-clear (not on TLS connections).
576 JH/09 Avoid using a temporary file during transport using dkim. Unless a
577 transport-filter is involved we can buffer the headers in memory for
578 creating the signature, and read the spool data file once for the
579 signature and again for transmission.
581 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
582 4.77 as the kernel support then wasn't solid, having issues in 64bit
583 mode. Now, it's been long enough. Add support for FreeBSD also.
585 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
586 case where the routing stage had gathered several addresses to send to
587 a host before calling the transport for the first, we previously failed
588 to close down TLS in the old transport process before passing the TCP
589 connection to the new process. The new one sent a STARTTLS command
590 which naturally failed, giving a failed delivery and bloating the retry
591 database. Investigation and fix prototype from Wolfgang Breyha.
593 JH/12 Fix check on SMTP command input synchronisation. Previously there were
594 false-negatives in the check that the sender had not preempted a response
595 or prompt from Exim (running as a server), due to that code's lack of
596 awareness of the SMTP input buffering.
598 PP/04 Add commandline_checks_require_admin option.
599 Exim drops privileges sanely, various checks such as -be aren't a
600 security problem, as long as you trust local users with access to their
601 own account. When invoked by services which pass untrusted data to
602 Exim, this might be an issue. Set this option in main configuration
603 AND make fixes to the calling application, such as using `--` to stop
606 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
607 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
608 responses to those, into a single TLS record each way (this usually means
609 a single packet). As a side issue, smtp_enforce_sync now works on TLS
612 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
613 affects you only if you're dancing at the edge of the param size limits.
614 If you are, and this message makes sense to you, then: raise the
615 configured limit or use OpenSSL 1.1. Nothing we can do for older
618 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
619 cmdline element to get the documented default one. Previously it was
622 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
623 are unsuable for non-message-specific reasons - eg. connection timeout,
624 banner-time rejection.
626 JH/16 Fix logging of delivery remote port, when specified by router, under
629 PP/06 Repair manualroute's ability to take options in any order, even if one
630 is the name of a transport.
633 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
635 JH/17 Change the list-building routines interface to use the expanding-string
636 triplet model, for better allocation and copying behaviour.
638 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
639 Previously it was constructed the first time a possibly-matching string
640 was met in the configuration file input during startup; now it is done
643 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
644 compatible one, to avoid the (poorly documented) possibility of a config
645 file in the working directory redirecting the DB files, possibly correpting
646 some existing file. CVE-2017-10140 assigned for BDB.
648 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
649 cache-hot. Previously, although the result was properly cached, the
650 initial verify call returned a defer.
652 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
653 the main verify for receipient in uncached-mode.
655 JH/22 Retire historical build files to an "unsupported" subdir. These are
656 defined as "ones for which we have no current evidence of testing".
658 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
659 if present. Previously it was ignored.
661 JH/24 Start using specified-initialisers in C structure init coding. This is
662 a C99 feature (it's 2017, so now considered safe).
664 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
665 if was a fixed-sized field and bitmask ops via macros; it is now more
668 PP/07 GitHub PR 56: Apply MariaDB build fix.
669 Patch provided by Jaroslav Škarvada.
671 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
672 during Coverity cleanups [4.87 JH/47]
673 Diagnosis and fix provided by Michael Fischer v. Mollard.
675 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
676 the right size to place the terminating semicolon on its own folded
677 line, the header hash was calculated to an incorrect value thanks to
678 the (relaxed) space the fold became.
680 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
681 and confused the parent.
683 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
684 which could crash as a result. This could lead to undeliverable messages.
686 JH/28 Logging: "next input sent too soon" now shows where input was truncated
689 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
690 matters on fast-turnover and PID-randomising systems, which were getting
691 out-of-order delivery.
693 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
694 a possibly-overlapping copy. The symptom was that "Remote host closed
695 connection in response to HELO" was logged instead of the actual 4xx
698 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
699 Previously only that bufferd was discarded, resulting in SYMTP command
702 JH/32 DKIM: when a message has multiple signatures matching an identity given
703 in dkim_verify_signers, run the dkim acl once for each. Previously only
704 one run was done. Bug 2189.
706 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
707 "panic the current process" to "deliberately defer". The panic log is
708 still written with the problem list name; the mail and reject logs now
709 get a temp-reject line for the message that was being handled, saying
710 something like "domains check lookup or other defer". The SMTP 451
711 message is still "Temporary local problem".
713 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
714 A crafted sequence of BDAT commands could result in in-use memory beeing
715 freed. CVE-2017-16943.
717 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
718 from SMTP input. Previously it was always done; now only done for DATA
719 and not BDAT commands. CVE-2017-16944.
721 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
722 to the message (such as an overlong header line). Previously this was
723 not done and we did not exit BDAT mode. Followon from the previous item
724 though a different problem.
730 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
731 than -2003 did; needs libidn2 in addition to libidn.
733 JH/02 The path option on a pipe transport is now expanded before use.
735 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
736 Patch provided by "Björn", documentation fix added too.
738 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
739 missing a wire-to-host endian conversion.
741 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
742 close after a BDAT command line could be taken as a following command,
743 giving a synch failure. Fix by only checking for synch immediately
744 before acknowledging the chunk.
746 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
747 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
748 macro. Patches provided by Josh Soref.
750 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
751 Previously we did not; the RFC seems ambiguous and VRFY is not listed
752 by IANA as a service extension. However, John Klensin suggests that we
755 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
756 the dkim code may be unix-mode line endings rather than smtp wire-format
757 CRLF, so prepend a CR to any bare LF.
759 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
760 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
762 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
763 the first were themselves being wrongly included in the feed into dkim
764 processing; with most chunk sizes in use this resulted in an incorrect
765 body hash calculated value.
767 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
768 DKIM signature block, for verification. Although advised against by
769 standards it is specifically not ruled illegal.
771 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
773 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
774 missing a body hash (the bh= tag).
776 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
777 It seems that HAProxy sends the Proxy Protocol information in clear and
778 only then does a TLS startup, so do the same.
780 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
781 TCP connections (such as for Spamd) unless the daemon successfully set
782 Fast Open mode on its listening sockets. This fixes breakage seen on
783 too-old kernels or those not configured for Fast Open, at the cost of
784 requiring both directions being enabled for TFO, and TFO never being used
785 by non-daemon-related Exim processes.
787 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
788 endings, at least on the first header line. Try to canonify any that get
789 past that check, despite the cost.
791 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
792 now limited to an arbitrary five deep, while parsing addresses with the
793 strip_excess_angle_brackets option enabled.
795 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
796 instead leave the unprompted TLS handshake in socket buffer for the
797 TLS library to consume.
799 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
801 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
803 JH/16 Drop variables when they go out of scope. Memory management drops a whole
804 region in one operation, for speed, and this leaves assigned pointers
805 dangling. Add checks run only under the testsuite which checks all
806 variables at a store-reset and panics on a dangling pointer; add code
807 explicitly nulling out all the variables discovered. Fixes one known
808 bug: a transport crash, where a dangling pointer for $sending_ip_address
809 originally assigned in a verify callout, is re-used.
811 PP/06 Drop '.' from @INC in various Perl scripts.
813 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
815 PP/08 Reduce a number of compilation warnings under clang; building with
816 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
817 should be warning-free.
819 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
821 HS/01 Fix portability problems introduced by PP/08 for platforms where
822 realloc(NULL) is not equivalent to malloc() [SunOS et al].
824 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
825 chunk. This allows us to accept broken chunked messages. We need a more
826 general solution here.
828 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
829 already-broken messages in the queue.
831 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
833 JH/19 Fix reference counting bug in routing-generated-address tracking.
839 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
840 supports it and a size is available (ie. the sending peer gave us one).
842 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
843 years of being deprecated). The replacements are the ACLs
844 acl_smtp_mime and acl_not_smtp_mime.
846 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
847 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
848 or even an in-clear connection were permitted. Now, if the host lookup
849 was dnssec and dane was requested then the host is only used if the
850 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
851 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
852 if one fails this test.
853 This means that a poorly-configured remote DNS will make it incommunicado;
854 but it protects against a DNS-interception attack on it.
856 JH/04 Bug 1810: make continued-use of an open smtp transport connection
857 non-noisy when a race steals the message being considered.
859 JH/05 If main configuration option tls_certificate is unset, generate a
860 self-signed certificate for inbound TLS connections.
862 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
863 in rewrites and routers.
865 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
866 and logged a warning sing 4.83; now they are a configuration file error.
868 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
869 (lacking @domain). Apply the same qualification processing as RCPT.
871 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
873 JH/10 Support ${sha256:} applied to a string (as well as the previous
876 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
877 a cutthrough deliver is pending, as we always want to make a connection.
878 This also avoids re-routing the message when later placing the cutthrough
879 connection after a verify cache hit.
880 Do not update it with the verify result either.
882 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
883 when routing results in more than one destination address.
885 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
886 signing (which inhibits the cutthrough capability). Previously only
887 the presence of an option was tested; now an expansion evaluating as
888 empty is permissible (obviously it should depend only on data available
889 when the cutthrough connection is made).
891 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
892 the relevant preceding SMTP command did not note the pipelining mode.
894 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
895 Previously they were not counted.
897 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
898 as one having no matching records. Previously we deferred the message
899 that needed the lookup.
901 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
902 distinguished as "(=".
904 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
905 for missing MX records. Previously it only worked for missing A records.
907 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
909 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
910 after the data-go-ahead and data-ack. Patch from Jason Betts.
912 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
913 even for a "none" policy. Patch from Tony Meyer.
915 JH/22 Fix continued use of a connection for further deliveries. If a port was
916 specified by a router, it must also match for the delivery to be
919 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
920 When one of the recipients of a message was unsuitable for the connection
921 (has no matching addresses), we lost track of needing to mark it
922 deferred. As a result mail would be lost.
924 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
926 JH/25 Decoding ACL controls is now done using a binary search; the source code
927 takes up less space and should be simpler to maintain. Merge the ACL
928 condition decode tables also, with similar effect.
930 JH/26 Fix problem with one_time used on a redirect router which returned the
931 parent address unchanged. A retry would see the parent address marked as
932 delivered, so not attempt the (identical) child. As a result mail would
935 JH/27 Fix a possible security hole, wherein a process operating with the Exim
936 UID can gain a root shell. Credit to http://www.halfdog.net/ for
937 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
940 JH/28 Enable {spool,log} filesystem space and inode checks as default.
941 Main config options check_{log,spool}_{inodes,space} are now
942 100 inodes, 10MB unless set otherwise in the configuration.
944 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
945 Previously it only applied to the main-section connection policy
948 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
950 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
951 by me. Added RFC7919 DH primes as an alternative.
953 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
954 are not in the system include path.
956 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
957 GnuTLS, when a session startup failed (eg because the client disconnected)
958 Exim did stdio operations after fclose. This was exposed by a recent
959 change which nulled out the file handle after the fclose.
961 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
962 signed directly by the cert-signing cert, rather than an intermediate
963 OCSP-signing cert. This is the model used by LetsEncrypt.
965 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
967 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
968 an incoming connection.
970 HS/02 Bug 1802: Do not half-close the connection after sending a request
973 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
974 fallback to "prime256v1".
976 JH/34 SECURITY: Use proper copy of DATA command in error message.
977 Could leak key material. Remotely exploitable. CVE-2016-9963.
983 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
984 and 3.4.4 - once the server is enabled to respond to an OCSP request
985 it does even when not requested, resulting in a stapling non-aware
986 client dropping the TLS connection.
988 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
989 support variable-length bit vectors. No functional change.
991 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
992 The I= interface field on outgoing lines is now after the H= remote
993 host field, same as incoming lines. There is a separate
994 outgoing_interface log selector which allows you to disable the
997 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
998 If not running log_selector +smtp_connection the mainlog would be held
999 open indefinitely after a "too many connections" event, including to a
1000 deleted file after a log rotate. Leave the per net connection logging
1001 leaving it open for efficiency as that will be quickly detected by the
1002 check on the next write.
1004 HS/01 Bug 1671: Fix post transport crash.
1005 Processing the wait-<transport> messages could crash the delivery
1006 process if the message IDs didn't exist for some reason. When
1007 using 'split_spool_directory=yes' the construction of the spool
1008 file name failed already, exposing the same netto behaviour.
1010 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1011 mime_regex ACL conditions.
1013 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1014 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1015 to HELO, local diagnostic string.
1017 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1018 log line to debug. Even when configured with a tls authenticator many
1019 client connections are expected to not authenticate in this way, so
1020 an authenticate fail is not an error.
1022 HS/02 Add the Exim version string to the process info. This way exiwhat
1023 gives some more detail about the running daemon.
1025 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1026 matter for fast-change records such as DNSBLs.
1028 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1029 retry record, even if constant. There may be multiple transports with
1030 different interface settings and the retry behaviour needs to be kept
1033 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1035 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1037 JH/10 Bug 840: fix log_defer_output option of pipe transport
1039 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1040 hosts_randomize. This matters a lot when combined with mua_wrapper.
1042 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1043 ${quote_pgsql:<string>} operator.
1045 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1047 JH/14 Change header file name for freeradius-client. Relevant if compiling
1048 with Radius support; from the Gentoo tree and checked under Fedora.
1050 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1052 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1053 option transport_filter, meaning no filtering.
1055 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1057 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1058 defaults to "*" (all hosts). The variable is now available when not built
1059 with TLS, default unset, mainly to enable keeping the testsuite sane.
1060 If a server certificate is not supplied (via tls_certificate) an error is
1061 logged, and clients will find TLS connections fail on startup. Presumably
1062 they will retry in-clear.
1063 Packagers of Exim are strongly encouraged to create a server certificate
1064 at installation time.
1066 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1067 with the $config_file variable.
1069 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1070 in transport context, after the attempt, and per-recipient. The latter type
1071 is per host attempted. The event data is the error message, and the errno
1072 information encodes the lookup type (A vs. MX) used for the (first) host,
1073 and the trailing two digits of the smtp 4xx response.
1075 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1076 to write to mainlog (or rejectlog, paniclog) in the window between file
1077 creation and permissions/ownership being changed. Particularly affects
1078 installations where exicyclog is run as root, rather than exim user;
1079 result is that the running daemon panics and dies.
1081 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1083 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1084 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1085 "pri" and "weight". Note that the previous implicit priority given by the
1086 list order is no longer honoured.
1088 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1089 for DKIM processing.
1091 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1092 by defining SUPPORT_SOCKS.
1094 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1095 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1096 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1097 variables are renamed to proxy_{local,external}_{address,port}.
1099 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1100 for a build by defining SUPPORT_I18N
1102 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1103 of the query string, and make ${quote_redis:} do that quoting.
1105 JH/29 Move Events support from Experimental to mainline, enabled by default
1106 and removable for a build by defining DISABLE_EVENT.
1108 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1110 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1111 cached by the daemon.
1113 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1114 by defining LOOKUP_REDIS. The libhiredis library is required.
1116 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1117 keys are given for lookup.
1119 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1120 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1121 only supported when built with TLS support. The PolarSSL SHA routines
1122 are still used when the TLS library is too old for convenient support.
1124 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1125 openssl_options), for security. OpenSSL forces this from version 1.1.0
1126 server-side so match that on older versions.
1128 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1129 allocation for $value could be released as the expansion processing
1130 concluded, but leaving the global pointer active for it.
1132 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1133 and to use the domains and local_parts ACL conditions.
1135 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1136 incorrectly not doubled on cutthrough transmission, hence seen as a
1137 body-termination at the receiving system - resulting in truncated mails.
1138 Commonly the sender saw a TCP-level error, and retransmitted the message
1139 via the normal store-and-forward channel. This could result in duplicates
1140 received - but deduplicating mailstores were liable to retain only the
1141 initial truncated version.
1143 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1145 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1147 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1148 we're in there, support oversigning also; bug 1309.
1150 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1152 HS/04 Add support for keep_environment and add_environment options.
1154 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1155 either intentional arithmetic overflow during PRNG, or testing config-
1158 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1159 delivery resulted in actual delivery. Cancel cutthrough before DATA
1162 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1163 rejects a recipient: pass the reject to the originator.
1165 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1166 Many were false-positives and ignorable, but it's worth fixing the
1169 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1170 for the new environment-manipulation done at startup. Move the routines
1171 from being local to tls.c to being global via the os.c file.
1173 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1174 an extract embedded as result-arg for a map, the first arg for extract
1175 is unavailable so we cannot tell if this is a numbered or keyed
1176 extraction. Accept either.
1182 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1185 JH/02 The smtp transport option "multi_domain" is now expanded.
1187 JH/03 The smtp transport now requests PRDR by default, if the server offers
1190 JH/04 Certificate name checking on server certificates, when exim is a client,
1191 is now done by default. The transport option tls_verify_cert_hostnames
1192 can be used to disable this per-host. The build option
1193 EXPERIMENTAL_CERTNAMES is withdrawn.
1195 JH/05 The value of the tls_verify_certificates smtp transport and main options
1196 default to the word "system" to access the system default CA bundle.
1197 For GnuTLS, only version 3.0.20 or later.
1199 JH/06 Verification of the server certificate for a TLS connection is now tried
1200 (but not required) by default. The verification status is now logged by
1201 default, for both outbound TLS and client-certificate supplying inbound
1204 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1207 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1208 Status Notification (bounce) messages are now MIME format per RFC 3464.
1209 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1210 under the control of the dsn_advertise_hosts option, and routers may
1211 have a dsn_lasthop option.
1213 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1214 default, modifiable by a malware= option. The list separator for
1215 the options can now be changed in the usual way. Bug 68.
1217 JH/10 The smtp_receive_timeout main option is now expanded before use.
1219 JH/11 The incoming_interface log option now also enables logging of the
1220 local interface on delivery outgoing connections.
1222 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1223 if the interface and destination host and port all match.
1225 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1228 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1229 Patch from Andrew Lewis.
1231 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1232 now supports optional time-restrictions, weighting, and priority
1233 modifiers per server. Patch originally by <rommer@active.by>.
1235 JH/16 The spamd_address main option now supports a mixed list of local
1236 and remote servers. Remote servers can be IPv6 addresses, and
1237 specify a port-range.
1239 JH/17 Bug 68: The spamd_address main option now supports an optional
1240 timeout value per server.
1242 JH/18 Bug 1581: Router and transport options headers_add/remove can
1243 now have the list separator specified.
1245 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1248 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1251 JH/21 Support for the A6 type of dns record is withdrawn.
1253 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1254 rather than the verbs used.
1256 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1257 from 255 to 1024 chars.
1259 JH/24 Verification callouts now attempt to use TLS by default.
1261 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1262 are generic router options now. The defaults didn't change.
1264 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1265 Original patch from Alexander Shikoff, worked over by JH.
1267 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1268 files if ZCAT_COMMAND is not executable.
1270 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1272 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1274 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1275 Normally benign, it bites when the pair was led to by a CNAME;
1276 modern usage is to not canonicalize the domain to a CNAME target
1277 (and we were inconsistent anyway for A-only vs AAAA+A).
1279 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1281 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1282 when evaluating $sender_host_dnssec.
1284 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1285 $sender_helo_dnssec variable.
1287 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1289 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1291 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1293 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1294 documented as working, but never had. Support all but $spam_report.
1296 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1297 added for tls authenticator.
1299 HS/03 Add perl_taintmode main config option
1305 TL/01 When running the test suite, the README says that variables such as
1306 no_msglog_check are global and can be placed anywhere in a specific
1307 test's script, however it was observed that placement needed to be near
1308 the beginning for it to behave that way. Changed the runtest perl
1309 script to read through the entire script once to detect and set these
1310 variables, reset to the beginning of the script, and then run through
1311 the script parsing/test process like normal.
1313 TL/02 The BSD's have an arc4random API. One of the functions to induce
1314 adding randomness was arc4random_stir(), but it has been removed in
1315 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1316 function when detected.
1318 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1319 cause callback expansion.
1321 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1322 syntax errors in an expansion can be treated as a string instead of
1323 logging or causing an error, due to the internal use of bool_lax
1324 instead of bool when processing it.
1326 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1327 server certificates when making smtp deliveries.
1329 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1331 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1333 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1335 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1336 Merged patch from Sebastian Wiedenroth.
1338 JH/05 Fix results-pipe from transport process. Several recipients, combined
1339 with certificate use, exposed issues where response data items split
1340 over buffer boundaries were not parsed properly. This eventually
1341 resulted in duplicates being sent. This issue only became common enough
1342 to notice due to the introduction of connection certificate information,
1343 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1345 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1346 size buffer was used, resulting in syntax errors when an expansion
1349 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1350 version 3.3.6 or later.
1352 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1353 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1354 both become "event_action", the variables become $event_name, $event_data
1355 and $event_defer_errno. There is a new variable $verify_mode, usable in
1356 routers, transports and related events. The tls:cert event is now also
1357 raised for inbound connections, if the main configuration event_action
1360 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1361 early OCSP support, but no stapling (appears to be less than 1.0.0).
1363 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1364 server certificate names available under the smtp transport option
1365 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1368 JH/10 Time-related extraction expansions from certificates now use the main
1369 option "timezone" setting for output formatting, and are consistent
1370 between OpenSSL and GnuTLS compilations. Bug 1541.
1372 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1373 encoded parameter in the incoming message. Bug 1558.
1375 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1376 include certificate info, eximon was claiming there were spoolfile
1379 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1381 JH/14 Log delivery-related information more consistently, using the sequence
1382 "H=<name> [<ip>]" wherever possible.
1384 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1385 are problematic for Debian distribution, omit them from the release
1388 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1390 JH/16 Fix string representation of time values on 64bit time_t architectures.
1393 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1394 output list separator was used.
1399 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1400 checkers that were complaining about end of non-void function with no
1403 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1404 This was a regression introduced in 4.83 by another bugfix.
1406 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1408 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1409 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1415 TF/01 Correctly close the server side of TLS when forking for delivery.
1417 When a message was received over SMTP with TLS, Exim failed to clear up
1418 the incoming connection properly after forking off the child process to
1419 deliver the message. In some situations the subsequent outgoing
1420 delivery connection happened to have the same fd number as the incoming
1421 connection previously had. Exim would try to use TLS and fail, logging
1422 a "Bad file descriptor" error.
1424 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1425 utilities have not been installed.
1427 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1428 temporary space as the ACL may create new global variables.
1430 TL/01 LDAP support uses per connection or global context settings, depending
1431 upon the detected version of the libraries at build time.
1433 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1434 to extract and use the src ip:port in logging and expansions as if it
1435 were a direct connection from the outside internet. PPv2 support was
1436 updated based on HAProxy spec change in May 2014.
1438 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1440 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1441 Properly escape header and check for NULL return.
1443 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1446 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1448 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1449 characters in header names, implemented as a verify condition.
1450 Contributed by Michael Fischer v. Mollard.
1452 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1453 results permerror and temperror. Previous values are deprecated but
1454 still accepted. In a future release, err_perm and err_temp will be
1455 completely removed, which will be a backward incompatibility if the
1456 ACL tests for either of these two old results. Patch contributed by
1457 user bes-internal on the mailing list.
1459 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1461 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1462 selectors, in both main and reject logs.
1464 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1467 JH/07 Add malware type "sock" for talking to simple daemon.
1469 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1471 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1472 routers/transports under cutthrough routing.
1474 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1475 numbers. Touch up "bool" conditional to keep the same definition.
1477 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1479 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1481 JH/12 Expand items in router/transport headers_add or headers_remove lists
1482 individually rather than the list as a whole. Bug 1452.
1484 Required for reasonable handling of multiple headers_ options when
1485 they may be empty; requires that headers_remove items with embedded
1486 colons must have them doubled (or the list-separator changed).
1488 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1489 view the policy declared in the DMARC record. Currently, $dmarc_status
1490 is a combined value of both the record presence and the result of the
1493 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1495 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1496 dnslookup router and the smtp transport (applying to the forward
1499 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1500 of ldap servers used for a specific lookup. Patch provided by Heiko
1503 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1504 New variable $lookup_dnssec_authenticated for observability.
1506 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1507 Patch submitted by Lars Timman.
1509 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1511 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1512 Requires trusted mode and valid format message id, aborts otherwise.
1513 Patch contributed by Heiko Schlichting.
1515 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1516 certextract with support for various fields. Bug 1358.
1518 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1519 is requested by default, modifiable by smtp transport option
1522 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1523 operate on certificate variables to give certificate fingerprints
1524 Also new ${sha256:cert_variable}.
1526 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1528 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1529 Christian Aistleitner.
1531 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1533 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1534 file. Patch from Wolfgang Breyha.
1536 JH/25 Expand the coverage of the delivery $host and $host_address to
1537 client authenticators run in verify callout. Bug 1476.
1539 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1540 align with daemon_smtp_ports. Bug 72.
1542 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1543 support and error reporting did not work properly.
1545 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1546 and is readable. Patch from Andrew Colin Kissa.
1548 TL/14 Enhance documentation of ${run expansion and how it parses the
1549 commandline after expansion, particularly in the case when an
1550 unquoted variable expansion results in an empty value.
1552 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1554 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1555 by renaming away from it. Addresses GH issue 15, reported by
1558 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1560 TL/15 SECURITY: prevent double expansion in math comparison functions
1561 (can expand unsanitized data). Not remotely exploitable.
1568 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1570 PP/02 Make -n do something, by making it not do something.
1571 When combined with -bP, the name of an option is not output.
1573 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1576 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1577 $sender_host_name and config options to manage this, and basic check
1580 PP/05 DSCP support for outbound connections and control modifier for inbound.
1582 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1583 (Only plugin which currently uses this is kerberos4, which nobody should
1584 be using, but we should make it available and other future plugins might
1585 conceivably use it, even though it would break NAT; stuff *should* be
1586 using channel bindings instead).
1588 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1589 name; added for Sendmail compatibility; requires admin caller.
1590 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1591 just ignore it); requires trusted caller.
1592 Also parse but ignore: -Ac -Am -X<logfile>
1595 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1597 TL/02 Add +smtp_confirmation as a default logging option.
1599 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1600 Patch by Magnus Holmgren from 2007-02-20.
1602 TL/04 Bugzilla 1281 - Spec typo.
1603 Bugzilla 1283 - Spec typo.
1604 Bugzilla 1290 - Spec grammar fixes.
1606 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1608 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1610 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1611 mailing list by by Dmitry Isaikin.
1613 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1615 JH/02 Support "G" suffix to numbers in ${if comparisons.
1617 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1619 NM/01 Bugzilla 1197 - Spec typo
1620 Bugzilla 1196 - Spec examples corrections
1622 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1624 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1625 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1628 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1629 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1631 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1632 "acl {{name}{arg}...}", and optional args on acl condition
1635 JH/05 Permit multiple router/transport headers_add/remove lines.
1637 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1639 JH/07 Avoid using a waiting database for a single-message-only transport.
1640 Performance patch from Paul Fisher. Bugzilla 1262.
1642 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1645 JH/09 Add $headers_added variable, with content from use of ACL modifier
1646 add_header (but not yet added to the message). Bugzilla 199.
1648 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1649 Pulled from Bugzilla 817 by Wolfgang Breyha.
1651 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1653 (nb: this is the same fix as in Exim 4.80.1)
1655 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1658 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1661 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1663 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1664 advertise SMTP AUTH mechanism to us, instead of a generic
1665 protocol violation error. Also, make Exim more robust to bad
1666 data from the Dovecot auth socket.
1668 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1670 When a queue runner is handling a message, Exim first routes the
1671 recipient addresses, during which it prunes them based on the retry
1672 hints database. After that it attempts to deliver the message to
1673 any remaining recipients. It then updates the hints database using
1676 So if a recipient address works intermittently, it can get repeatedly
1677 deferred at routing time. The retry hints record remains fresh so the
1678 address never reaches the final cutoff time.
1680 This is a fairly common occurrence when a user is bumping up against
1681 their storage quota. Exim had some logic in its local delivery code
1682 to deal with this. However it did not apply to per-recipient defers
1683 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1685 This change adds a proper retry rule check during routing so that the
1686 final cutoff time is checked against the message's age. We only do
1687 this check if there is an address retry record and there is not a
1688 domain retry record; this implies that previous attempts to handle
1689 the address had the retry_use_local_parts option turned on. We use
1690 this as an approximation for the destination being like a local
1691 delivery, as in LMTP.
1693 I suspect this new check makes the old local delivery cutoff check
1694 redundant, but I have not verified this so I left the code in place.
1696 TF/02 Correct gecos expansion when From: is a prefix of the username.
1698 Test 0254 submits a message to Exim with the header
1702 When I ran the test suite under the user fanf2, Exim expanded
1703 the header to contain my full name, whereas it should have added
1704 a Resent-Sender: header. It erroneously treats any prefix of the
1705 username as equal to the username.
1707 This change corrects that bug.
1709 GF/01 DCC debug and logging tidyup
1710 Error conditions log to paniclog rather than rejectlog.
1711 Debug lines prefixed by "DCC: " to remove any ambiguity.
1713 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1715 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1716 Bug spotted by Jeremy Harris; was flawed since initial commit.
1717 Would have resulted in OCSP responses post-SNI triggering an Exim
1718 NULL dereference and crash.
1720 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1722 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1723 Bug detection, analysis and fix by Samuel Thibault.
1724 Bugzilla 1331, Debian bug #698092.
1726 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1728 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1729 Server implementation by Todd Lyons, client by JH.
1730 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1731 config variable "prdr_enable" controls whether the server
1732 advertises the facility. If the client requests PRDR a new
1733 acl_data_smtp_prdr ACL is called once for each recipient, after
1734 the body content is received and before the acl_smtp_data ACL.
1735 The client is controlled by both of: a hosts_try_prdr option
1736 on the smtp transport, and the server advertisement.
1737 Default client logging of deliveries and rejections involving
1738 PRDR are flagged with the string "PRDR".
1740 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1741 fclose(). Diagnosis by Todd Lyons.
1743 PP/17 Update configure.default to handle IPv6 localhost better.
1744 Patch by Alain Williams (plus minor tweaks).
1747 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1748 This is now consistent with GnuTLS, and is now documented: the
1749 previous undocumented portable approach to treating the option as
1750 unset was to force an expansion failure. That still works, and
1751 an empty string is now equivalent.
1753 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1754 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1755 not performing validation itself.
1757 PP/20 Added force_command boolean option to pipe transport.
1758 Patch from Nick Koston, of cPanel Inc.
1760 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1763 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1765 PP/21 Fix eximon continuous updating with timestamped log-files.
1766 Broken in a format-string cleanup in 4.80, missed when I repaired the
1767 other false fix of the same issue.
1768 Report and fix from Heiko Schlichting.
1771 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1772 Report from Prashanth Katuri.
1774 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1775 It's SecureTransport, so affects any MacOS clients which use the
1776 system-integrated TLS libraries, including email clients.
1778 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1779 using a MIME ACL for non-SMTP local injection.
1780 Report and assistance in diagnosis by Warren Baker.
1782 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1784 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1786 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1787 last id that failed so it may be referenced in subsequent ACL's.
1789 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1792 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1793 ldap library initialization, allowing self-signed CA's to be
1794 used. Also properly sets require_cert option later in code by
1795 using NULL (global ldap config) instead of ldap handle (per
1796 session). Bug diagnosis and testing by alxgomz.
1798 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1799 the src/util/ subdirectory.
1801 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1802 renamed to Transport Post Delivery Action by Jeremy Harris, as
1805 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1806 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1807 redis_servers = needs to be configured which will be used by the redis
1808 lookup. Patch from Warren Baker, of The Packet Hub.
1810 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1812 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1813 hostname or reverse DNS when processing a host list. Used suggestions
1814 from multiple comments on this bug.
1816 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1818 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1819 Missed a few lines, added it to make the runtest require no keyboard
1822 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1823 contains upper case chars. Make router use caseful_local_part.
1825 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1826 support when GnuTLS has been built with p11-kit.
1832 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1834 This, or similar/improved, will also be change PP/11 of 4.82.
1840 PP/01 Handle short writes when writing local log-files.
1841 In practice, only affects FreeBSD (8 onwards).
1842 Bugzilla 1053, with thanks to Dmitry Isaikin.
1844 NM/01 Bugzilla 949 - Documentation tweak
1846 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1849 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1851 PP/02 Implemented gsasl authenticator.
1853 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1855 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1856 `pkg-config foo` for cflags/libs.
1858 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1859 with rest of GSASL and with heimdal_gssapi.
1861 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1862 `pkg-config foo` for cflags/libs for the TLS implementation.
1864 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1865 properties get this fed in as external SSF. A number of robustness
1866 and debugging improvements to the cyrus_sasl authenticator.
1868 PP/08 cyrus_sasl server now expands the server_realm option.
1870 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1871 Patch by Jeremy Harris.
1873 PP/10 Added dbmjz lookup type.
1875 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1877 PP/12 MAIL args handles TAB as well as SP, for better interop with
1878 non-compliant senders.
1879 Analysis and variant patch by Todd Lyons.
1881 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1882 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1883 Patch from Dirk Mueller <dmueller@suse.com>
1885 PP/13 tls_peerdn now print-escaped for spool files.
1886 Observed some $tls_peerdn in wild which contained \n, which resulted
1887 in spool file corruption.
1889 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1890 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1891 or write after TLS renegotiation, which otherwise led to messages
1894 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1895 as a tracking header (ie: a signed header comes before the signature).
1896 Patch from Wolfgang Breyha.
1898 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1899 comma-sep list; embedded commas doubled.
1901 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1903 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1905 Report and patch from Dmitry Banschikov.
1907 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1908 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1909 failures. We appear to now support TLS1.1+ with Exim.
1911 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1912 lets Exim select keys and certificates based upon TLS SNI from client.
1913 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1914 before an outbound SMTP session. New log_selector, +tls_sni.
1916 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1917 NULL dereference. Report and patch from Alun Jones.
1919 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1920 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1921 Not seeing resolver debug output on NetBSD, but suspect this is a
1922 resolver implementation change.
1924 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1925 Left warnings. Added "eximon gdb" invocation mode.
1927 PP/21 Defaulting "accept_8bitmime" to true, not false.
1929 PP/22 Added -bw for inetd wait mode support.
1931 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1932 locate the relevant includes and libraries. Made this the default.
1934 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1935 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1937 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1938 This may cause build issues on older platforms.
1940 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1941 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1942 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1943 Added SNI support via GnuTLS too.
1944 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1946 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1948 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1949 Applied second patch from Janne, implementing suggestion to default
1950 multiple-strings-in-record handling to match SPF spec.
1952 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1954 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1955 read-only, out of scope).
1956 Patch from Wolfgang Breyha, report from Stuart Northfield.
1958 PP/29 Fix three issues highlighted by clang analyser static analysis.
1959 Only crash-plausible issue would require the Cambridge-specific
1960 iplookup router and a misconfiguration.
1961 Report from Marcin Mirosław.
1963 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1965 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1966 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1967 As part of this, removing so much warning spew let me fix some minor
1968 real issues in debug logging.
1970 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1971 assignment on my part. Fixed.
1973 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1974 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1975 Janne Snabb (who went above and beyond: thank you).
1977 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1978 string otherwise requires a connection and a bunch more work and it's
1979 relatively easy to get wrong. Should also expose TLS library linkage
1982 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1983 64-bit ${eval} (JH/03).
1985 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1986 GNU libc to support some of the 64-bit stuff, should not lead to
1987 conflicts. Defined before os.h is pulled in, so if a given platform
1988 needs to override this, it can.
1990 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1991 protection layer was required, which is not implemented.
1992 Bugzilla 1254, patch from Wolfgang Breyha.
1994 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1995 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1996 tls_dhparam take prime identifiers. Also unbreak combination of
1997 OpenSSL+DH_params+TLSSNI.
1999 PP/39 Disable SSLv2 by default in OpenSSL support.
2005 PP/01 Solaris build fix for Oracle's LDAP libraries.
2006 Bugzilla 1109, patch from Stephen Usher.
2008 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2010 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2013 TF/02 Fix a couple more cases where we did not log the error message
2014 when unlink() failed. See also change 4.74-TF/03.
2016 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2017 lock up or crash if it happened to be inside a call to libc when it
2018 got a SIGUSR1 from exiwhat.
2020 The SIGUSR1 handler appends the current process status to the process
2021 log which is later printed by exiwhat. It used to use the general
2022 purpose logging code to do this, but several functions it calls are
2023 not safe for signals.
2025 The new output code in the SIGUSR1 handler is specific to the process
2026 log, and simple enough that it's easy to inspect for signal safety.
2027 Removing some special cases also simplifies the general logging code.
2028 Removing the spurious timestamps from the process log simplifies
2031 TF/04 Improved ratelimit ACL condition.
2033 The /noupdate option has been deprecated in favour of /readonly which
2034 has clearer semantics. The /leaky, /strict, and /readonly update modes
2035 are mutually exclusive. The update mode is no longer included in the
2036 database key; it just determines when the database is updated. (This
2037 means that when you upgrade Exim will forget old rate measurements.)
2039 Exim now checks that the per_* options are used with an update mode that
2040 makes sense for the current ACL. For example, when Exim is processing a
2041 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2042 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2043 must specify per_mail/readonly. If you omit the update mode it defaults to
2044 /leaky where that makes sense (as before) or /readonly where required.
2046 The /noupdate option is now undocumented but still supported for
2047 backwards compatibility. It is equivalent to /readonly except that in
2048 ACLs where /readonly is required you may specify /leaky/noupdate or
2049 /strict/noupdate which are treated the same as /readonly.
2051 A useful new feature is the /count= option. This is a generalization
2052 of the per_byte option, so that you can measure the throughput of other
2053 aggregate values. For example, the per_byte option is now equivalent
2054 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2056 The per_rcpt option has been generalized using the /count= mechanism
2057 (though it's more complicated than the per_byte equivalence). When it is
2058 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2059 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2060 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2061 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2062 Note that using per_rcpt with a non-readonly update mode in more than
2063 one ACL will cause the recipients to be double-counted. (The per_mail
2064 and per_byte options don't have this problem.)
2066 The handling of very low rates has changed slightly. If the computed rate
2067 is less than the event's count (usually one) then this event is the first
2068 after a long gap. In this case the rate is set to the same as this event's
2069 count, so that the first message of a spam run is counted properly.
2071 The major new feature is a mechanism for counting the rate of unique
2072 events. The new per_addr option counts the number of different
2073 recipients that someone has sent messages to in the last time period. It
2074 behaves like per_rcpt if all the recipient addresses are different, but
2075 duplicate recipient addresses do not increase the measured rate. Like
2076 the /count= option this is a general mechanism, so the per_addr option
2077 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2078 example, measure the rate that a client uses different sender addresses
2079 with the options per_mail/unique=$sender_address. There are further
2080 details in the main documentation.
2082 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2084 TF/06 Removed a few PCRE remnants.
2086 TF/07 Automatically extract Exim's version number from tags in the git
2087 repository when doing development or release builds.
2089 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2090 Bugzilla 879. Patch from Paul Fisher.
2092 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2093 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2096 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2098 PP/05 Variable $av_failed, true if the AV scanner deferred.
2099 Bugzilla 1078. Patch from John Horne.
2101 PP/06 Stop make process more reliably on build failure.
2102 Bugzilla 1087. Patch from Heiko Schlittermann.
2104 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2105 Bugzilla 1089. Patch from Heiko Schlittermann.
2107 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2108 Bugzilla 1131. Patch from Holger Weiß.
2110 PP/09 Handle IPv6 addresses with SPF.
2111 Bugzilla 860. Patch from Wolfgang Breyha.
2113 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2115 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2118 PP/11 match_* no longer expand right-hand-side by default.
2119 New compile-time build option, EXPAND_LISTMATCH_RHS.
2120 New expansion conditions, "inlist", "inlisti".
2122 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2124 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2126 PP/14 fix log_write() format string regression from TF/03.
2127 Bugzilla 1152. Patch from Dmitry Isaikin.
2133 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2135 PP/02 Harmonised TLS library version reporting; only show if debugging.
2136 Layout now matches that introduced for other libraries in 4.74 PP/03.
2138 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2140 PP/04 New "dns_use_edns0" global option.
2142 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2145 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2146 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2148 TK/01 Updated PolarSSL code to 0.14.2.
2149 Bugzilla 1097. Patch from Andreas Metzler.
2151 PP/07 Catch divide-by-zero in ${eval:...}.
2152 Fixes bugzilla 1102.
2154 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2157 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2158 format-string attack -- SECURITY: remote arbitrary code execution.
2160 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2161 time unintentionally subject to list matching rules, letting the header
2162 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2163 arbitrary string expansion). This allowed for information disclosure.
2165 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2166 INT_MIN/-1 -- value coerced to INT_MAX.
2172 NM/01 Workaround for PCRE version dependency in version reporting
2175 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2176 This fixes portability to compilers other than gcc, notably
2177 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2179 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2180 makefiles for portability to HP-UX and POSIX correctness.
2182 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2183 Also via indented variable definition in the Makefile.
2184 (Debugging by Oliver Heesakkers).
2186 PP/02 Restore caching of spamd results with expanded spamd_address.
2187 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2189 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2190 Improves build reliability. Fix from: Frank Elsner
2192 NM/02 Fix wide character breakage in the rfc2047 coding
2193 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2195 NM/03 Allow underscore in dnslist lookups
2196 Fixes bug 1026. Patch from Graeme Fowler
2198 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2199 Code patches from Adam Ciarcinski of NetBSD.
2201 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2204 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2205 is logged, to avoid truncation. Patch from John Horne.
2207 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2208 Patch from Jakob Hirsch.
2210 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2211 SQL string expansion failure details.
2212 Patch from Andrey Oktyabrski.
2214 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2215 Patch from Simon Arlott.
2217 PP/09 New lookups functionality failed to compile on old gcc which rejects
2218 extern declarations in function scope.
2219 Patch from Oliver Fleischmann
2221 PP/10 Use sig_atomic_t for flags set from signal handlers.
2222 Check getgroups() return and improve debugging.
2223 Fixed developed for diagnosis in bug 927 (which turned out to be
2226 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2227 Patch from Mark Zealey.
2229 PP/12 Bugzilla 1056: Improved spamd server selection.
2230 Patch from Mark Zealey.
2232 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2233 Based on patch from Heiko Schlittermann.
2235 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2236 Patch from Uwe Doering, sign-off by Michael Haardt.
2238 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2239 variable declaration deep within a block. Bug and patch from
2242 PP/15 lookups-Makefile IRIX compatibility coercion.
2244 PP/16 Make DISABLE_DKIM build knob functional.
2246 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2247 Patch by Simon Arlott
2249 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2250 variable argument macros. Our copy now differs from upstream.
2256 TF/01 Failure to get a lock on a hints database can have serious
2257 consequences so log it to the panic log.
2259 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2260 controlled using the smtp_confirmation log selector.
2262 TF/03 Include the error message when we fail to unlink a spool file.
2264 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2265 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2266 for maintaining out-of-tree patches for some time.
2268 PP/01 Bugzilla 139: Documentation and portability issues.
2269 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2270 Handle per-OS dynamic-module compilation flags.
2272 PP/02 Let /dev/null have normal permissions.
2273 The 4.73 fixes were a little too stringent and complained about the
2274 permissions on /dev/null. Exempt it from some checks.
2275 Reported by Andreas M. Kirchwitz.
2277 PP/03 Report version information for many libraries, including
2278 Exim version information for dynamically loaded libraries. Created
2279 version.h, now support a version extension string for distributors
2280 who patch heavily. Dynamic module ABI change.
2282 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2283 privilege escalation vulnerability whereby the Exim run-time user
2284 can cause root to append content of the attacker's choosing to
2287 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2290 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2291 If dropping privileges for untrusted macros, we disabled normal logging
2292 on the basis that it would fail; for the Exim run-time user, this is not
2293 the case, and it resulted in successful deliveries going unlogged.
2294 Fixed. Reported by Andreas Metzler.
2300 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2301 only prepend for the Resent-* case. Fixes regression introduced in
2302 Exim 4.70 by NM/22 for Bugzilla 607.
2304 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2305 increasing numbers of administrators be bitten by this.
2307 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2309 PP/03 Bugzilla 994: added openssl_options main configuration option.
2311 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2313 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2315 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2317 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2318 without a peer certificate, leading to a segfault because of an
2319 assumption that peers always have certificates. Be a little more
2320 paranoid. Problem reported by Martin Tscholak.
2322 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2323 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2324 NB: ClamAV planning to remove STREAM in "middle of 2010".
2325 CL also introduces -bmalware, various -d+acl logging additions and
2326 more caution in buffer sizes.
2328 PP/09 Implemented reverse_ip expansion operator.
2330 PP/10 Bugzilla 937: provide a "debug" ACL control.
2332 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2334 PP/12 Bugzilla 973: Implement --version.
2336 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2338 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2340 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2342 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2343 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2344 ignore trailing whitespace.
2346 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2348 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2351 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2352 Notification from Dr Andrew Aitchison.
2354 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2355 ExtendedDetectionInfo response format.
2356 Notification from John Horne.
2358 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2361 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2362 XSL and documented dependency on system catalogs, with examples of how
2365 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2368 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2369 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2370 configuration file which is writeable by the Exim user or group.
2372 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2373 of configuration files to cover files specified with the -C option if
2374 they are going to be used with root privileges, not just the default
2377 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2378 option (effectively making it always true).
2380 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2381 files to be used while preserving root privileges.
2383 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2384 that rogue child processes cannot use them.
2386 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2387 run-time user, instead of root.
2389 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2390 Exim run-time user without dropping privileges.
2392 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2393 result string, instead of calling string_vformat() twice with the same
2396 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2397 for other users. Others should always drop root privileges if they use
2398 -C on the command line, even for a whitelisted configure file.
2400 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2402 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2408 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2409 $data_path, and $header_path variables; fixed documentation bugs and
2412 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2413 exipick to access non-standard spools, including the "frozen" queue
2416 NM/01 Bugzilla 965: Support mysql stored procedures.
2417 Patch from Alain Williams
2419 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2421 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2422 Patch from Andreas Metzler
2424 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2425 Patch from Kirill Miazine
2427 NM/05 Bugzilla 671: Added umask to procmail example.
2429 JJ/03 installed exipick 20100323.0, fixing doc bug
2431 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2432 directory. Notification and patch from Dan Rosenberg.
2434 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2436 TK/02 Improve log output when DKIM signing operation fails.
2438 MH/01 Treat the transport option dkim_domain as a colon separated
2439 list, not as a single string, and sign the message with each element,
2440 omitting multiple occurences of the same signer.
2442 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2443 Bugzilla 985, 986. Patch by Simon Arlott
2445 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2446 Patch by Simon Arlott
2448 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2449 MBX locking. Notification from Dan Rosenberg.
2455 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2457 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2459 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2461 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2463 NM/04 Bugzilla 915: Flag broken perl installation during build.
2469 TK/01 Added patch by Johannes Berg that expands the main option
2470 "spamd_address" if it starts with a dollar sign.
2472 TK/02 Write list of recipients to X-Envelope-Sender header when building
2473 the mbox-format spool file for content scanning (suggested by Jakob
2476 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2477 (http://www.dcc-servers.net/) support via dccifd. Activated by
2478 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2480 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2481 by Mark Daniel Reidel <mr@df.eu>.
2483 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2484 When building exim an external PCRE library is now needed -
2485 PCRE is a system library on the majority of modern systems.
2486 See entry on PCRE_LIBS in EDITME file.
2488 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2489 conversation. Added nologin parameter to request.
2490 Patch contributed by Kirill Miazine.
2492 TF/01 Do not log submission mode rewrites if they do not change the address.
2494 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2496 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2497 log files in place. Contributed by Roberto Lima.
2499 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2501 TF/03 Bugzilla 615: When checking the local_parts router precondition
2502 after a local_part_suffix or local_part_prefix option, Exim now
2503 does not use the address's named list lookup cache, since this
2504 contains cached lookups for the whole local part.
2506 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2507 Robert Millan. Documentation is in experimental-spec.txt.
2509 TF/04 Bugzilla 668: Fix parallel build (make -j).
2511 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2513 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2514 Patch provided by Jan Srzednicki.
2516 TF/05 Leading white space used to be stripped from $spam_report which
2517 wrecked the formatting. Now it is preserved.
2519 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2520 that they are available at delivery time.
2522 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2524 TF/08 TLS error reporting now respects the incoming_interface and
2525 incoming_port log selectors.
2527 TF/09 Produce a more useful error message if an SMTP transport's hosts
2528 setting expands to an empty string.
2530 NM/06 Bugzilla 744: EXPN did not work under TLS.
2531 Patch provided by Phil Pennock.
2533 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2534 Patch provided by Richard Godbee.
2536 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2537 acl_smtp_notquit, added index entry.
2539 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2540 Patch provided by Eugene Bujak.
2542 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2543 accept(). Patch provided by Maxim Dounin.
2545 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2546 Patch provided by Phil Pennock.
2548 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2550 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2551 Patch provided by Brad "anomie" Jorsch.
2553 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2554 Patch provided by Dean Brooks.
2556 TK/05 Add native DKIM support (does not depend on external libraries).
2558 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2559 Patch provided by Graeme Fowler.
2561 NM/16 Bugzilla 851: Documentation example syntax fix.
2563 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2565 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2568 NM/19 Bugzilla 745: TLS version reporting.
2569 Patch provided by Phil Pennock.
2571 NM/20 Bugzilla 167: bool: condition support.
2572 Patch provided by Phil Pennock.
2574 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2575 clients. Patch provided by Phil Pennock.
2577 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2578 Patch provided by Brad "anomie" Jorsch.
2580 NM/23 Bugzilla 687: Fix misparses in eximstats.
2581 Patch provided by Heiko Schlittermann.
2583 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2584 Patch provided by Heiko Schlittermann.
2586 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2587 plus update to original patch.
2589 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2591 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2592 Patch provided by David Brownlee.
2594 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2596 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2598 NM/30 Bugzilla 888: TLS documentation bugfixes.
2600 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2602 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2603 Unlike the original bugzilla I have changed all shell scripts in src tree.
2605 NM/33 Bugzilla 898: Transport filter timeout fix.
2606 Patch by Todd Rinaldo.
2608 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2609 Patch by Serge Demonchaux.
2611 NM/35 Bugzilla 39: Base64 decode bug fixes.
2612 Patch by Jakob Hirsch.
2614 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2616 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2618 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2620 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2626 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2627 ALT-N's libdkim that I have put here:
2628 http://duncanthrax.net/exim-experimental/
2630 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2631 were called 'true' and it seems that C99 defines that as a reserved
2632 keyword to be used with 'bool' variable types. That means you could
2633 not include C99-style headers which use bools without triggering
2634 build errors in sieve.c.
2636 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2637 as mailq or other aliases. Changed the --help handling significantly
2638 to do whats expected. exim_usage() emits usage/help information.
2640 SC/01 Added the -bylocaldomain option to eximstats.
2642 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2644 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2646 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2652 PH/01 Another patch from the Sieve maintainer.
2654 PH/02 When an IPv6 address is converted to a string for single-key lookup
2655 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2656 dots are used instead of colons so that keys in lsearch files need not
2657 contain colons. This was done some time before quoting was made available
2658 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2659 (notated using the quote facility) so as to distinguish them from IPv4
2660 keys. This meant that lookups for IP addresses in host lists did not work
2661 for iplsearch lookups.
2663 This has been fixed by arranging for IPv6 addresses to be expressed with
2664 colons if the lookup type is iplsearch. This is not incompatible, because
2665 previously such lookups could never work.
2667 The situation is now rather anomalous, since one *can* have colons in
2668 ordinary lsearch keys. However, making the change in all cases is
2669 incompatible and would probably break a number of configurations.
2671 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2674 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2675 conversion specification without a maximum field width, thereby enabling
2676 a rogue spamd server to cause a buffer overflow. While nobody in their
2677 right mind would setup Exim to query an untrusted spamd server, an
2678 attacker that gains access to a server running spamd could potentially
2679 exploit this vulnerability to run arbitrary code as the Exim user.
2681 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2682 $primary_hostname instead of what libspf2 thinks the hosts name is.
2684 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2685 a directory entry by the name of the lookup key. Previously, if a
2686 symlink pointed to a non-existing file or a file in a directory that
2687 Exim lacked permissions to read, a lookup for a key matching that
2688 symlink would fail. Now it is enough that a matching directory entry
2689 exists, symlink or not. (Bugzilla 503.)
2691 PH/03 The body_linecount and body_zerocount variables are now exported in the
2694 PH/04 Added the $dnslist_matched variable.
2696 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2697 This means they are set thereafter only if the connection becomes
2700 PH/06 Added the client_condition to authenticators so that some can be skipped
2701 by clients under certain conditions.
2703 PH/07 The error message for a badly-placed control=no_multiline_responses left
2704 "_responses" off the end of the name.
2706 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2708 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2709 (without spaces) instead of just copying the configuration text.
2711 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2713 PH/11 Added $max_received_linelength.
2715 PH/12 Added +ignore_defer and +include_defer to host lists.
2717 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2718 way in which PCRE > 7.0 is built.
2720 PH/14 Implemented queue_only_load_latch.
2722 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2723 MAIL command. The effect was to mangle the value on 64-bit systems.
2725 PH/16 Another patch from the Sieve maintainer.
2727 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2729 PH/18 If a system quota error occurred while trying to create the file for
2730 a maildir delivery, the message "Mailbox is full" was not appended to the
2731 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2732 only to a quota excession during the actual writing of the file.
2734 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2735 characters?) which causes problems in log lines. The DN values are now
2736 passed through string_printing() before being added to log lines.
2738 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2739 and InterBase are left for another time.)
2741 PH/21 Added message_body_newlines option.
2743 PH/22 Guard against possible overflow in moan_check_errorcopy().
2745 PH/23 POSIX allows open() to be a macro; guard against that.
2747 PH/24 If the recipient of an error message contained an @ in the local part
2748 (suitably quoted, of course), incorrect values were put in $domain and
2749 $local_part during the evaluation of errors_copy.
2755 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2756 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2759 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2760 issue a MAIL command.
2762 PH/02 In an ACL statement such as
2764 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2766 if a client was not listed at all, or was listed with a value other than
2767 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2768 the condition was not true (as it should be), so access was not denied.
2769 The bug was that the ! inversion was incorrectly passed on to the second
2770 item. This has been fixed.
2772 PH/03 Added additional dnslists conditions == and =& which are different from
2773 = and & when the dns lookup returns more than one IP address.
2775 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2776 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2778 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2779 FSYNC, which compiles an option called disable_fsync that allows for
2780 bypassing fsync(). The documentation is heavily laced with warnings.
2782 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2784 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2785 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2786 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2787 including adding "make clean"; (3) Added -fPIC when compiling the test
2788 dynamically loaded module, to get rid of a warning.
2790 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2791 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2792 = 0s. The bug is otherwise harmless.
2794 PH/07 There was a bug in the dovecot authenticator such that the value of
2795 $auth1 could be overwritten, and so not correctly preserved, after a
2796 successful authentication. This usually meant that the value preserved by
2797 the server_setid option was incorrect.
2799 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2801 PH/09 Installed PCRE release 7.0.
2803 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2804 run for batched SMTP input. It is now run at the start of every message
2805 in the batch. While fixing this I discovered that the process information
2806 (output by running exiwhat) was not always getting set for -bs and -bS
2807 input. This is fixed, and it now also says "batched" for BSMTP.
2809 PH/11 Added control=no_pipelining.
2811 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2812 patch, slightly modified), and move the expansion of helo_data till after
2813 the connection is made in the smtp transport (so it can use these
2816 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2818 PH/14 Added log_selector = +pid.
2820 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2822 PH/16 Add ${if forany and ${if forall.
2824 PH/17 Added dsn_from option to vary the From: line in DSNs.
2826 PH/18 Flush SMTP output before performing a callout, unless control =
2827 no_callout_flush is set.
2829 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2830 was true (the default) a successful delivery failed to delete the retry
2831 item, thus causing premature timeout of the address. The bug is now
2834 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2836 PH/21 Long custom messages for fakedefer and fakereject are now split up
2837 into multiline responses in the same way that messages for "deny" and
2838 other ACL rejections are.
2840 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2841 with slight modification.
2843 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2844 draft, changing the syntax and factoring some duplicate code".
2846 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2847 for deliveries of the second and subsequent messages over the same SMTP
2850 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2851 ${reduce, with only minor "tidies".
2853 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2855 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2856 expansion side effects.
2858 PH/27 When a message times out after an over-quota error from an Exim-imposed
2859 quota, the bounce message says "mailbox is full". This message was not
2860 being given when it was a system quota that was exceeded. It now should
2863 MH/03 Made $recipients available in local_scan(). local_scan() already has
2864 better access to the recipient list through recipients_list[], but
2865 $recipients can be useful in postmaster-provided expansion strings.
2867 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2868 in the case of a MAIL command with additional options following the
2869 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2870 were accidentally chopped off.
2872 PH/29 SMTP synchronization checks are implemented when a command is read -
2873 there is a check that no more input is waiting when there shouldn't be
2874 any. However, for some commands, a delay in an ACL can mean that it is
2875 some time before the response is written. In this time, more input might
2876 arrive, invalidly. So now there are extra checks after an ACL has run for
2877 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2878 pipelining has not been advertised.
2880 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2882 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2883 returned in the SMTP response when smtp_return_error_details was false.
2884 This has been fixed.
2886 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2887 instead of the C I/O that was originally supplied, because problems were
2888 reported on Solaris.
2890 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2891 Exim which did not show up earlier: it was assuming that a call to
2892 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2893 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2894 was a macro that became an assignment, so it seemed to work. This has
2895 changed to a proper function call with a void return, hence the compile
2896 error. Exim's code has been fixed.
2898 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2901 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2903 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2905 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2906 because it turns out to be too big or there is a timeout) there is no
2907 "Completed" line in the log. When some messages of this type were
2908 selected by exigrep, they were listed as "not completed". Others were
2909 picked up by some special patterns. I have improved the selection
2910 criteria to be more general.
2912 PH/38 The host_find_failed option in the manualroute router can now be set
2913 to "ignore", to completely ignore a host whose IP address cannot be
2914 found. If all hosts are ignored, the behaviour is controlled by the new
2915 host_all_ignored option.
2917 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2918 homing or because of multiple MX records with /mx) generated more than
2919 one IP address, and the following item turned out to be the local host,
2920 all the secondary addresses of the first item were incorrectly removed
2921 from the list, along with the local host and any following hosts (which
2922 is what is supposed to happen).
2924 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2925 whoever called Exim into the -H file. In the case of the daemon it was
2926 behaving confusingly. When first started, it used values for whoever
2927 started the daemon, but after a SIGHUP it used the Exim user (because it
2928 calls itself on a restart). I have changed the code so that it now always
2931 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2932 message are rejected with the same error (e.g. no authentication or bad
2933 sender address), and a DATA command is nevertheless sent (as can happen
2934 with PIPELINING or a stupid MUA), the error message that was given to the
2935 RCPT commands is included in the rejection of the DATA command. This is
2936 intended to be helpful for MUAs that show only the final error to their
2939 PH/42 Another patch from the Sieve maintainer.
2941 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2942 Eximstats - Fixed some broken HTML links and added missing column headers
2944 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2945 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2947 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2949 PH/43 Yet another patch from the Sieve maintainer.
2951 PH/44 I found a way to check for a TCP/IP connection going away before sending
2952 the response to the final '.' that terminates a message, but only in the
2953 case where the client has not sent further data following the '.'
2954 (unfortunately, this is allowed). However, in many cases there won't be
2955 any further data because there won't be any more messages to send. A call
2956 to select() can be used: if it shows that the input is "ready", there is
2957 either input waiting, or the socket has been closed. An attempt to read
2958 the next input character can distinguish the two cases. Previously, Exim
2959 would have sent an OK response which the client would never have see.
2960 This could lead to message repetition. This fix should cure that, at
2961 least in a lot of common cases.
2963 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2964 advertised in response to EHLO.
2970 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2971 fixed by 4.65/MH/01 (is this a record?) are fixed:
2973 (i) An empty string was always treated as zero by the numeric comparison
2974 operators. This behaviour has been restored.
2976 (ii) It is documented that the numeric comparison operators always treat
2977 their arguments as decimal numbers. This was broken in that numbers
2978 starting with 0 were being interpreted as octal.
2980 While fixing these problems I realized that there was another issue that
2981 hadn't been noticed. Values of message_size_limit (both the global option
2982 and the transport option) were treated as octal if they started with 0.
2983 The documentation was vague. These values are now always treated as
2984 decimal, and I will make that clear in the documentation.
2990 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2991 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2994 MH/01 Don't check that the operands of numeric comparison operators are
2995 integers when their expansion is in "skipping" mode (fixes bug
2996 introduced by 4.64-PH/07).
2998 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2999 child addresses, Exim now panics and dies. Previously, because the count
3000 is held in a short int, deliveries were likely to be lost. As such a
3001 large number of recipients for a single message is ridiculous
3002 (performance will be very, very poor), I have chosen to impose a limit
3003 rather than extend the field.
3009 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3010 leftover -K file (the existence of which was triggered by #402).
3011 While we were at it, introduced process PID as part of the -K
3012 filename. This should rule out race conditions when creating
3015 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3016 processing considerably. Previous code took too long for large mails,
3017 triggering a timeout which in turn triggers #401.
3019 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3020 in the DK code in transports.c. sendfile() is not really portable,
3021 hence the _LINUX specificness.
3023 TF/01 In the add_headers option to the mail command in an Exim filter,
3024 there was a bug that Exim would claim a syntax error in any
3025 header after the first one which had an odd number of characters
3028 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3029 callout verification, Exim cached a "reject" for the entire domain. This
3030 is correct for most verifications, but it is not correct for a recipient
3031 verification with use_sender or use_postmaster set, because in that case
3032 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3033 case of MAIL FROM:<> rejection from other early rejections (e.g.
3034 rejection of HELO). When verifying a recipient using a non-null MAIL
3035 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3036 Whatever the result of the callout, the value of the domain cache is
3037 left unchanged (for any other kind of callout, getting as far as trying
3038 RCPT means that the domain itself is ok).
3040 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3043 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3044 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3045 session, the daemon ignores ECONNECT errors and logs others; it now
3046 ignores EPIPE as well.
3048 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3049 (quoted-printable decoding).
3051 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3052 later the small subsequent patch to fix an introduced bug.
3054 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3056 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3058 PH/08 An error is now given if message_size_limit is specified negative.
3060 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3061 to be given (somewhat) arbitrary names.
3063 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3066 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3067 miscellaneous code fixes
3069 PH/10 Added the log_reject_target ACL modifier to specify where to log
3072 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3073 hostname. This is wrong, because it relates to the incoming message (and
3074 probably the interface on which it is arriving) and not to the outgoing
3075 callout (which could be using a different interface). This has been
3076 changed to use the value of the helo_data option from the smtp transport
3077 instead - this is what is used when a message is actually being sent. If
3078 there is no remote transport (possible with a router that sets up host
3079 addresses), $smtp_active_hostname is used.
3081 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3082 tweaks were necessary in order to get it to work (see also 21 below):
3083 (a) The code assumed that strncpy() returns a negative number on buffer
3084 overflow, which isn't the case. Replaced with Exim's string_format()
3086 (b) There were several signed/unsigned issues. I just did the minimum
3087 hacking in of casts. There is scope for a larger refactoring.
3088 (c) The code used strcasecmp() which is not a standard C function.
3089 Replaced with Exim's strcmpic() function.
3090 (d) The code set only $1; it now sets $auth1 as well.
3091 (e) A simple test gave the error "authentication client didn't specify
3092 service in request". It would seem that Dovecot has changed its
3093 interface. Fortunately there's a specification; I followed it and
3094 changed what the client sends and it appears to be working now.
3096 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3099 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3100 address A is aliased to B and C, where B exists and C does not. Without
3101 -v the output is "A verified" because verification stops after a
3102 successful redirection if more than one address is generated. However,
3103 with -v the child addresses are also verified. Exim was outputting "A
3104 failed to verify" and then showing the successful verification for C,
3105 with its parentage. It now outputs "B failed to verify", showing B's
3106 parentage before showing the successful verification of C.
3108 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3109 look up a TXT record in a specific list after matching in a combined
3112 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3113 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3114 they consult the DNS. I had assumed they would set it the way they
3115 wanted; and indeed my experiments on Linux seem to show that in some
3116 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3117 To be on the safe side, however, I have now made the interface to
3118 host_find_byname() similar to host_find_bydns(), with an argument
3119 containing the DNS resolver options. The host_find_byname() function now
3120 sets these options at its start, just as host_find_bydns() does. The smtp
3121 transport options dns_qualify_single and dns_search_parents are passed to
3122 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3123 of host_find_byname() use the default settings of RES_DEFNAMES
3124 (qualify_single) but not RES_DNSRCH (search_parents).
3126 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3127 spool_read_header() do less string testing, by means of a preliminary
3128 switch on the second character of optional "-foo" lines. (This is
3129 overdue, caused by the large number of possibilities that now exist.
3130 Originally there were few.) While I was there, I also converted the
3131 str(n)cmp tests so they don't re-test the leading "-" and the first
3132 character, in the hope this might squeeze out yet more improvement.
3134 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3135 flag allowing group syntax was set by the header_syntax check but not
3136 turned off, possible causing trouble later; (2) The flag was not being
3137 set at all for the header_verify test, causing "group"-style headers to
3138 be rejected. I have now set it in this case, and also caused header_
3139 verify to ignore an empty address taken from a group. While doing this, I
3140 came across some other cases where the code for allowing group syntax
3141 while scanning a header line wasn't quite right (mostly, not resetting
3142 the flag correctly in the right place). These bugs could have caused
3143 trouble for malformed header lines. I hope it is now all correct.
3145 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3146 with the "reply" argument non-NULL. The code, however (which originally
3147 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3148 but it didn't always do it. This confused somebody who was copying the
3149 code for some other use. I have removed all the tests.
3151 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3152 feature that was used to support insecure browsers during the U.S. crypto
3153 embargo. It requires special client support, and Exim is probably the
3154 only MTA that supported it -- and would never use it because real RSA is
3155 always available. This code has been removed, because it had the bad
3156 effect of slowing Exim down by computing (never used) parameters for the
3157 RSA_EXPORT functionality.
3159 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3160 authenticator to fail if there's a tab character in the incoming data
3161 (there should never be unless someone is messing about, as it's supposed
3162 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3163 if the connection is using TLS or if the remote IP is the same as the
3164 local IP, and the "valid-client-cert option" if a client certificate has
3167 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3168 authenticators. This can be used for authorization after authentication
3169 succeeds. (In the case of plaintext, it servers for both authentication
3172 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3173 if any retry times were supplied.
3175 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3176 connection, where there is no client IP address to check. In this
3177 situation, the verify now always succeeds.
3179 PH/25 Applied John Jetmore's -Mset patch.
3181 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3183 PH/27 In a string expansion for a processed (not raw) header when multiple
3184 headers of the same name were present, leading whitespace was being
3185 removed from all of them, but trailing whitespace was being removed only
3186 from the last one. Now trailing whitespace is removed from each header
3187 before concatenation. Completely empty headers in a concatenation (as
3188 before) are ignored.
3190 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3191 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3193 PH/29 [Removed. This was a change that I later backed out, and forgot to
3194 correct the ChangeLog entry (that I had efficiently created) before
3195 committing the later change.]
3197 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3198 address errors (4xx response to RCPT) over the same connection as other
3199 messages routed to the same hosts. Such deliveries are always "forced",
3200 so retry times are not inspected. This resulted in far too many retries
3201 for the affected addresses. The effect occurred only when there were more
3202 hosts than the hosts_max_try setting in the smtp transport when it had
3203 the 4xx errors. Those hosts that it had tried were not added to the list
3204 of hosts for which the message was waiting, so if all were tried, there
3205 was no problem. Two fixes have been applied:
3207 (i) If there are any address or message errors in an SMTP delivery, none
3208 of the hosts (tried or untried) are now added to the list of hosts
3209 for which the message is waiting, so the message should not be a
3210 candidate for sending over the same connection that was used for a
3211 successful delivery of some other message. This seems entirely
3212 reasonable: after all the message is NOT "waiting for some host".
3213 This is so "obvious" that I'm not sure why it wasn't done
3214 previously. Hope I haven't missed anything, but it can't do any
3215 harm, as the worst effect is to miss an optimization.
3217 (ii) If, despite (i), such a delivery is accidentally attempted, the
3218 routing retry time is respected, so at least it doesn't keep
3219 hammering the server.
3221 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3222 in ${readsocket because some servers need this prod.
3224 PH/32 Added some extra debug output when updating a wait-xxx database.
3226 PH/33 The hint "could be header name not terminated by colon", which has been
3227 given for certain expansion errors for a long time, was not being given
3228 for the ${if def:h_colon_omitted{... case.
3230 PH/34 The spec says: "With one important exception, whenever a domain list is
3231 being scanned, $domain contains the subject domain." There was at least
3232 one case where this was not true.
3234 PH/35 The error "getsockname() failed: connection reset by peer" was being
3235 written to the panic log as well as the main log, but it isn't really
3236 panic-worthy as it just means the connection died rather early on. I have
3237 removed the panic log writing for the ECONNRESET error when getsockname()
3240 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3241 runs only) independently of the message's sender address. This meant
3242 that, if the 4xx error was in fact related to the sender, a different
3243 message to the same recipient with a different sender could confuse
3244 things. In particular, this can happen when sending to a greylisting
3245 server, but other circumstances could also provoke similar problems.
3246 I have changed the default so that the retry time for these errors is now
3247 based a combination of the sender and recipient addresses. This change
3248 can be overridden by setting address_retry_include_sender=false in the
3251 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3252 remote server are returned as part of bounce messages. This was not
3253 happening for LMTP over a pipe (the lmtp transport), but now it is the
3254 same for both kinds of LMTP.
3256 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3257 in header lines that were in fact CNAMEs. This is no longer the case.
3259 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3260 and queue runs started by the daemon processed all messages. This has
3261 been fixed so that -R and -S can now usefully be given with -q<time>.
3263 PH/40 Import PCRE release 6.7 (fixes some bugs).
3265 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3267 PH/42 Give an error if -q is specified more than once.
3269 PH/43 Renamed the variables $interface_address and $interface_port as
3270 $received_ip_address and $received_port, to make it clear that these
3271 values apply to message reception, and not to the outgoing interface when
3272 a message is delivered. (The old names remain recognized, of course.)
3274 PH/44 There was no timeout on the connect() call when using a Unix domain
3275 socket in the ${readsocket expansion. There now is.
3277 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3278 be meaningful with "accept".
3280 SC/01 Eximstats V1.43
3281 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3283 SC/02 Eximstats V1.44
3284 Use a glob alias rather than an array ref in the generated
3285 parser. This improves both readability and performance.
3287 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3288 Collect SpamAssassin and rejection statistics.
3289 Don't display local sender or destination tables unless
3290 there is data to show.
3291 Added average volumes into the top table text output.
3293 SC/04 Eximstats V1.46
3294 Collect data on the number of addresses (recipients)
3295 as well as the number of messages.
3297 SC/05 Eximstats V1.47
3298 Added 'Message too big' to the list of mail rejection
3299 reasons (thanks to Marco Gaiarin).
3301 SC/06 Eximstats V1.48
3302 Mainlog lines which have GMT offsets and are too short to
3303 have a flag are now skipped.
3305 SC/07 Eximstats V1.49 (Alain Williams)
3306 Added the -emptyok flag.
3308 SC/08 Eximstats V1.50
3309 Fixes for obtaining the IP address from reject messages.
3311 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3312 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3313 whitespace changes from 4.64-PH/27
3315 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3318 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3319 are found, allow negative numbers in numeric criteria)
3321 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3323 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3324 to match changes made in 4.64-PH/43
3326 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3328 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3329 "domain" is actually an IP address, and force a failure. This locks out
3330 those revolvers/nameservers that support "A-for-A" lookups, in
3331 contravention of the specifications.
3333 PH/48 When a host name was looked up from an IP address, and the subsequent
3334 forward lookup of the name timed out, the host name was left in
3335 $sender_host_name, contrary to the specification.
3337 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3338 restricted to single-key lookups, Exim was not diagnosing an error if
3339 * or *@ was used with a query-style lookup.
3341 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3343 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3344 long ago, but noone interested enough thought of it. Let's just say that
3345 the "1.1" means that there are some new functions that weren't there at
3346 some point in the past.
3348 PH/51 Error processing for expansion failure of helo_data from an smtp
3349 transport during callout processing was broken.
3351 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3352 tested/used via the -bh/-bhc/-bs options.
3354 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3355 bug, fixed in subsequent PCRE releases).
3357 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3358 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3360 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3366 SC/01 Use a glob alias rather than an array ref in eximstats generated
3367 parser. This improves both readability and performance.
3369 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3370 Don't display local sender or destination tables in eximstats unless
3371 there is data to show.
3372 Added average volumes into the eximstats top table text output.
3374 SC/03 Collect data on the number of addresses (recipients) as well
3375 as the number of messages in eximstats.
3377 TF/01 Correct an error in the documentation for the redirect router. Exim
3378 does (usually) call initgroups() when daemonizing.
3380 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3381 with consistent privilege compared to when running as a daemon.
3383 TF/03 Note in the spec that $authenticated_id is not set for local
3384 submissions from trusted users.
3386 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3387 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3389 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3390 by adding some example configuration directives to the default
3391 configuration file. A little bit of work is required to uncomment the
3392 directives and define how usernames and passwords are checked, but
3393 there is now a framework to start from.
3395 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3396 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3397 without this. I don't know how relevant this is to other LDAP libraries.
3399 PH/02 Add the verb name to the "unknown ACL verb" error.
3401 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3403 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3405 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3406 directory not expanded when it should be if an expanded home directory
3407 was set for the address (which is overridden by the transport).
3409 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3412 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3413 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3414 because it is too late at that time, and has no effect.
3416 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3417 security issue with \' (bugzilla #107). I could not use the
3418 PQescapeStringConn() function, because it needs a PGconn value as one of
3421 PH/08 When testing addresses using -bt, indicate those final addresses that
3422 are duplicates that would not cause an additional delivery. At least one
3423 person was confused, thinking that -bt output corresponded to deliveries.
3424 (Suppressing duplicates isn't a good idea as you lose the information
3425 about possibly different redirections that led to the duplicates.)
3427 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3428 systems where poll() doesn't work, in particular OS X.
3430 PH/10 Added more information to debugging output for retry time not reached.
3432 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3433 operations in malware.c.
3435 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3438 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3439 syslog_duplication set false, log lines that would normally be written
3440 both the the main log and to the reject log were not written to syslog at
3443 PH/14 In the default configuration, change the use of "message" in ACL warn
3444 statements to "add_header".
3446 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3447 not followed by a command (e.g. "seen endif").
3449 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3450 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3453 PH/17 Added extra conditions to the default value of delay_warning_condition
3457 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3458 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3459 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3462 The Auto-Submitted: and various List- headers are standardised, whereas I
3463 don't think Precedence: ever was.
3465 PH/18 Refactored debugging code in route_finduser() to show more information,
3466 in particular, the error code if getpwnam() issues one.
3468 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3469 This is apparently needed in addition to the PH/07 change above to avoid
3470 any possible encoding problems.
3472 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3473 but not after initializing Perl.
3475 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3476 output them only if debugging. By default they are written stderr,
3477 apparently, which is not desirable.
3479 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3482 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3485 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3487 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3488 authenticated or an ident call has been made. Suppress the default
3489 values for $authenticated_id and $authenticated_sender (but permit -oMai
3490 and -oMas) when testing with -bh.
3492 PH/24 Re-jigged the order of the tests in the default configuration so that the
3493 tests for valid domains and recipients precede the DNS black list and CSA
3494 tests, on the grounds that those ones are more expensive.
3496 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3497 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3498 This bug exists in every version of Exim that I still have, right back to
3501 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3502 However, an attempt to turn on case-sensitivity in a regex key by
3503 including (?-i) didn't work because the subject string was already
3504 lowercased, and the effects were non-intuitive. It turns out that a
3505 one-line patch can be used to allow (?-i) to work as expected.
3511 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3512 other effects) broke the use of negated acl sub-conditions.
3514 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3517 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3518 "Deny" causes Exim to reject the incoming connection with a 554 error.
3519 Unfortunately, if there is a major crisis, such as a disk failure,
3520 tcp-wrappers gives "deny", whereas what one would like would be some
3521 kind of temporary error. A kludge has been added to help with this.
3522 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3523 554 error is used if errno is still zero or contains ENOENT (which occurs
3524 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3527 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3529 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3530 errors. Otherwise a message that provokes a temporary error (when other
3531 messages do not) can cause a whole host to time out.
3533 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3534 addresses were routed directly to files or pipes from a redirect router.
3535 File deliveries just didn't batch; pipe deliveries might have suffered
3538 PH/06 A failure to get a lock for a hints database would erroneously always say
3539 "Failed to get write lock", even when it was really a read lock.
3541 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3542 of 0600. This has been changed to use the value of the lockfile_mode
3543 option (which defaults to 0600).
3545 PH/08 Applied small patch from the Sieve maintainer.
3547 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3548 folder from quota calculations, a direct delivery into this folder messed
3549 up the contents of the maildirsize file. This was because the regex was
3550 used only to exclude .Trash (or whatever) when the size of the mailbox
3551 was calculated. There was no check that a delivery was happening into an
3552 excluded directory. This bug has been fixed by ignoring all quota
3553 processing for deliveries into excluded directories.
3555 PH/10 Added the maildirfolder_create_regex option to appendfile.
3561 PH/01 The code for finding all the local interface addresses on a FreeBSD
3562 system running IPv6 was broken. This may well have applied to all BSD
3563 systems, as well as to others that have similar system calls. The broken
3564 code found IPv4 interfaces correctly, but gave incorrect values for the
3565 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3566 that it would not match correctly against @[] and not recognize the IPv6
3569 PH/02 The ipliteral router was not recognizing addresses of the form user@
3570 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3572 PH/03 Added disable_ipv6.
3574 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3575 decoded form, because it is most often used to construct To: headers
3576 lines in autoreplies, and the decoded form may well be syntactically
3577 invalid. However, $reply_address has leading white space removed, and all
3578 newlines turned into spaces so that the autoreply transport does not
3581 PH/05 If group was specified without a user on a router, and no group or user
3582 was specified on a transport, the group from the router was ignored.
3584 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3585 for visible compile-time settings that can be used to change these
3586 numbers, for those that want even more. Backwards compatibility with old
3587 spool files has been maintained. However, going back to a previous Exim
3588 release will lost any variables that are in spool files.
3590 PH/07 Two small changes when running in the test harness: increase delay when
3591 passing a TCP/IP connection to a new process, in case the original
3592 process has to generate a bounce, and remove special handling of
3593 127.0.0.2 (sic), which is no longer necessary.
3595 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3596 be the same on different OS.
3598 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3601 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3602 whether --show-vars was specified or not
3604 JJ/02 exipick: Added support for new ACL variable spool format introduced
3607 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3608 syntactically invalid From: or Reply-to: line, and a filter used this to
3609 generate an autoreply, and therefore failed to obtain an address for the
3610 autoreply, Exim could try to deliver to a non-existent relative file
3611 name, causing unrelated and misleading errors. What now happens is that
3612 it logs this as a hard delivery error, but does not attempt to create a
3615 PH/11 The exinext utility has a -C option for testing purposes, but although
3616 the given file was scanned by exinext itself; it wasn't being passed on
3617 when Exim was called.
3619 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3620 an end-of-file indication when reading a command response.
3622 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3623 compiled. In many other places in Exim, IPv6 addresses are always
3624 recognized, so I have changed this. It also means that IPv4 domain
3625 literals of the form [IPV4:n.n.n.n] are now always recognized.
3627 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3628 used if the router is not running as root, for example, when verifying at
3629 ACL time, or when using -bh. The debugging output from this situation was
3630 non-existent - all you got was a failure to exec. I have made two
3633 (a) Failures to set uid/gid, the current directory, or a process leader
3634 in a subprocess such as that created by queryprogram now generate
3635 suitable debugging output when -d is set.
3637 (b) The queryprogram router detects when it is not running as root,
3638 outputs suitable debugging information if -d is set, and then runs
3639 the subprocess without attempting to change uid/gid.
3641 PH/15 Minor change to Makefile for building test_host (undocumented testing
3644 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3645 additional section of a DNS packet that returns MX or SRV records.
3646 Instead, it always explicitly searches for A/AAAA records. This avoids
3647 major problems that occur when a DNS server includes only records of one
3648 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3649 fixed another bug: if SRV records were looked up and the corresponding
3650 address records were *not* found in the additional section, the port
3651 values from the SRV records were lost.
3653 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3654 using the correct key (the original address) when searching the retry
3655 rules in order to find which one to use for generating the retry hint.
3657 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3658 adding the default one. Similarly, if it contains a Reply-To: header, the
3659 errors_reply_to option, if set, is not used.
3661 PH/19 When calculating a retry time, Exim used to measure the "time since
3662 failure" by looking at the "first failed" field in the retry record. Now
3663 it does not use this if it is later than than the arrival time of the
3664 message. Instead it uses the arrival time. This makes for better
3665 behaviour in cases where some deliveries succeed, thus re-setting the
3666 "first failed" field. An example is a quota failure for a huge message
3667 when small messages continue to be delivered. Without this change, the
3668 "time since failure" will always be short, possible causing more frequent
3669 delivery attempts for the huge message than are intended.
3670 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3672 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3673 $1, $2, $3) because the numerical variables can be reset during some
3674 expansion items (e.g. "match"), thereby losing the authentication data.
3676 PH/21 Make -bV show the size of off_t variables so that the test suite can
3677 decide whether to run tests for quotas > 2G.
3679 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3680 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3681 filecount value is greater than 2G or if a quota value is greater than 2G
3682 on a system where the size of off_t is not greater than 4, a panic error
3685 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3686 never match. The debug and -bh output now contains an explicit error
3687 message indicating a malformed IPv4 address or mask.
3689 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3690 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3691 PH/23 above applies.
3693 PH/25 Do not write to syslog when running in the test harness. The only
3694 occasion when this arises is a failure to open the main or panic logs
3695 (for which there is an explicit test).
3697 PH/26 Added the /no_tell option to "control=freeze".
3699 PH/27 If a host name lookup failed very early in a connection, for example, if
3700 the IP address matched host_lookup and the reverse lookup yielded a name
3701 that did not have a forward lookup, an error message of the form "no IP
3702 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3703 could be logged. Now it outputs the IP address instead of "NULL".
3705 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3706 allows the sender and the authenticated sender to be set when
3707 submitting a message from within Exim. Since child_open_exim() is
3708 documented for local_scan(), the new function should be too.
3710 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3711 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3712 results in an empty string is now treated as unset.
3714 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3716 PH/31 Added sender_verify_fail logging option.
3718 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3719 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3720 tidied the source and removed it altogether.
3722 PH/33 When a queue run was abandoned because the load average was too high, a
3723 log line was always written; now it is written only if the queue_run log
3724 selector is set. In addition, the log line for abandonment now contains
3725 information about the queue run such as the pid. This is always present
3726 in "start" and "stop" lines but was omitted from the "abandon" line.
3728 PH/34 Omit spaces between a header name and the colon in the error message that
3729 is given when verify = headers_syntax fails (if there are lots of them,
3730 the message gets confusing).
3732 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3733 names, as there are now some PTR records that contain slashes. This check
3734 is only to protect against broken name servers that fall over on strange
3735 characters, so the fact that it applies to all lookups doesn't matter.
3737 PH/36 Now that the new test suite is complete, we can remove some of the
3738 special code in Exim that was needed for the old test suite. For example,
3739 sorting DNS records because real resolvers return them in an arbitrary
3740 order. The new test suite's fake resolver always returns records in the
3743 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3744 bounces) except when queue_only is set, to avoid logging races between
3745 the different processes.
3747 PH/38 Panic-die if .include specifies a non-absolute path.
3749 PH/39 A tweak to the "H" retry rule from its user.
3751 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3752 a label. They prevented compilation on older perls.
3754 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3755 a warning to be raised on newish perls.
3757 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3758 on queue. Changes to match documented behaviour of showing count of
3759 messages matching specified criteria.
3761 PH/40 Changed the default ident timeout from 30s to 5s.
3763 PH/41 Added support for the use of login_cap features, on those BSD systems
3764 that have them, for controlling the resources used by pipe deliveries.
3766 PH/42 The content-scanning code uses fopen() to create files in which to put
3767 message data. Previously it was not paying any attention to the mode of
3768 the files. Exim runs with umask(0) because the rest of the code creates
3769 files with open(), and sets the required mode explicitly. Thus, these
3770 files were ending up world-writeable. This was not a big issue, because,
3771 being within the spool directory, they were not world-accessible. I have
3772 created a function called modefopen, which takes an additional mode
3773 argument. It sets umask(777), creates the file, chmods it to the required
3774 mode, then resets the umask. All the relevant calls to fopen() in the
3775 content scanning code have been changed to use this function.
3777 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3778 to 24 hours. This avoids potential overflow problems when processing G
3779 and H retry rules. I suspect nobody ever tinkers with this value.
3781 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3783 PH/45 When the plaintext authenticator is running as a client, the server's
3784 challenges are checked to ensure they are valid base64 strings. By
3785 default, the authentication attempt is cancelled if an invalid string is
3786 received. Setting client_ignore_invalid_base64 true ignores these errors.
3787 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3788 they are received. Thus, the responses can be made to depend on the
3789 challenges. If an invalid string is ignored, an empty string is placed in
3792 PH/46 Messages that are created by the autoreply transport now contains a
3793 References: header, in accordance with RFCs 2822 and 3834.
3795 PH/47 Added authenticated_sender_force to the smtp transport.
3797 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3799 PH/49 Installed latest patch from the Sieve maintainer.
3801 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3802 was also set, the appendfile transport was unnecessarily scanning a
3803 directory of message files (e.g. for maildir delivery) to find the count
3804 of files (along with the size), even though it did not need this
3805 information. It now does the scan only if it needs to find either the
3806 size of the count of files.
3808 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3810 PH/52 Two bugs concerned with error handling when the smtp transport is
3813 (i) Exim was not creating retry information for temporary errors given
3814 for individual recipients after the DATA command when the smtp transport
3815 was used in LMTP mode. This meant that they could be retried too
3816 frequently, and not timed out correctly.
3818 (ii) Exim was setting the flag that allows error details to be returned
3819 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3820 recipients that were returned after the DATA command.
3822 PH/53 This is related to PH/52, but is more general: for any failing address,
3823 when detailed error information was permitted to be returned to the
3824 sender, but the error was temporary, then after the final timeout, only
3825 "retry timeout exceeded" was returned. Now it returns the full error as
3826 well as "retry timeout exceeded".
3828 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3829 do this, and (what is worse) MTAs that accept it.
3831 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3832 will now be deprecated.
3834 PH/56 New os.c-cygwin from the Cygwin maintainer.
3836 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3837 formats (previously only available in exim formats via -bpr, -bpru,
3838 and -bpra. Now also available in native and exiqgrep formats)
3840 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3841 with very large, slow to parse queues
3843 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3845 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3847 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3848 responses to authentication challenges, though it was showing the
3849 challenges; (ii) I've removed the CR characters from the debug output for
3852 PH/58 Allow for the insertion of a newline as well as a space when a string
3853 is turned into more than one encoded-word during RFC 2047 encoding. The
3854 Sieve code now uses this.
3856 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3857 data_4xx, lost_connection, tls_required.
3859 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3860 message was being sent as an SMTP response.
3862 PH/61 Add -l and -k options to exicyclog.
3864 PH/62 When verifying, if an address was redirected to one new address, so that
3865 verification continued, and the new address failed or deferred after
3866 having set something in $address_data, the value of $address_data was not
3867 passed back to the ACL. This was different to the case when no
3868 redirection occurred. The value is now passed back in both cases.
3870 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3871 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3872 use login_cap.h, so on its own it isn't the distinguishing feature. The
3873 new name refers directly to the setclassresources() function.
3875 PH/65 Added configuration files for NetBSD3.
3877 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3879 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3880 is preferred over IPv4.
3882 PH/68 The bounce_return_message and bounce_return_body options were not being
3883 honoured for bounces generated during the reception of non-SMTP messages.
3884 In particular, this applied to messages rejected by the ACL. This bug has
3885 been fixed. However, if bounce_return_message is true and bounce_return_
3886 body is false, the headers that are returned for a non-SMTP message
3887 include only those that have been read before the error was detected.
3888 (In the case of an ACL rejection, they have all been read.)
3890 PH/69 The HTML version of the specification is now built in a directory called
3891 spec_html instead of spec.html, because the latter looks like a path with
3892 a MIME-type, and this confuses some software.
3894 PH/70 Catch two compiler warnings in sieve.c.
3896 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3897 function verify_get_ident() calls ip_connect() to connect a socket, but
3898 if the "connect()" function timed out, ip_connect() used to close the
3899 socket. However, verify_get_ident() also closes the socket later, and in
3900 between Exim writes to the log, which may get opened at this point. When
3901 the socket was closed in ip_connect(), the log could get the same file
3902 descriptor number as the socket. This naturally causes chaos. The fix is
3903 not to close the socket in ip_connect(); the socket should be closed by
3904 the function that creates it. There was only one place in the code where
3905 this was missing, in the iplookup router, which I don't think anybody now
3906 uses, but I've fixed it anyway.
3908 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3909 well as to direct DNS lookups. Otherwise the handling of names in host
3910 lists is inconsistent and therefore confusing.
3916 PH/01 Two changes to the default runtime configuration:
3918 (1) Move the checks for relay_from_hosts and authenticated clients from
3919 after to before the (commented out) DNS black list checks.
3921 (2) Add control=submission to the relay_from_hosts and authenticated
3922 clients checks, on the grounds that messages accepted by these
3923 statements are most likely to be submissions.
3925 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3927 (1) Generate an error if the third argument for the ${prvs expansion is
3930 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3933 (3) Reset the variables that are obtained from the first argument of
3934 ${prvscheck and used in the second argument before leaving the code,
3935 because their memory is reclaimed, so using them afterwards may do
3938 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3939 one (it's much easier than Tom thought :-).
3941 (5) Because of (4), we can now allow for the use of $prvscheck_result
3942 inside the third argument.
3944 PH/03 For some reason, the default setting of PATH when running a command from
3945 a pipe transport was just "/usr/bin". I have changed it to
3948 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3949 anything to be listed in the output from -bV.
3951 PH/05 When a filter generated an autoreply, the entire To: header line was
3952 quoted in the delivery log line, like this:
3954 => >A.N.Other <ano@some.domain> <original@ddress> ...
3956 This has been changed so that it extracts the operative address. There
3957 may be more than one such address. If so, they are comma-separated, like
3960 => >ano@some.domain,ona@other.domain <original@ddress> ...
3962 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3963 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3964 not being looked up in the rDNS to get a host name, Exim was showing the
3965 IP address twice in Received: lines, even though the IP addresses were
3966 identical. For example:
3968 Received: from [1.2.3.4] (helo=[1.2.3.4])
3970 However, if the real host name was known, it was omitting the HELO data
3971 if it matched the actual IP address. This has been tidied up so that it
3972 doesn't show the same IP address twice.
3974 PH/07 When both +timestamp and +memory debugging was on, the value given by
3975 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3976 called by the debug printing, thereby overwriting the timestamp buffer.
3977 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3979 PH/08 When the original message was included in an autoreply transport, it
3980 always said "this is a copy of the message, including all the headers",
3981 even if body_only or headers_only was set. It now gives an appropriate
3984 PH/09 Applied a patch from the Sieve maintainer which:
3986 o fixes some comments
3987 o adds the (disabled) notify extension core
3988 o adds some debug output for the result of if/elsif tests
3989 o points to the current vacation draft in the documentation
3990 and documents the missing references header update
3994 o fixes a bug in processing the envelope test (when testing
3995 multiple envelope elements, the last element determined the
3998 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3999 Electronic Mail") by including:
4001 Auto-submitted: auto-generated
4003 in the messages that it generates (bounce messages and others, such as
4004 warnings). In the case of bounce messages for non-SMTP messages, there was
4005 also a typo: it was using "Auto_submitted" (underscore instead of
4006 hyphen). Since every message generated by Exim is necessarily in response
4007 to another message, thes have all been changed to:
4009 Auto-Submitted: auto-replied
4011 in accordance with these statements in the RFC:
4013 The auto-replied keyword:
4015 - SHOULD be used on messages sent in direct response to another
4016 message by an automatic process,
4018 - MUST NOT be used on manually-generated messages,
4020 - MAY be used on Delivery Status Notifications (DSNs) and Message
4021 Disposition Notifications (MDNs),
4023 - MUST NOT be used on messages generated by automatic or periodic
4024 processes, except for messages which are automatic responses to
4027 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4028 to the default Received: header definition.
4030 PH/12 Added log selector acl_warn_skipped (default on).
4032 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4033 variables because (a) they are in the wrong storage pool and (b) even if
4034 they were copied, it wouldn't work properly because of the caching.
4036 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4037 checking when decoding. Apparently there are clients that generate
4038 overlong encoded strings. Why am I not surprised?
4040 PH/15 If the first argument of "${if match_address" was not empty, but did not
4041 contain an "@" character, Exim crashed. Now it writes a panic log message
4042 and treats the condition as false.
4044 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4046 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4047 extension "envelope-auth". The code is finished and in agreement with
4048 other implementations, but there is no documentation so far and in fact,
4049 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4050 not changing the active code.
4052 Print executed "if" and "elsif" statements when debugging is used. This
4053 helps a great deal to understand what a filter does.
4055 Document more things not specified clearly in RFC3028. I had all this
4056 sorted out, when out of a sudden new issues came to my mind. Oops."
4058 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4061 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4062 canonical form (as documented). However, after a host name lookup from
4063 the IP address, check_host() was doing a simple string comparison with
4064 addresses acquired from the DNS when checking that the found name did
4065 have the original IP as one of its addresses. Since any found IPv6
4066 addresses are likely to be in abbreviated form, the comparison could
4067 fail. Luckily, there already exists a function for doing the comparison
4068 by converting both addresses to binary, so now that is used instead of
4069 the text comparison.
4071 PH/20 There was another similar case to PH/19, when a complete host name was
4072 given in a host list; looking up its IP address could give an abbreviated
4073 form, whereas the current host's name might or might not be abbreviated.
4074 The same fix has been applied.
4080 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4081 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4084 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4085 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4087 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4089 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4090 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4091 Also, for "Auto-Submitted", treat anything other than "no" as
4092 non-personal, in accordance with RFC 3834. (Previously it treated
4093 anything starting "auto-" as non-personal.)
4095 TF/01 The control=submission/name=... option had a problem with syntax
4096 errors if the name included a slash character. The /name= option
4097 now slurps the rest of the string, so it can include any characters
4098 but it must come last in the list of options (after /sender_retain
4101 PH/05 Some modifications to the interface to the fake nameserver for the new
4109 TK/01 Added the "success_on_redirect" address verification option. See
4110 NewStuff for rationale and an example.
4112 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4114 PH/02 Patch to exigrep to allow it to work on syslog lines.
4116 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4117 fread() to skip over the body file's header line, because in Cygwin the
4118 header line is locked and is inaccessible.
4120 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4121 co-exist for some time) to make it clear that it is the Exim ID that is
4122 referenced, not the Message-ID: header line.
4124 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4125 string_format() function, because snprintf() does not exist on all
4128 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4129 ${stat: expansion item.
4131 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4132 protocol synchronization error", to keep the pedants happy.
4134 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4135 well as for IRIX systems, when gcc is being used. See the host.c source
4138 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4140 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4143 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4144 transport and to the smtp transport in LMTP mode.
4146 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4148 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4149 run a filter in a subprocess. This could lead to confusion in subsequent
4150 lookups in the parent process. There should also be a search_tidyup() at
4151 the end of the subprocess.
4153 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4154 only if the host matched helo_try_verify_hosts, which caused the
4155 verification to occur when the EHLO/HELO command was issued. The ACL just
4156 tested the remembered result. Now, if a previous verification attempt has
4157 not happened, "verify = helo" does it there and then.
4159 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4161 TK/03 Fix log output including CR from clamd.
4163 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4164 exist provoked a memory error which could cause a segfault.
4166 PH/15 Installed PCRE 6.2
4168 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4170 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4171 of the problem. Specifically, suggested +O2 rather than +O1 for the
4174 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4176 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4177 record, the port was not correctly logged when the outgoing_port log
4178 selector was set (it logged the transort's default port).
4180 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4181 fallback_hosts, and "hosts" in the smtp transport.
4183 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4184 host errors such as "Connection refused".
4186 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4187 authentication with radiusclient 0.4.9:
4189 - Error returned from rc_read_config was caught wrongly
4190 - Username/password not passed on to radius server due to wrong length.
4192 The presumption is that some radiusclient API changes for 4.51/PH/17
4193 were not taken care of correctly. The code is still untested by me (my
4194 Linux distribution still has 0.3.2 of radiusclient), but it was
4195 contributed by a Radius user.
4197 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4198 expanding the "port" option of the smtp transport.
4200 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4201 while reading a MIME header. Thanks to Tom Hughes for a patch.
4203 PH/24 Include config.h inside local_scan.h so that configuration settings are
4206 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4207 that in an ACL for RCPT (for example), you can examine exactly what was
4210 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4211 commands, but it was not correctly comparing the address with the actual
4212 client host address. Thus, it would show the EHLO address in Received:
4213 header lines when this was not necessary.
4215 PH/27 Added the % operator to ${eval:}.
4217 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4218 it should be ignoring failures (because with -C, for example, it has lost
4219 privilege). It wasn't ignoring creation failures other than "already
4222 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4225 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4226 because an input line was too long, either on its own, or by virtue of
4227 too many continuations, the temporary file was not being removed, and the
4228 return code was incorrect.
4230 PH/31 Missing "BOOL" in function definition in filtertest.c.
4232 PH/32 Applied Sieve patches from the maintainer.
4234 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4236 PH/33 Added "verify = not_blind".
4238 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4239 Local/Makefile (with some defaults set). These are used in built scripts
4240 such as exicyclog, but they have never been used in the exim_install
4241 script (though there are many overriding facilities there). I have
4242 arranged that the exim_install script now takes note of these two
4245 PH/35 Installed configuration files for Dragonfly.
4247 PH/36 When a locally submitted message by a trusted user did not contain a
4248 From: header, and the sender address was obtained from -f or from an SMTP
4249 MAIL command, and the trusted user did not use -F to supply a sender
4250 name, $originator_name was incorrectly used when constructing a From:
4251 header. Furthermore, $originator_name was used for submission mode
4252 messages from external hosts without From: headers in a similar way,
4253 which is clearly wrong.
4255 PH/37 Added control=suppress_local_fixups.
4257 PH/38 When log_selector = +received_sender was set, and the addition of the
4258 sender made the log line's construction buffer exactly full, or one byte
4259 less than full, an overflow happened when the terminating "\n" was
4262 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4263 when the result of a list match is failure because a DNS lookup failed.
4265 PH/40 RM_COMMAND is now used in the building process.
4267 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4268 the "build-* directories that it finds.
4270 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4271 address in a domain literal was a prefix of an interface address.
4273 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4274 when verifying a sender address, unless rewrite_headers is false.
4276 PH/44 Wrote a long comment about why errors_to addresses are verified as
4277 recipients, not senders.
4279 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4280 the ratelimit ACL was added.
4282 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4284 PH/46 Added extra information about PostgreSQL errors to the error string.
4286 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4287 avoiding the need to install special zones in a real server. This is
4288 backwards compatible; if it can't find the fake resolver, it drops back.
4289 Thus, both old and new test suites can be run.
4291 TF/02 Added util/ratelimit.pl
4293 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4294 clock is set back in time.
4296 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4297 Candler <B.Candler@pobox.com>.
4299 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4300 OK for addresses that are the result of redirections.
4302 PH/48 A number of further additions for the benefit of the new test suite,
4303 including a fake gethostbyname() that interfaces to the fake DNS resolver
4306 TF/06 The fix for widen_domains has also been applied to qualify_single and
4307 search_parents which are the other dnslookup options that can cause
4310 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4313 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4315 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4316 service" if it encountered an MX record with an empty target hostname.
4317 The message is now "an MX or SRV record indicated no SMTP service".
4319 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4320 if the DNS of the sending site is misconfigured. This is quite a
4321 common situation. This change restores the behaviour of treating a
4322 helo verification defer as a failure.
4324 PH/51 If self=fail was set on a router, the bounce message did not include the
4325 actual error message.
4331 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4333 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4334 command itself ended in error, the underlying message about the transport
4335 filter timeout was being overwritten with the pipe command error. Now the
4336 underlying error message should be appended to the second error message.
4338 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4340 PH/02 Reduce the amount of output that "make" produces by default. Full output
4341 can still be requested.
4343 PH/03 The warning log line about a condition test deferring for a "warn" verb
4344 was being output only once per connection, rather than after each
4345 occurrence (because it was using the same function as for successful
4346 "warn" verbs). This seems wrong, so I have changed it.
4348 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4349 it should not have, which might have caused a crash in the right
4350 circumstances, but probably never did.
4352 PH/04 Installed a modified version of Tony Finch's patch to make submission
4353 mode fix the return path as well as the Sender: header line, and to
4354 add a /name= option so that you can make the user's friendly name appear
4357 TF/03 Added the control = fakedefer ACL modifier.
4359 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4360 Mark Lowes for thorough testing.
4362 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4364 TK/03 Merged latest SRS patch from Miles Wilton.
4366 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4367 with the definition in sysexits.h (which is #included earlier).
4368 Fortunately, Exim does not actually use EX_OK. The code used to try to
4369 preserve the sysexits.h value, by assuming that macro definitions were
4370 scanned for macro replacements. I have been disabused of this notion,
4371 so now the code just undefines EX_OK before #including unistd.h.
4373 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4374 in the smtp transport. When a block could not be written in a single
4375 write() function, the timeout was being re-applied to each part-write.
4376 This seems wrong - if the receiver was accepting one byte at a time it
4377 would take for ever. The timeout is now adjusted when this happens. It
4378 doesn't have to be particularly precise.
4380 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4381 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4383 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4384 without a domain if the check to <postmaster@domain> fails.
4386 SC/01 Eximstats: added -xls and the ability to specify output files
4387 (patch written by Frank Heydlauf).
4389 SC/02 Eximstats: use FileHandles for outputting results.
4391 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4393 SC/04 Eximstats: fixed display of large numbers with -nvr option
4395 SC/05 Eximstats: fixed merging of reports with empty tables.
4397 SC/06 Eximstats: added the -include_original_destination flag
4399 SC/07 Eximstats: removed tabs and trailing whitespace.
4401 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4403 TK/06 MBOX spool code: Add real "From " MBOX separator line
4404 so the .eml file is really in mbox format (even though
4405 most programs do not really care). Patch from Alex Miller.
4407 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4408 The latter is generated from $received_to and is only set if the
4409 message has one envelope recipient. SA can use these headers,
4410 obviously out-of-the-box. Patch from Alex Miller.
4412 PH/08 The ${def test on a variable was returning false if the variable's
4413 value was "0", contrary to what the specification has always said!
4414 The result should be true unless the variable is empty.
4416 PH/09 The syntax error of a character other than { following "${if
4417 def:variable_name" (after optional whitespace) was not being diagnosed.
4418 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4419 accidental colon was present, for example, could give incorrect results.
4421 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4422 result is used (not including appendfile, where other changes are about
4425 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4426 This involved changing a lot of size variables from int to off_t. It
4427 should work with maildirs and everything.
4429 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4430 spamd dying while we are connected to it.
4432 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4435 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4436 function for BDB 4.3.
4438 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4440 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4441 the sources for PCRE have been re-arranged and more files are now
4444 PH/15 The code I had for printing potentially long long variables in PH/11
4445 above was not the best (it lost precision). The length of off_t variables
4446 is now inspected at build time, and an appropriate printing format (%ld
4447 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4448 to be "long long int" or "long int". This is needed for the internal
4449 formatting function string_vformat().
4451 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4452 the configuration file to be ":syslog", then the script "guesses" where
4453 the logs files are, rather than using the compiled in default. In our
4454 case the guess is not the same as the compiled default, so the script
4455 suddenly stopped working when I started to use syslog. The patch checks
4456 to see if log_file_path is "". If so, it attempts to read it from exim
4457 with no configuration file to get the compiled in version, before it
4458 falls back to the previous guessing code."
4460 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4461 implementing BATV in an Exim configuration. See NewStuff for the gory
4464 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4465 Makefile that are specific to HP-UX.
4467 PH/18 If the "use_postmaster" option was set for a recipient callout together
4468 with the "random" option, the postmaster address was used as the MAIL
4469 FROM address for the random test, but not for the subsequent recipient
4470 test. It is now used for both.
4472 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4473 patch removes a few documentation additions to RFC 3028, because the
4474 latest draft now contains them. It adds the new en;ascii-case comparator
4475 and a new error check for 8bit text in MIME parts. Comparator and
4476 require names are now matched exactly. I enabled the subaddress
4477 extension, but it is not well tested yet (read: it works for me)."
4479 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4480 rework some of the code of TK/09 above to avoid the hardwired use of
4481 "%lld" and "long long". Replaced the call to snprintf() with a call to
4484 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4485 records point to non-existent hosts", "retry timeout exceeded", and
4486 "retry time not reached for any host after a long failure period".
4488 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4489 experimental DomainKeys support:
4491 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4492 (2) On an error such as an illegally used "control", the wrong name for
4493 the control was given.
4495 These problems did NOT occur unless DomainKeys support was compiled.
4497 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4499 PH/24 Added ${if match_ip condition.
4501 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4502 databases so that it will be absolutely obvious if a crash occurs in the
4503 DB library. This is a regular occurrence (often caused by mis-matched
4506 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4507 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4508 picked up on a user's system that detects such things. There doesn't seem
4509 to be a gcc warning option for this - only an attribute that has to be
4510 put on the function's prototype. It seems that in Fedora Core 4 they have
4511 set this on a number of new functions. No doubt there will be more in due
4514 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4515 specify a transport. However, if an address that was verified by such a
4516 router was the subject of a callout, Exim crashed because it tried to
4517 read the rcpt_include_affixes from the non-existent transport. Now it
4518 just assumes that the setting of that option is false. This bug was
4519 introduced by 4.51/PH/31.
4521 PH/28 Changed -d+all to exclude +memory, because that information is very
4522 rarely of interest, but it makes the output a lot bigger. People tend to
4523 do -d+all out of habit.
4525 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4526 code in os-type was giving problems when libc.so lives in lib64, like on
4529 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4530 aren't the modern standard, and it seems that some systems' include files
4531 don't always have them. Exim was already checking for some of the newer
4532 ones like T_AAAA, and defining it itself. I've added checks for all the
4533 record types that Exim uses.
4535 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4536 not automatically generating a new one, as it is supposed to. This
4537 prevented TLS from working. If the file did exist, but contained invalid
4538 data, a new version was generated, as expected. It was only the case of a
4539 non-existent file that was broken.
4541 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4542 with a change in libdomainkeys > 0.64.
4544 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4545 from DNS. If the selector record carries the flag, it now has
4546 precedence over the domain-wide flag.
4548 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4550 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4551 the use of a port name that isn't defined in /etc/services) occurred, the
4552 message was deferred as in a normal delivery, and thus remained on the
4553 spool, instead of being failed because of the mua_wrapper setting. This
4554 is now fixed, and I tidied up some of the mua_wrapper messages at the
4557 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4558 the messages in a hash of arrays rather than using individual hashes.
4559 This is a bit cleaner and results in dramatic memory savings, albeit
4560 at a slight CPU cost.
4562 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4563 as requested by Marc Sherman.
4565 SC/10 Eximstats: added histograms for user specified patterns as requested
4568 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4570 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4571 fopen() in the content-scanning modules that did not already have it.
4577 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4578 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4580 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4582 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4584 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4585 to test Sieve filters that use "vacation".
4587 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4588 that changes the way the GnuTLS parameters are stored in the cache file.
4589 The new format can be generated externally. For backward compatibility,
4590 if the data in the cache doesn't make sense, Exim assumes it has read an
4591 old-format file, and it generates new data and writes a new file. This
4592 means that you can't go back to an older release without removing the
4595 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4596 work if there are any delivery delays because "one_time" forces the
4597 parent to be marked "delivered", so its unseen clone is never tried
4598 again. For this reason, Exim now forbids the simultaneous setting of
4601 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4602 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4603 entry if you want to know the details. The fix, however, neglected to
4604 consider the case when local delivery batching is involved. The test for
4605 "previously delivered" was not happening when checking to see if an
4606 address could be batched with a previous (undelivered) one; under
4607 certain circumstances this could lead to multiple deliveries to the same
4610 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4611 in its include files, and this causes problems building Exim.
4613 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4614 header_syntax) but Exim was just ignoring anything given after a slash.
4615 In particular, this caused confusion with an attempt to use "verify =
4616 reverse_host_lookup/defer_ok". An error is now given when options are
4617 supplied for verify items that do not have them. (Maybe reverse_host_
4618 lookup should have a defer_ok option, but that's a different point.)
4620 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4621 defined by RFC 821) to 2048, because there were problems with some AUTH
4622 commands, and RFC 1869 says the size should be increased for extended
4623 SMTP commands that take arguments.
4625 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4628 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4629 "unknown" error; now it says that the functionality isn't in the binary.
4631 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4632 an address' error message when a string expansion fails (syntax or
4633 whatever). Otherwise the password may appear in the log. Following change
4634 PH/42 below, there is no longer a chance of it appearing in a bounce
4637 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4639 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4640 of its addresses was ever tried. (Bugzilla bug #2.)
4642 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4643 the result incorrectly in the debug output. (It correctly added a newline
4644 to what was transported.)
4646 TF/01 Added $received_time.
4648 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4649 commented out examples of how to interface to a virus scanner and to
4650 SpamAssassin. Also added commented examples of av_scanner and
4651 spamd_address settings.
4653 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4654 and controls are allowed in which ACLs. There were a couple of minor
4655 errors. Some of the entries in the conditions table (which is a table of
4656 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4657 negation of where the condition IS allowed.
4659 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4661 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4662 header file does not have a version number, so I've had to invent a new
4663 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4664 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4665 radiusclient), but it was contributed by a Radius user.
4667 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4668 files or directories, for OpenSSL.
4670 PH/19 When an Exim process that is running as root has to create an Exim log
4671 file, it does so in a subprocess that runs as exim:exim so as to get the
4672 ownership right at creation (otherwise, other Exim processes might see
4673 the file with the wrong ownership). There was no test for failure of this
4674 fork() call, which would lead to the process getting stuck as it waited
4675 for a non-existent subprocess. Forks do occasionally fail when resources
4676 run out. I reviewed all the other calls to fork(); they all seem to check
4679 PH/20 When checking for unexpected SMTP input at connect time (before writing
4680 the banner), Exim was not dealing correctly with a non-positive return
4681 from the read() function. If the client had disconnected by this time,
4682 the result was a log entry for a synchronization error with an empty
4683 string after "input=" when read() returned zero. If read() returned -1
4684 (an event I could not check), uninitialized data bytes were printed.
4685 There were reports of junk text (parts of files, etc) appearing after
4688 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4690 PH/22 Added support for macro redefinition, and (re)definition in between
4691 driver and ACL definitions.
4693 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4694 forgetting to use the resulting value; it was using the unexpanded value.
4696 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4697 hadn't been configured. The fix is from Juergen Kreileder, who
4698 understands it better than I do:
4700 "Here's what I see happening with three configured cyrus_sasl
4701 authenticators configured (plain, login, cram-md5):
4703 On startup auth_cyrus_sasl_init() gets called for each of these.
4704 This means three calls to sasl_listmech() without a specified mech_list.
4705 => SASL tests which mechs of all available mechs actually work
4706 => three warnings about OTP not working
4707 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4709 With the patch, sasl_listmech() also gets called three times. But now
4710 SASL's mech_list option is set to the server_mech specified in the the
4711 authenticator. Or in other words, the answer from sasl_listmech()
4712 gets limited to just the mech you're testing for (which is different
4714 => the return list contains just 'plain' or 'login', 'cram-md5' or
4715 nothing depending on the value of ob->server_mech.
4717 I've just tested the patch: Authentication still works fine,
4718 unavailable mechs specified in the exim configuration are still
4719 caught, and the auth.log warnings about OTP are gone."
4721 PH/25 When debugging is enabled, the contents of the command line are added
4722 to the debugging output, even when log_selector=+arguments is not
4725 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4726 answer is "GNU", and only if the return is "GNU/something" is the answer
4729 PH/27 $acl_verify_message is now set immediately after the failure of a
4730 verification in an ACL, and so is available in subsequent modifiers. In
4731 particular, the message can be preserved by coding like this:
4733 warn !verify = sender
4734 set acl_m0 = $acl_verify_message
4736 Previously, $acl_verify_message was set only while expanding "message"
4737 and "log_message" when a very denied access.
4739 PH/28 Modified OS/os.c-Linux with
4741 -#ifndef OS_LOAD_AVERAGE
4742 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4744 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4745 nomenclature these days.)
4747 PH/29 Installed patch from the Sieve maintainer that adds the options
4748 sieve_useraddress and sieve_subaddress to the redirect router.
4750 PH/30 In these circumstances:
4751 . Two addresses routed to the same list of hosts;
4752 . First host does not offer TLS;
4753 . First host accepts first address;
4754 . First host gives temporary error to second address;
4755 . Second host offers TLS and a TLS session is established;
4756 . Second host accepts second address.
4757 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4758 and peerdn, if requested) that were in fact used only for the second
4761 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4762 attention to any local part prefix or suffix that was matched by the
4763 router that accepted the address. It now behaves in the same way as it
4764 does for delivery: the affixes are removed from the local part unless
4765 rcpt_include_affixes is set on the transport.
4767 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4768 timeout during the DATA phase of an incoming message.
4770 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4771 applied a patch sanctioned by the Sieve maintainer.
4773 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4774 the uid or gid is negative. A case of a negative gid caused this to be
4775 noticed. The fix allows for either to be negative.
4777 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4778 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4781 PH/36 The change PH/12 above was broken. Fixed it.
4783 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4784 the grounds that routing the same address twice would always produce the
4785 same answer. This might have been true once, but it is certainly no
4786 longer true now. Routing a child address may depend on the previous
4787 routing that produced that child. Some complicated redirection strategies
4788 went wrong when messages had multiple recipients, and made Exim's
4789 behaviour dependent on the order in which the addresses were given.
4791 I have moved the duplicate checking until after the routing is complete.
4792 Exim scans the addresses that are assigned to local and remote
4793 transports, and removes any duplicates. This means that more work will be
4794 done, as duplicates will always all be routed, but duplicates are
4795 presumably rare, so I don't expect this is of any significance.
4797 For deliveries to pipes, files, and autoreplies, the duplicate checking
4798 still happens during the routing process, since they are not going to be
4801 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4802 It corrects a timeout issue with spamd. This is Ian's comment: "The
4803 background is that sometimes spamd either never reads data from a
4804 connection it has accepted, or it never writes response data. The exiscan
4805 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4806 blindly assumes that writes won't block so it may never time out."
4808 PH/39 Allow G after quota size as well as K and M.
4810 PH/40 The value set for $authenticated_id in an authenticator may not contain
4811 binary zeroes or newlines because the value is written to log lines and
4812 to spool files. There was no check on this. Now the value is run through
4813 the string_printing() function so that such characters are converted to
4814 printable escape sequences.
4816 PH/41 $message_linecount is a new variable that contains the total number of
4817 lines in the message. Compare $body_linecount, which is the count for the
4820 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4821 bounce and delay warning messages, except in certain special cases, which
4824 (a) An SMTP error message from a remote host;
4825 (b) A message specified in a :fail: redirection;
4826 (c) A message specified in a "fail" command in a system filter;
4827 (d) A message specified in a FAIL return from the queryprogram router;
4828 (e) A message specified by the cannot_route_message router option.
4830 In these cases only, Exim does include the error details in bounce and
4831 warning messages. There are also a few cases where bland messages such
4832 as "unrouteable address" or "local delivery error" are given.
4834 PH/43 $value is now also set for the "else" part of a ${run expansion.
4836 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4837 being worked on, but at least Exim now implements the latest version to
4840 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4841 process to complete was treated as a delivery failure, a timeout while
4842 writing the message to the pipe was logged, but erroneously treated as a
4843 successful delivery. Such timeouts include transport filter timeouts. For
4844 consistency with the overall process timeout, these timeouts are now
4845 treated as errors, giving rise to delivery failures by default. However,
4846 there is now a new Boolean option for the pipe transport called
4847 timeout_defer, which, if set TRUE, converts the failures into defers for
4848 both kinds of timeout. A transport filter timeout is now identified in
4851 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4852 systems where "make" and "gmake" are different, calling "gmake" at top
4853 level broke things. I've arranged for the value of $(MAKE) to be passed
4854 from the Makefile to this script so that it can call the same version of
4858 A note about Exim versions 4.44 and 4.50
4859 ----------------------------------------
4861 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4862 changes of various kinds. As a consequence, a big documentation update was
4863 needed. This delayed the release for rather longer than seemed good, especially
4864 in the light of a couple of (minor) security issues. Therefore, the changes
4865 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4866 release. So 4.44 and 4.50 are in effect two different branches that both start
4869 I have left the 4.50 change log unchanged; it contains all the changes since
4870 4.43. The change log for 4.44 is below; many of its items are identical to
4871 those for 4.50. This seems to be the most sensible way to preserve the
4872 historical information.
4878 1. Minor wording change to the doc/README.SIEVE file.
4880 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4881 computation of the current number of files was incorrect.
4883 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4884 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4887 4. Give more explanation in the error message when the command for a transport
4888 filter fails to execute.
4890 5. There are several places where Exim runs a non-Exim command in a
4891 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4892 was being done only for the command run by the queryprogram router. It is
4893 now done for all such subprocesses. The other cases are: ${run, transport
4894 filters, and the commands run by the lmtp and pipe transports.
4896 6. Added CONFIGURE_GROUP build-time option.
4898 7. Some older OS have a limit of 256 on the maximum number of file
4899 descriptors. Exim was using setrlimit() to set 1000 as a large value
4900 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4901 systems. I've change it so that if it can't get 1000, it tries for 256.
4903 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4904 was an oversight, and furthermore, ever since the addition of extra
4905 controls (e.g. 4.43/32), the checks on when to allow different forms of
4906 "control" were broken. There should now be diagnostics for all cases when a
4907 control that does not make sense is encountered.
4909 9. Added the /retain_sender option to "control=submission".
4911 10. $recipients is now available in the predata ACL (oversight).
4913 11. Tidy the search cache before the fork to do a delivery from a message
4914 received from the command line. Otherwise the child will trigger a lookup
4915 failure and thereby defer the delivery if it tries to use (for example) a
4916 cached ldap connection that the parent has called unbind on.
4918 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4919 of $address_data from the recipient verification was clobbered by the
4920 sender verification.
4922 13. The value of address_data from a sender verification is now available in
4923 $sender_address_data in subsequent conditions in the ACL statement.
4925 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4927 15. Added a new option "connect=<time>" to callout options, to set a different
4930 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4931 was its contents. (It was OK if the option was not defined at all.)
4933 17. A "Completed" log line is now written for messages that are removed from
4934 the spool by the -Mrm option.
4936 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4937 information about exactly what failed.
4939 19. Added -dd to debug only the daemon process.
4941 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4942 handles timeouts, both on the server side and network timeouts. Renamed the
4943 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4945 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4946 It is now set to "smtps".
4948 22. $host_address is now set to the target address during the checking of
4949 ignore_target_hosts.
4951 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4952 being passed; this would have caused $sender_host_name to have been used if
4953 matching the list had actually called for a host name (not very likely,
4954 since this list is usually IP addresses). A host name is now passed as
4957 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4958 code that specifies a non-restarting handler (typically sigaction() in
4959 modern systems) in an attempt to fix a rare and obscure crash bug.
4961 25. Narrowed the window for a race in the daemon that could cause it to ignore
4962 SIGCHLD signals. This is not a major problem, because they are used only to
4963 wake it up if nothing else does.
4965 26. A malformed maildirsize file could cause Exim to calculate negative values
4966 for the mailbox size or file count. Odd effects could occur as a result.
4967 The maildirsize information is now recalculated if the size or filecount
4970 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4971 support for a long time. Removed HAVE_SYS_VFS_H.
4973 28. Installed the latest version of exipick from John Jetmore.
4975 29. In an address list, if the pattern was not a regular expression, an empty
4976 subject address (from a bounce message) matched only if the pattern was an
4977 empty string. Non-empty patterns were not even tested. This was the wrong
4978 because it is perfectly reasonable to use an empty address as part of a
4979 database query. An empty address is now tested by patterns that are
4980 lookups. However, all the other forms of pattern expect the subject to
4981 contain a local part and a domain, and therefore, for them, an empty
4982 address still always fails if the pattern is not itself empty.
4984 30. Exim went into a mad DNS loop when attempting to do a callout where the
4985 host was specified on an smtp transport, and looking it up yielded more
4986 than one IP address.
4988 31. Re-factored the code for checking spool and log partition space into a
4989 function that finds that data and another that does the check. The former
4990 is then used to implement four new variables: $spool_space, $log_space,
4991 $spool_inodes, and $log_inodes.
4993 32. The RFC2047 encoding function was originally intended for short strings
4994 such as real names; it was not keeping to the 75-character limit for
4995 encoded words that the RFC imposes. It now respects the limit, and
4996 generates multiple encoded words if necessary. To be on the safe side, I
4997 have increased the buffer size for the ${rfc2047: expansion operator from
5000 33. It is now permitted to omit both strings after an "if" condition; if the
5001 condition is true, the result is "true". As before, when the second string
5002 is omitted, a false condition yields an empty string. This makes it less
5003 cumbersome to write custom ACL and router conditions.
5005 34. Failure to deliver a bounce message always caused it to be frozen, even if
5006 there was an errors_to setting on the router. The errors_to setting is now
5009 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5010 canonical form (fully expanded) before being placed in
5011 $sender_host_address.
5013 36. The table in the code that translates DNS record types into text (T_A to
5014 "A" for instance) was missing entries for NS and CNAME. It is just possible
5015 that this could have caused confusion if both these types were looked up
5016 for the same domain, because the text type is used as part of Exim's
5017 per-process caching. But the chance of anyone hitting this buglet seems
5020 37. The dnsdb lookup has been extended in a number of ways.
5022 (1) There is a new type, "zns", which walks up the domain tree until it
5023 finds some nameserver records. It should be used with care.
5025 (2) There is a new type, "mxh", which is like "mx" except that it returns
5026 just the host names, not the priorities.
5028 (3) It is now possible to give a list of domains (or IP addresses) to be
5029 looked up. The behaviour when one of the lookups defers can be
5030 controlled by a keyword.
5032 (4) It is now possible to specify the separator character for use when
5033 multiple records are returned.
5035 38. The dnslists ACL condition has been extended: it is now possible to supply
5036 a list of IP addresses and/or domains to be looked up in a particular DNS
5039 39. Added log_selector=+queue_time_overall.
5041 40. When running the queue in the test harness, wait just a tad after forking a
5042 delivery process, to get repeatability of debugging output.
5044 41. Include certificate and key file names in error message when GnuTLS fails
5045 to set them up, because the GnuTLS error message doesn't include the name
5046 of the failing file when there is a problem reading it.
5048 42. Allow both -bf and -bF in the same test run.
5050 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5052 44. The "Exiscan patch" is now merged into the mainline Exim source.
5054 45. Sometimes the final signoff response after QUIT could fail to get
5055 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5056 < 0 before doing a fflush(). This bug looks as though it goes back to the
5057 introduction of TLS in release 3.20, but "sometimes" must have been rare
5058 because the tests only now provoked it.
5060 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5061 (this can affect the format of dates).
5063 47. exim_tidydb, when checking for the continued existence of a message for
5064 which it has found a message-specific retry record, was not finding
5065 messages that were in split spool directories. Consequently, it was
5066 deleting retry records that should have stayed in existence.
5068 48. Steve fixed some bugs in eximstats.
5070 49. The SPA authentication driver was not abandoning authentication and moving
5071 on to the next authenticator when an expansion was forced to fail,
5072 contradicting the general specification for all authenticators. Instead it
5073 was generating a temporary error. It now behaves as specified.
5075 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5076 (the order specifies the preference for clients). The order is now AES256,
5077 AES128, 3DES, ARCFOUR128.
5079 51. Small patch to Sieve code - explicitly set From: when generating an
5082 52. Exim crashed if a remote delivery caused a very long error message to be
5083 recorded - for instance if somebody sent an entire SpamAssassin report back
5084 as a large number of 550 error lines. This bug was coincidentally fixed by
5085 increasing the size of one of Exim's internal buffers (big_buffer) that
5086 happened as part of the Exiscan merge. However, to be on the safe side, I
5087 have made the code more robust (and fixed the comments that describe what
5090 53. Now that there can be additional text after "Completed" in log lines (if
5091 the queue_time_overall log selector is set), a one-byte patch to exigrep
5092 was needed to allow it to recognize "Completed" as not the last thing in
5095 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5096 patch that reportedly fixes this has been added. I am not expert enough to
5097 create a test for it. This is what the patch creator wrote:
5099 "I found a little strange behaviour of ldap code when working with
5100 Windows 2003 AD Domain, where users was placed in more than one
5101 Organization Units. When I tried to give exim partial DN, the exit code
5102 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5103 But simultaneously result of request was absolutely normal ldap result,
5104 so I produce this patch..."
5106 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5107 REFERENCE, so I have modified the code to exclude the patch when that macro
5110 55. Some experimental protocols are using DNS PTR records for new purposes. The
5111 keys for these records are domain names, not reversed IP addresses. The
5112 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5113 leaves it alone. Component reversal etc. now happens only for IP addresses.
5116 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5118 57. Double the size of the debug message buffer (to 2048) so that more of very
5119 long debug lines gets shown.
5121 58. The exicyclog utility now does better if the number of log files to keep
5122 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5124 59. Two changes related to the smtp_active_hostname option:
5126 (1) $smtp_active_hostname is now available as a variable.
5127 (2) The default for smtp_banner uses $smtp_active_hostname instead
5128 of $primary_hostname.
5130 60. The host_aton() function is supposed to be passed a string that is known
5131 to be a valid IP address. However, in the case of IPv6 addresses, it was
5132 not checking this. This is a hostage to fortune. Exim now panics and dies
5133 if the condition is not met. A case was found where this could be provoked
5134 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5135 components; fortuitously, this particular loophole had already been fixed
5136 by change 4.50/55 above.
5138 If there are any other similar loopholes, the new check in host_aton()
5139 itself should stop them being exploited. The report I received stated that
5140 data on the command line could provoke the exploit when Exim was running as
5141 exim, but did not say which command line option was involved. All I could
5142 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5143 running as the user.
5146 61. There was a buffer overflow vulnerability in the SPA authentication code
5147 (which came originally from the Samba project). I have added a test to the
5148 spa_base64_to_bits() function which I hope fixes it.
5151 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5152 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5154 63. The daemon start-up calls getloadavg() while still root for those OS that
5155 need the first call to be done as root, but it missed one case: when
5156 deliver_queue_load_max is set with deliver_drop_privilege. This is
5157 necessary for the benefit of the queue runner, because there is no re-exec
5158 when deliver_drop_privilege is set.
5160 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5161 This has been fixed.
5163 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5164 was in use, was not putting the data itself into the right store pool;
5165 consequently, it could be overwritten for a subsequent message in the same
5166 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5169 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5171 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5172 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5173 the function were treating the return as a boolean value, which happened to
5174 work because 0=false and not-0=true, but is not correct code.
5176 68. The host_aton() function was not handling scoped IPv6 addresses (those
5177 with, for example, "%eth0" on the end) correctly.
5179 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5180 negated items (that is, ~something) in unsigned ints. Some compilers
5181 apparently mutter when there is no cast.
5183 70. If an address verification called from an ACL failed, and did not produce a
5184 user-specific message (i.e. there was only a "system" message), nothing was
5185 put in $acl_verify_message. In this situation, it now puts the system
5188 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5189 session; change 4.31/43 added the unwanted input to the log line - except
5190 that it did not do this in the start of session case. It now does.
5192 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5193 This is wrong and can cause the other end to generate a synchronization
5194 error if it is another Exim or anything else that does the synchronization
5195 check. A QUIT command is no longer sent after a timeout.
5197 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5198 during host lookups.
5200 74. The defer_ok option of callout verification was not working if it was used
5201 when verifying addresses in header lines, that is, for this case:
5203 verify = header_sender/callout=defer_ok
5205 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5206 those file descriptors could be used for SMTP connections. If anything
5207 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5208 could be sent to the SMTP client, causing chaos. The daemon now opens
5209 stdin, stdout, and stderr to /dev/null when it puts itself into the
5212 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5213 log by default. The user can override this with suitable Perl magic.
5215 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5216 the log message when discard triggers, was not working for the DATA ACL or
5217 for the non-SMTP ACL.
5219 78. Error message wording change in sieve.c.
5221 79. If smtp_accept_max_per_host was set, the number of connections could be
5222 restricted to fewer than expected, because the daemon was trying to set up
5223 a new connection before checking whether the processes handling previous
5224 connections had finished. The check for completed processes is now done
5225 earlier. On busy systems, this bug wouldn't be noticed because something
5226 else would have woken the daemon, and it would have reaped the completed
5229 80. If a message was submitted locally by a user whose login name contained one
5230 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5231 It caused a spool format error. I have fixed the spool reading code. A
5232 related problem was that the "from" clause in the Received: line became
5233 illegal because of the space(s). It is now covered by ${quote_local_part.
5235 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5238 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5240 83. Patch from Sieve maintainer to fix unterminated string problem in
5241 "vacation" handling.
5243 84. Some minor changes to the Linux configuration files to help with other
5244 OS variants using glibc.
5246 85. One more patch for Sieve to update vacation handling to latest spec.
5249 ----------------------------------------------------
5250 See the note above about the 4.44 and 4.50 releases.
5251 ----------------------------------------------------
5257 1. Change 4.43/35 introduced a bug that caused file counts to be
5258 incorrectly computed when quota_filecount was set in an appendfile
5261 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5262 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5265 3. Give more explanation in the error message when the command for a transport
5266 filter fails to execute.
5268 4. There are several places where Exim runs a non-Exim command in a
5269 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5270 was being done only for the command run by the queryprogram router. It is
5271 now done for all such subprocesses. The other cases are: ${run, transport
5272 filters, and the commands run by the lmtp and pipe transports.
5274 5. Some older OS have a limit of 256 on the maximum number of file
5275 descriptors. Exim was using setrlimit() to set 1000 as a large value
5276 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5277 systems. I've change it so that if it can't get 1000, it tries for 256.
5279 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5280 was an oversight, and furthermore, ever since the addition of extra
5281 controls (e.g. 4.43/32), the checks on when to allow different forms of
5282 "control" were broken. There should now be diagnostics for all cases when a
5283 control that does not make sense is encountered.
5285 7. $recipients is now available in the predata ACL (oversight).
5287 8. Tidy the search cache before the fork to do a delivery from a message
5288 received from the command line. Otherwise the child will trigger a lookup
5289 failure and thereby defer the delivery if it tries to use (for example) a
5290 cached ldap connection that the parent has called unbind on.
5292 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5293 of $address_data from the recipient verification was clobbered by the
5294 sender verification.
5296 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5297 was its contents. (It was OK if the option was not defined at all.)
5299 11. A "Completed" log line is now written for messages that are removed from
5300 the spool by the -Mrm option.
5302 12. $host_address is now set to the target address during the checking of
5303 ignore_target_hosts.
5305 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5306 being passed; this would have caused $sender_host_name to have been used if
5307 matching the list had actually called for a host name (not very likely,
5308 since this list is usually IP addresses). A host name is now passed as
5311 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5312 code that specifies a non-restarting handler (typically sigaction() in
5313 modern systems) in an attempt to fix a rare and obscure crash bug.
5315 15. Narrowed the window for a race in the daemon that could cause it to ignore
5316 SIGCHLD signals. This is not a major problem, because they are used only to
5317 wake it up if nothing else does.
5319 16. A malformed maildirsize file could cause Exim to calculate negative values
5320 for the mailbox size or file count. Odd effects could occur as a result.
5321 The maildirsize information is now recalculated if the size or filecount
5324 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5325 support for a long time. Removed HAVE_SYS_VFS_H.
5327 18. Updated exipick to current release from John Jetmore.
5329 19. Allow an empty sender to be matched against a lookup in an address list.
5330 Previously the only cases considered were a regular expression, or an
5333 20. Exim went into a mad DNS lookup loop when doing a callout where the
5334 host was specified on the transport, if the DNS lookup yielded more than
5337 21. The RFC2047 encoding function was originally intended for short strings
5338 such as real names; it was not keeping to the 75-character limit for
5339 encoded words that the RFC imposes. It now respects the limit, and
5340 generates multiple encoded words if necessary. To be on the safe side, I
5341 have increased the buffer size for the ${rfc2047: expansion operator from
5344 22. Failure to deliver a bounce message always caused it to be frozen, even if
5345 there was an errors_to setting on the router. The errors_to setting is now
5348 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5349 canonical form (fully expanded) before being placed in
5350 $sender_host_address.
5352 24. Updated eximstats to version 1.33
5354 25. Include certificate and key file names in error message when GnuTLS fails
5355 to set them up, because the GnuTLS error message doesn't include the name
5356 of the failing file when there is a problem reading it.
5358 26. Expand error message when OpenSSL has problems setting up cert/key files.
5361 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5362 (this can affect the format of dates).
5364 28. exim_tidydb, when checking for the continued existence of a message for
5365 which it has found a message-specific retry record, was not finding
5366 messages that were in split spool directories. Consequently, it was
5367 deleting retry records that should have stayed in existence.
5369 29. eximstats updated to version 1.35
5370 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5371 1.35 - bugfix such that pie charts by volume are generated correctly
5373 30. The SPA authentication driver was not abandoning authentication and moving
5374 on to the next authenticator when an expansion was forced to fail,
5375 contradicting the general specification for all authenticators. Instead it
5376 was generating a temporary error. It now behaves as specified.
5378 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5379 (the order specifies the preference for clients). The order is now AES256,
5380 AES128, 3DES, ARCFOUR128.
5382 31. Small patch to Sieve code - explicitly set From: when generating an
5385 32. Exim crashed if a remote delivery caused a very long error message to be
5386 recorded - for instance if somebody sent an entire SpamAssassin report back
5387 as a large number of 550 error lines. This bug was coincidentally fixed by
5388 increasing the size of one of Exim's internal buffers (big_buffer) that
5389 happened as part of the Exiscan merge. However, to be on the safe side, I
5390 have made the code more robust (and fixed the comments that describe what
5393 33. Some experimental protocols are using DNS PTR records for new purposes. The
5394 keys for these records are domain names, not reversed IP addresses. The
5395 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5396 leaves it alone. Component reversal etc. now happens only for IP addresses.
5399 34. The host_aton() function is supposed to be passed a string that is known
5400 to be a valid IP address. However, in the case of IPv6 addresses, it was
5401 not checking this. This is a hostage to fortune. Exim now panics and dies
5402 if the condition is not met. A case was found where this could be provoked
5403 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5404 components; fortuitously, this particular loophole had already been fixed
5405 by change 4.50/55 or 4.44/33 above.
5407 If there are any other similar loopholes, the new check in host_aton()
5408 itself should stop them being exploited. The report I received stated that
5409 data on the command line could provoke the exploit when Exim was running as
5410 exim, but did not say which command line option was involved. All I could
5411 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5412 running as the user.
5415 35. There was a buffer overflow vulnerability in the SPA authentication code
5416 (which came originally from the Samba project). I have added a test to the
5417 spa_base64_to_bits() function which I hope fixes it.
5420 36. The daemon start-up calls getloadavg() while still root for those OS that
5421 need the first call to be done as root, but it missed one case: when
5422 deliver_queue_load_max is set with deliver_drop_privilege. This is
5423 necessary for the benefit of the queue runner, because there is no re-exec
5424 when deliver_drop_privilege is set.
5426 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5427 was in use, was not putting the data itself into the right store pool;
5428 consequently, it could be overwritten for a subsequent message in the same
5429 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5432 38. Sometimes the final signoff response after QUIT could fail to get
5433 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5434 < 0 before doing a fflush(). This bug looks as though it goes back to the
5435 introduction of TLS in release 3.20, but "sometimes" must have been rare
5436 because the tests only now provoked it.
5442 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5443 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5444 Now it returns an integer. A number of calls were still expecting a T/F
5445 return. Fortuitously, in all cases, the tests worked in OK situations,
5446 which is the norm. However, things would have gone wrong on any write
5447 failures on the smtp file descriptor. This function is used when sending
5448 messages over SMTP and also when doing verify callouts.
5450 2. When Exim is called to do synchronous delivery of a locally submitted
5451 message (the -odf or -odi options), it no longer closes stderr before doing
5454 3. Implemented the mua_wrapper option.
5456 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5458 5. Implemented the functions header_remove(), header_testname(),
5459 header_add_at_position(), and receive_remove_recipient(), and exported them
5462 6. If an ACL "warn" statement specified the addition of headers, Exim already
5463 inserted X-ACL-Warn: at the start if there was no header name. However, it
5464 was not making this test for the second and subsequent header lines if
5465 there were newlines in the string. This meant that an invalid header could
5466 be inserted if Exim was badly configured.
5468 7. Allow an ACL "warn" statement to add header lines at the start or after all
5469 the Received: headers, as well as at the end.
5471 8. Added the rcpt_4xx retry error code.
5473 9. Added postmaster_mailfrom=xxx to callout verification option.
5475 10. Added mailfrom=xxxx to the callout verification option, for verify=
5478 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5479 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5481 12. Inserted some casts to stop certain compilers warning when using pointer
5482 differences as field lengths or precisions in printf-type calls (mostly
5483 affecting debugging statements).
5485 13. Added optional readline() support for -be (dynamically loaded).
5487 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5488 same clock tick as a message's arrival, so that its received time was the
5489 same as the "first fail" time on the retry record, and that message
5490 remained on the queue past the ultimate address timeout, every queue runner
5491 would try a delivery (because it was past the ultimate address timeout) but
5492 after another failure, the ultimate address timeout, which should have then
5493 bounced the address, did not kick in. This was a "< instead of <=" error;
5494 in most cases the first failure would have been in the next clock tick
5495 after the received time, and all would be well.
5497 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5498 being recognized when the domain list was tested by the match_domain
5499 condition in an expansion string.
5501 16. Added the ${str2b64: operator.
5503 17. Exim was always calling setrlimit() to set a large limit for the number of
5504 processes, without checking whether the existing limit was already
5505 adequate. (It did check for the limit on file descriptors.) Furthermore,
5506 errors from getrlimit() and setrlimit() were being ignored. Now they are
5507 logged to the main and panic logs, but Exim does carry on, to try to do its
5508 job under whatever limits there are.
5510 18. Imported PCRE 5.0.
5512 19. Trivial typo in log message " temporarily refused connection" (the leading
5515 20. If the log selector return_path_on_delivery was set and an address was
5516 redirected to /dev/null, the delivery process crashed because it assumed
5517 that a return path would always be set for a "successful" delivery. In this
5518 case, the whole delivery is bypassed as an optimization, and therefore no
5521 21. Internal re-arrangement: the function for sending a challenge and reading
5522 a response while authentication was assuming a zero-terminated challenge
5523 string. It's now changed to take a pointer and a length, to allow for
5524 binary data in such strings.
5526 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5528 23. Exim was not respecting finduser_retries when seeking the login of the
5529 uid under which it was called; it was always trying 10 times. (The default
5530 setting of finduser_retries is zero.) Also, it was sleeping after the final
5531 failure, which is pointless.
5533 24. Implemented tls_on_connect_ports.
5535 25. Implemented acl_smtp_predata.
5537 26. If the domain in control=submission is set empty, Exim assumes that the
5538 authenticated id is a complete email address when it generates From: or
5539 Sender: header lines.
5541 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5542 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5543 chown and chgrp in /bin and hostname in /usr/bin.
5545 28. Exim was keeping the "process log" file open after each use, just as it
5546 does for the main log. This opens the possibility of it remaining open for
5547 long periods when the USR1 signal hits a daemon. Occasional processlog
5548 errors were reported, that could have been caused by this. Anyway, it seems
5549 much more sensible not to leave this file open at all, so that is what now
5552 29. The long-running daemon process does not normally write to the log once it
5553 has entered its main loop, and it closes the log before doing so. This is
5554 so that log files can straightforwardly be renamed and moved. However,
5555 there are a couple of unusual error situations where the daemon does write
5556 log entries, and I had neglected to close the log afterwards.
5558 30. The text of an SMTP error response that was received during a remote
5559 delivery was being truncated at 512 bytes. This is too short for some of
5560 the long messages that one sometimes sees. I've increased the limit to
5563 31. It is now possible to make retry rules that apply only when a message has a
5564 specific sender, in particular, an empty sender.
5566 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5567 it possible to be selective about when SMTP synchronization is enforced.
5569 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5571 32. Implemented hosts_connection_nolog.
5573 33. Added an ACL for QUIT.
5575 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5578 35. Added mailbox_size and mailbox_filecount to appendfile.
5580 36. Added control = no_multiline_responses to ACLs.
5582 37. There was a bug in the logic of the code that waits for the clock to tick
5583 in the case where the clock went backwards by a substantial amount such
5584 that the microsecond fraction of "now" was more than the microsecond
5585 fraction of "then" (but the whole seconds number was less).
5587 38. Added support for the libradius Radius client library this is found on
5588 FreeBSD (previously only the radiusclient library was supported).
5594 1. When certain lookups returned multiple values in the form name=value, the
5595 quoting of the values was not always being done properly. Specifically:
5596 (a) If the value started with a double quote, but contained no whitespace,
5598 (b) If the value contained whitespace other than a space character (i.e.
5599 tabs or newlines or carriage returns) it was not quoted.
5600 This fix has been applied to the mysql and pgsql lookups by writing a
5601 separate quoting function and calling it from the lookup code. The fix
5602 should probably also be applied to nisplus, ibase and oracle lookups, but
5603 since I cannot test any of those, I have not disturbed their existing code.
5605 2. A hit in the callout cache for a specific address caused a log line with no
5606 reason for rejecting RCPT. Now it says "Previous (cached) callout
5607 verification failure".
5609 3. There was an off-by-one bug in the queryprogram router. An over-long
5610 return line was truncated at 256 instead of 255 characters, thereby
5611 overflowing its buffer with the terminating zero. As well as fixing this, I
5612 have increased the buffer size to 1024 (and made a note to document this).
5614 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5615 when Exim is waiting for an SMTP response from a remote server, Exim
5616 restarts its select() call on the socket, thereby resetting its timeout.
5617 This is not a problem when such interrupts are rare. Somebody set up a cron
5618 job to run exiwhat every 2 minutes, which is less than the normal select()
5619 timeout (5 or 10 minutes). This meant that the select() timeout never
5620 kicked in because it was always reset. I have fixed this by comparing the
5621 time when an interrupt arrives with the time at the start of the first call
5622 to select(). If more time than the timeout has elapsed, the interrupt is
5623 treated as a timeout.
5625 5. Some internal re-factoring in preparation for the addition of Sieve
5626 extensions (by MH). In particular, the "personal" test is moved to a
5627 separate function, and given an option for scanning Cc: and Bcc: (which is
5628 not set for Exim filters).
5630 6. When Exim created an email address using the login of the caller as the
5631 local part (e.g. when creating a From: or Sender: header line), it was not
5632 quoting the local part when it contained special characters such as @.
5634 7. Installed new OpenBSD configuration files.
5636 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5637 try to make them clearer.
5639 9. Callout options, other than the timeout value, were being ignored when
5640 verifying sender addresses in header lines. For example, when using
5642 verify = header_sender/callout=no_cache
5644 the cache was (incorrectly) being used.
5646 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5647 only the Cygwin environment.
5649 11. When return_path_on_delivery was set as a log selector, if different remote
5650 addresses in the same message used different return paths and parallel
5651 remote delivery occurred, the wrong values would sometimes be logged.
5652 (Whenever a remote delivery process finished, the return path value from
5653 the most recently started remote delivery process was logged.)
5655 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5656 lines when AUTH and/or TLS are in use. This is the "received protocol"
5657 field. Exim used to use "asmtp" for authenticated SMTP, without any
5658 indication (in the protocol name) for TLS use. Now it follows the RFC and
5659 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5660 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5661 names appear in log lines as well as in Received: header lines.
5663 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5664 extensions, and comparison tests, and to fix some bugs.
5666 14. Changes to the "personal" filter test:
5668 (1) The test was buggy in that it was just doing the equivalent of
5669 "contains" tests on header lines. For example, if a user's address was
5670 anne@some.where, the "personal" test would incorrectly be true for
5672 To: susanne@some.where
5674 This test is now done by extracting each address from the header in turn,
5675 and checking the entire address. Other tests that are part of "personal"
5676 are now done using regular expressions (for example, to check local parts
5677 of addresses in From: header lines).
5679 (2) The list of non-personal local parts in From: addresses has been
5680 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5681 taken from the Sieve specification recommendations.
5683 (3) If the message contains any header line starting with "List-" it is
5684 treated as non-personal.
5686 (4) The test for "circular" in the Subject: header line has been removed
5687 because it now seems ill-conceived.
5689 15. Minor typos in src/EDITME comments corrected.
5691 16. Installed latest exipick from John Jetmore.
5693 17. If headers_add on a router specified a text string that was too long for
5694 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5695 of string_sprintf() is now avoided.
5697 18. $message_body_size was not set (it was always zero) when running the DATA
5698 ACL and the local_scan() function.
5700 19. For the "mail" command in an Exim filter, no default was being set for
5701 the once_repeat time, causing a random time value to be used if "once" was
5702 specified. (If the value happened to be <= 0, no repeat happened.) The
5703 default is now 0s, meaning "never repeat". The "vacation" command was OK
5704 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5705 (I found it when inspecting the code).
5707 20. There is now an overall timeout for performing a callout verification. It
5708 defaults to 4 times the callout timeout, which applies to individual SMTP
5709 commands during the callout. The overall timeout applies when there is more
5710 than one host that can be tried. The timeout is checked before trying the
5711 next host. This prevents very long delays if there are a large number of
5712 hosts and all are timing out (e.g. when the network connections are timing
5713 out). The value of the overall timeout can be changed by specifying an
5714 additional sub-option for "callout", called "maxwait". For example:
5716 verify = sender/callout=5s,maxwait=20s
5718 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5719 to the end before writing, but this should make it even safer).
5721 22. Exim was forgetting that it had advertised PIPELINING for the second and
5722 subsequent messages on an SMTP connection. It was also not resetting its
5723 memory on STARTTLS and an internal HELO.
5725 23. When Exim logs an SMTP synchronization error within a session, it now
5726 records whether PIPELINING has been advertised or not.
5728 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5729 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5730 rather than long int.
5732 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5734 26. Added the never_mail option to autoreply.
5740 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5741 crash if the getsockname() call failed; this can happen if a connection is
5742 closed very soon after it is established. The problem was simply in the
5743 order in which certain operations were done, causing Exim to try to write
5744 to the SMTP stream before it had set up the file descriptor. The bug has
5745 been fixed by making things happen in the correct order.
5751 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5752 before the connection was closed, thus losing the rejection response.
5754 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5755 some early Solaris releases, but causes trouble in current releases where
5756 socklen_t is defined.
5758 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5761 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5764 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5765 that releasing the top of it at the end releases what was used for sub-
5766 expansions (unless the block got too big). However, discard this block if
5767 the first thing is a variable or header, so that we can use its block when
5768 it is dynamic (useful for very large $message_headers, for example).
5770 6. Lookups now cache *every* query, not just the most recent. A new, separate
5771 store pool is used for this. It can be recovered when all lookup caches are
5772 flushed. Lookups now release memory at the end of their result strings.
5773 This has involved some general refactoring of the lookup sources.
5775 7. Some code has been added to the store_xxx() functions to reduce the amount
5776 of flapping under certain conditions.
5778 8. log_incoming_interface used to affect only the <= reception log lines. Now
5779 it causes the local interface and port to be added to several more SMTP log
5780 lines, for example "SMTP connection from", and rejection lines.
5782 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5784 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5786 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5787 could be overwritten at the end of the current message (or the start of a
5788 new message if it was set in a HELO ACL). The value is now preserved for
5789 the duration of the SMTP connection.
5791 12. If a transport had a headers_rewrite setting, and a matching header line
5792 contained an unqualified address, that address was qualified, even if it
5793 did not match any rewriting rules. The underlying bug was that the values
5794 of the flags that permit the existence of unqualified sender and recipient
5795 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5796 non-local messages, and by -bnq for local messages) were not being
5797 preserved with the message after it was received.
5799 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5800 "next input=" as part of the text comprising the host identity instead of
5801 the correct text. The code was using the same buffer for two different
5802 strings. However, depending on which order the printing function evaluated
5803 its arguments, the bug did not always show up. Under Linux, for example, my
5804 test suite worked just fine.
5806 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5807 doesn't work with some older versions of Perl. It has been changed to "my",
5808 which in any case is probably the better facility to use.
5810 15. A really picky compiler found some instances of statements for creating
5811 error messages that either had too many or two few arguments for the format
5814 16. The size of the buffer for calls to the DNS resolver has been increased
5815 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5816 for addresses that have a lot of PTR records. This alleviates a problem; it
5817 does not fully solve it.
5819 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5820 buffer now truncates the list and logs the incident, which is the same
5821 action as happens when Exim is looking up a host name and its aliases.
5822 Previously in this situation something unpredictable would happen;
5823 sometimes it was "internal error: store_reset failed".
5825 18. If a server dropped the connection unexpectedly when an Exim client was
5826 using GnuTLS and trying to read a response, the client delivery process
5827 crashed while trying to generate an error log message.
5829 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5830 string, for example:
5832 warn message = H1: something\nH2: something
5834 the text was added as a single header line from Exim's point of view
5835 though it ended up OK in the delivered message. However, searching for the
5836 second and subsequent header lines using $h_h2: did not work. This has been
5837 fixed. Similarly, if a system filter added multiple headers in this way,
5838 the routers could not see them.
5840 20. Expanded the error message when iplsearch is called with an invalid key to
5841 suggest using net-iplsearch in a host list.
5843 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5844 ACLs are no longer actually imposed (and a message to that effect is
5847 22. If a "gecos" field in a passwd entry contained escaped characters, in
5848 particular, if it contained a \" sequence, Exim got it wrong when building
5849 a From: or a Sender: header from that name. A second bug also caused
5850 incorrect handling when an unquoted " was present following a character
5851 that needed quoting.
5853 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5854 was not being matched caselessly.
5856 24. Arranged for all hyphens in the exim.8 source to be escaped with
5859 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5860 properly. Recipient callout cache records were still being keyed to include
5861 the sender, even when use_sender was set false. This led to far more
5862 callouts that were necessary. The sender is no longer included in the key
5863 when use_sender is false.
5865 26. Added "control = submission" modifier to ACLs.
5867 27. Added the ${base62d: operator to decode base 62 numbers.
5869 28. dnsdb lookups can now access SRV records.
5871 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5872 the configuration file.
5874 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5875 (-v) mode. This makes the output for a verbose queue run more intelligible.
5877 31. Added a use_postmaster feature to recipient callouts.
5879 32. Added the $body_zerocount variable, containing the number of binary zero
5880 bytes in the message body.
5882 33. The time of last modification of the "new" subdirectory is now used as the
5883 "mailbox time last read" when there is a quota error for a maildir
5886 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5888 35. Added +ignore_unknown as a special item in host lists.
5890 36. Code for decoding IPv6 addresses in host lists is now included, even if
5891 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5892 address was recognized as an IP address, but was then not correctly decoded
5893 into binary, causing unexpected and incorrect effects when compared with
5900 1. Very minor rewording of debugging text in manualroute to say "list of
5901 hosts" instead of "hostlist".
5903 2. If verify=header_syntax was set, and a header line with an unqualified
5904 address (no domain) and a large number of spaces between the end of the
5905 name and the colon was received, the reception process suffered a buffer
5906 overflow, and (when I tested it) crashed. This was caused by some obsolete
5907 code that should have been removed. The fix is to remove it!
5909 3. When running in the test harness, delay a bit after writing a bounce
5910 message to get a bit more predictability in the log output.
5912 4. Added a call to search_tidyup() just before forking a reception process. In
5913 theory, someone could use a lookup in the expansion of smtp_accept_max_
5914 per_host which, without the tidyup, could leave open a database connection.
5916 5. Added the variables $recipient_data and $sender_data which get set from a
5917 lookup success in an ACL "recipients" or "senders" condition, or a router
5918 "senders" option, similar to $domain_data and $local_part_data.
5920 6. Moved the writing of debug_print from before to after the "senders" test
5923 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5924 problems for message scanning, either using a data ACL, or using
5925 local_scan() because the Received: header was not generated till after they
5926 were called (in order to set the time as the time of reception completion).
5927 I have revised the way this works. The header is now generated after the
5928 body is received, but before the ACL or local_scan() are called. After they
5929 are run, the timestamp in the header is updated.
5935 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5936 before starting a queue runner without re-exec. This happened only when
5937 deliver_drop_privilege was set or when the Exim user was set to root. The
5938 effect of the bug was that timeouts during subsequent deliveries caused
5939 crashes instead of being properly handled. The handler is now left at its
5940 default (and expected) setting.
5942 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5943 message, again when deliver_drop_privilege is set or Exim is run as root.
5944 The bug described in (1) was not present in this case, but the tidying up
5945 of the other signals was missing. I have made the two cases consistent.
5947 3. The ignore_target_hosts setting on a manualroute router was being ignored
5948 for hosts that were looked up using the /MX notation.
5950 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5953 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5954 operated on the sender address. After changing the $sender_address to <>
5955 for the sender address verify, Exim was re-instated it as the original
5956 (before rewriting) address, but remembering that it had rewritten it, so it
5957 wasn't rewriting it again. This bug also had the effect of breaking the
5958 sender address verification caching when the sender address was rewritten.
5960 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5961 This has been changed so that if the ip literal address matches
5962 ignore_target_hosts, the router declines.
5964 7. Added expansion conditions match_domain, match_address, and match_local_
5965 part (NOT match_host).
5967 8. The placeholder for the Received: header didn't have a length field set.
5969 9. Added code to Exim itself and to exim_lock to test for a specific race
5970 condition that could lead to file corruption when using MBX delivery. The
5971 issue is with the lockfile that is created in /tmp. If this file is removed
5972 after a process has opened it but before that process has acquired a lock,
5973 there is the potential for a second process to recreate the file and also
5974 acquire a lock. This could lead to two Exim processes writing to the file
5975 at the same time. The added code performs the same test as UW imapd; it
5976 checks after acquiring the lock that its file descriptor still refers to
5977 the same named file.
5979 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5980 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5983 11. Added the smtp_active_hostname option. If used, this will typically be made
5984 to depend on the incoming interface address. Because $interface_address is
5985 not set up until the daemon has forked a reception process, error responses
5986 that can happen earlier (such as "too many connections") no longer contain
5989 12. If an expansion in a condition on a "warn" statement fails because a lookup
5990 defers, the "warn" statement is abandoned, and the next ACL statement is
5991 processed. Previously this caused the whole ACL to be aborted.
5993 13. Added the iplsearch lookup type.
5995 14. Added ident_timeout as a log selector.
5997 15. Added tls_certificate_verified as a log selector.
5999 16. Added a global option tls_require_ciphers (compare the smtp transport
6000 option of the same name). This controls incoming TLS connections.
6002 17. I finally figured out how to make tls_require_ciphers do a similar thing
6003 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6004 before starting the TLS session.
6006 18. Tabs are now shown as \t in -bP output.
6008 19. If the log selector return_path_on_delivery was set, Exim crashed when
6009 bouncing a message because it had too many Received: header lines.
6011 20. If two routers both had headers_remove settings, and the first one included
6012 a superfluous trailing colon, the final name in the first list and the
6013 first name in the second list were incorrectly joined into one item (with a
6014 colon in the middle).
6020 1. Added -C and -D options to the exinext utility, mainly to make it easier
6021 to include in the automated testing, but these could be helpful when
6022 multiple configurations are in use.
6024 2. The exinext utility was not formatting the output nicely when there was
6025 an alternate port involved in the retry record key, nor when there was a
6026 message id as well (for retries that were specific to a specific message
6027 and a specific host). It was also confused by IPv6 addresses, because of
6028 the additional colons they contain. I have fixed the IPv4 problem, and
6029 patched it up to do a reasonable job for IPv6.
6031 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6032 delivery, the log line now contains "pipelined" if PIPELINING was used.
6034 4. An SMTP transport process used to panic and die if the bind() call to set
6035 an explicit outgoing interface failed. This has been changed; it is now
6036 treated in the same way as a connect() failure.
6038 5. A reference to $sender_host_name in the part of a conditional expansion
6039 that was being skipped was still causing a DNS lookup. This no longer
6042 6. The def: expansion condition was not recognizing references to header lines
6043 that used bh_ and bheader_.
6045 7. Added the _cache feature to named lists.
6047 8. The code for checking quota_filecount in the appendfile transport was
6048 allowing one more file than it should have been.
6050 9. For compatibility with Sendmail, the command line option
6058 and sets the incoming protocol and host name (for trusted callers). The
6059 host name and its colon can be omitted when only the protocol is to be set.
6060 Note the Exim already has two private options, -pd and -ps, that refer to
6061 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6062 "s", but I don't think that's a major issue.
6064 10. A number of refactoring changes to the code, none of which should affect
6067 (a) The number of logging options was getting close to filling up the
6068 32-bit word that was used as a bit map. I have split them into two classes:
6069 those that are passed in the argument to log_write(), and those that are
6070 only ever tested independently outside of that function. These are now in
6071 separate 32-bit words, so there is plenty of room for expansion again.
6072 There is no change in the user interface or the logging behaviour.
6074 (b) When building, for example, log lines, the code previously used a
6075 macro that called string_cat() twice, in order to add two strings. This is
6076 not really sufficiently general. Furthermore, there was one instance where
6077 it was actually wrong because one of the argument was used twice, and in
6078 one call a function was used. (As it happened, calling the function twice
6079 did not affect the overall behaviour.) The macro has been replaced by a
6080 function that can join an arbitrary number of extra strings onto a growing
6083 (c) The code for expansion conditions now uses a table and a binary chop
6084 instead of a serial search (which was left over from when there were very
6085 few conditions). Also, it now recognizes conditions like "pam" even when
6086 the relevant support is not compiled in: a suitably worded error message is
6087 given if an attempt is made to use such a condition.
6089 11. Added ${time_interval:xxxxx}.
6091 12. A bug was causing one of the ddress fields not to be passed back correctly
6092 from remote delivery subprocesses. The field in question was not being
6093 subsequently used, so this caused to problems in practice.
6095 13. Added new log selectors queue_time and deliver_time.
6097 14. Might have fixed a bug in maildirsizefile handling that threw up
6098 "unexpected character" debug warnings, and recalculated the data
6099 unnecessarily. In any case, I expanded the warning message to give more
6102 15. Added the message "Restricted characters in address" to the statements in
6103 the default ACL that block characters like @ and % in local parts.
6105 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6106 Three changes have been made:
6108 (a) There was a serious bug; a negative response to MAIL caused the whole
6109 recipient domain to be cached as invalid, thereby blocking all messages
6110 to all local parts at the same domain, from all senders. This bug has
6111 been fixed. The domain is no longer cached after a negative response to
6112 MAIL if the sender used is not empty.
6114 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6117 (c) A new callout option, "use_sender" has been added for people who want
6118 the modified behaviour.
6124 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6127 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6128 indeed breaks things for older releases.
6130 3. Added additional logging to the case where there is a problem reading data
6131 from a filter that is running in a subprocess using a pipe, in order to
6132 try to track down a specific problem.
6134 4. Testing facility fudge: when running in the test harness and attempting
6135 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6136 getting "No route to host". Convert this to a timeout.
6138 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6141 6. Some OS don't have socklen_t but use size_t instead. This affects the
6142 fifth argument of getsockopt() amongst other things. This is now
6143 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6144 can be set for individual OS. I have set it for SunOS5, OSF1, and
6145 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6146 some earlier ones do not.
6148 7. Change 4.30/15 was not doing the test caselessly.
6150 8. The standard form for an IPv6 address literal was being rejected by address
6151 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6152 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6153 this, as well as the form without the "IPv6" on the front (but only when
6154 address literals are enabled, of course).
6156 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6158 10. Exim crashed if a message with an empty sender address specified by -f
6159 encountered a router with an errors_to setting. This could be provoked only
6160 by a command such as
6164 where an empty string was supplied; "<>" did not hit this bug.
6166 11. Installed PCRE release 4.5.
6168 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6169 remained set. It is now erased.
6171 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6172 times from message ids (which are base 36 rather than the normal 62).
6174 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6175 were being counted as actual protocol errors, and logged if the log
6176 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6177 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6178 and DATA following a set of rejected RCPTs do not count as protocol errors.
6179 In other words, Exim assumes they were pipelined, though this may not
6180 actually be the case. Of course, in all cases the client gets an
6181 appropriate error code.
6183 15. If a lookup fails in an ACL condition, a message about the failure may
6184 be available; it is used if testing the ACL cannot continue, because most
6185 such messages specify what the cause of the deferral is. However, some
6186 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6187 that caused an old message to be retained and used if a later statement
6188 caused a defer, replacing the real cause of the deferral.
6190 16. If an IP address had so many PTR records that the DNS lookup buffer
6191 was not large enough to hold them, Exim could crash while trying to process
6192 the truncated data. It now detects and logs this case.
6194 17. Further to 4.21/58, another change has been made: if (and only if) the
6195 first line of a message (the first header line) ends with CRLF, a bare LF
6196 in a subsequent header line has a space inserted after it, so as not to
6197 terminate the header.
6199 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6200 unnecessarily, used atoi() instead of strtol(), and didn't check the
6201 termination when getting file sizes from file names by regex.
6203 19. Completely re-implemented the support for maildirsize files, in the light
6204 of a number of problems with the previous contributed implementation
6205 (4.30/29). In particular:
6207 . If the quota is zero, the maildirsize file is maintained, but no quota is
6210 . If the maildir directory does not exist, it is created before any attempt
6211 to write a maildirsize file.
6213 . The quota value in the file is just a cache; if the quota is changed in
6214 the transport, the new value overrides.
6216 . A regular expression is available for excluding directories from the
6219 20. The autoreply transport checks the characters in options that define the
6220 message's headers; it allows continued headers, but it was checking with
6221 isspace() after an embedded newline instead of explicitly looking for a
6224 21. If all the "regular" hosts to which an address was routed had passed their
6225 expiry times, and had not reached their retry times, the address was
6226 bounced, even if fallback hosts were defined. Now Exim should go on to try
6229 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6230 equivalent code in the SMTP transport. Some hosts send humungous responses
6231 to HELO/EHLO, more than 1024 it seems.
6233 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6234 gives compiler warnings in some environments. I've now done it "properly",
6237 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6238 (because of problems with the built-in one) was declared to return uschar *
6239 instead of char *, causing compiler failure.
6241 25. Fixed a file descriptor leak when processing alias/forward files.
6243 26. Fixed a minor format string issue in dbfn.c.
6245 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6247 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6248 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6251 29. When a sender address is verified, it is cached, to save repeating the test
6252 when there is more than one recipient in a message. However, when the
6253 verification involves a callout, it is possible for different callout
6254 options to be set for different recipients. It is too complicated to keep
6255 track of this in the cache, so now Exim always runs a verification when a
6256 callout is required, relying on the callout cache for the optimization.
6257 The overhead is duplication of the address routing, but this should not be
6260 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6261 to be verified with callout=postmaster, and the main callout worked but the
6262 postmaster check failed, the verification correctly failed. However, if a
6263 subsequent RCPT command asked for sender verification *without* the
6264 postmaster check, incorrect caching caused this verification also to fail,
6267 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6268 it was not caching the DNS options (qualify_single, search_parents) that
6269 were used when the lookup failed. A subsequent lookup with different
6270 options therefore always gave the same answer, though there were cases
6271 where it should not have. (Example: a "domains = !$mx_any" option on a
6272 dnslookup router: the "domains" option is always processed without any
6273 widening, but the router might have qualify_single set.) Now Exim uses the
6274 cached value only when the same options are set.
6276 32. Added John Jetmore's "exipick" utility to the distribution.
6278 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6279 than a timeout (e.g. a certificate is required, and is not provided), an
6280 Exim server now closes the connection immediately. Previously it waited for
6281 the client to close - but if the client is SSL, it seems that they each
6282 wait for each other, leading to a delay before one of them times out.
6284 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6285 maintained 0.8.x compatibility because I don't think many are using it, and
6286 it is clearly obsolete.
6288 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6291 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6292 client certificate was expired. A simple patch fixes this, though I don't
6293 understand the full logic of why the verify callback is called multiple
6296 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6297 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6298 which causes problems with some clients (such as the Certicom SSL Plus
6299 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6300 disables the coutermeasure allowing Eudora to connect."
6302 38. Exim was not checking that a write() to a log file succeeded. This could
6303 lead to Bad Things if a log got too big, in particular if it hit a file
6304 size limit. Exim now panics and dies if it cannot write to a log file, just
6305 as it does if it cannot open a log file.
6307 39. Modified OS/Makefile-Linux so that it now contains
6309 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6311 The two -D definitions ensure that Exim is compiled with large file
6312 support, which makes it possible to handle log files that are bigger than
6315 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6316 instance) a domain was checked against a named list that involved a lookup,
6317 causing $domain_data to be set, then another domain was checked against the
6318 same list, then the first domain was re-checked, the value of $domain_data
6319 after the final check could be wrong. In particular, if the second check
6320 failed, it could be set empty. This bug probably also applied to
6323 41. The strip_trailing_dot option was not being applied to the address given
6324 with the -f command-line option.
6326 42. The code for reading a message's header from the spool was incrementing
6327 $received_count, but never initializing it. This meant that the value was
6328 incorrect (doubled) while delivering a message in the same process in which
6329 it was received. In the most common configuration of Exim, this never
6330 happens - a fresh exec is done - but it can happen when
6331 deliver_drop_privilege is set.
6333 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6334 it now includes up to 150 characters of the unexpected data in the log
6337 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6338 and building data strings. The size of both of these buffers was 10 000
6339 bytes - far larger than anybody would *ever* want, thought I. Needless to
6340 say, somebody hit the limit. I have increased the maximum line length to
6341 20 000 and the maximum data length of concatenated lines to 100 000. I have
6342 also fixed two bugs, because there was no checking on these buffers. Tsk,
6343 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6344 buffer is too small.
6346 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6347 lsearch lookups. Now it does.
6349 46. When parsing a route_list item in a manualroute router, a fixed-length
6350 buffer was used for the list of hosts. I made this 1024 bytes long,
6351 thinking that nobody would ever have a list of hosts that long. Wrong.
6352 Somebody had a whole pile of complicated expansion conditions, and the
6353 string was silently truncated, leading to an expansion error. It turns out
6354 that it is easier to change to an unlimited length (owing to other changes
6355 that have happened since this code was originally written) than to build
6356 structure for giving a limitation error. The length of the item that
6357 expands into the list of hosts is now unlimited.
6359 47. The lsearch lookup could not handle data where the length of text line was
6360 more than 4095 characters. Such lines were truncated, leading to shortened
6361 data being returned. It should now handle lines of any length.
6363 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6364 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6367 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6368 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6369 no longer happens in comment lines. A list of replacements is now placed
6370 at the head of all of the source files, except those whose only change is
6371 to replace PERL_COMMAND in the very first #! line.
6373 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6374 messages on the queue, with a bottom-up merge sort, using code contributed
6375 by Michael Haardt. This should make operations like -bp somewhat faster on
6376 large queues. It won't affect queue runners, except when queue_run_in_order
6379 51. Installed eximstats 1.31 in the distribution.
6381 52. Added support for SRV lookups to the dnslookup router.
6383 53. If an ACL referred to $message_body or $message_body_end, the value was not
6384 reset for any messages that followed in the same SMTP session.
6386 54. The store-handling optimization for building very long strings was not
6387 differentiating between the different store pools. I don't think this
6388 actually made any difference in practice, but I've tidied it.
6390 55. While running the routers to verify a sender address, $sender_address
6391 was still set to the sender address. This is wrong, because when routing to
6392 send a bounce to the sender, it would be empty. Therefore, I have changed
6393 it so that, while verifying a sender address, $sender_address is set to <>.
6394 (There is no change to what happens when verifying a recipient address.)
6396 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6397 target A or AAAA records (if not already returned) without resetting the
6398 qualify_single or search_parents options of the DNS resolver. These are
6399 inappropriate in this case because the targets of MX and SRV records must
6400 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6401 target that, when qualified, matched something in the local domain. These
6402 two options are now turned off when doing these lookups.
6404 57. It seems that at least some releases of Reiserfs (which does not have the
6405 concept of a fixed number of inodes) returns zero and not -1 for the
6406 number of available inodes. This interacted badly with check_spool_inodes,
6407 which assumed that -1 was the "no such thing" setting. What I have done is
6408 to check that the total number of inodes is greater than zero before doing
6409 the test of how many are available.
6411 58. When a "warn" ACL statement has a log_message modifier, the message is
6412 remembered, and not repeated. This is to avoid a lot of repetition when a
6413 message has many recipients that cause the same warning to be written.
6414 However, Exim was preserving the list of already written lines for an
6415 entire SMTP session, which doesn't seem right. The memory is now reset if a
6416 new message is started.
6418 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6419 debugging output unless log_rewrite was also set.
6421 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6422 of (int)(handle) when we know that handle contains (void *)(-1).
6424 61. The Exim daemon panic-logs an error return when it closes the incoming
6425 connection. However "connection reset by peer" seems to be common, and
6426 isn't really an error worthy of noting specially, so that particular error
6429 62. When Exim is trying to find all the local interfaces, it used to panic and
6430 die if the ioctl to get the interface flags failed. However, it seems that
6431 on at least one OS (Solaris 9) it is possible to have an interface that is
6432 included in the list of interfaces, but for which you get a failure error
6433 for this call. This happens when the interface is not "plumbed" into a
6434 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6435 failure of the "get flags" call assumes that the interface is down.
6437 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6438 makes life easier for people who are doing arithmetic on fields extracted
6439 from dates, where you often get leading zeros that should not be
6440 interpreted as octal.
6442 64. Added qualify_domain to the redirect router, to override the global
6445 65. If a pathologically long header line contained very many addresses (the
6446 report of this problem mentioned 10 000) and each of them was rewritten,
6447 Exim could use up a very large amount of memory. (It kept on making new
6448 copies of the header line as it rewrote, and never released the old ones.)
6449 At the expense of a bit more processing, the header rewriting function has
6450 been changed so that it no longer eats memory in this way.
6452 66. The generation of the Received: header has been moved from the time that a
6453 message starts to be received, to the time that it finishes. The timestamp
6454 in the Received: header should now be very close to that of the <= log
6455 line. There are two side-effects of this change:
6457 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6458 logged header lines no longer include the local Received: line, because
6459 it has not yet been created. The same applies to a copy of the message
6460 that is returned to a non-SMTP sender when a message is rejected.
6462 (b) When a filter file is tested using -bf, no additional Received: header
6463 is added to the test message. After some thought, I decided that this
6466 This change does not affect the value of $received_for. It is still set
6467 after address rewriting, but before local_scan() is called.
6469 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6471 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6472 gave an unhelpful panic error message, and a defer error. I have managed to
6473 change this behaviour so that it now rejects any supplied certificate,
6474 which seems right, as the list of acceptable certificates is empty.
6476 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6477 gave an unhelpful defer error. I have not managed to make this reject any
6478 supplied certificates, but the error message it gives is "no certificate
6479 supplied", which is not helpful.
6481 70. exigrep's output now also includes lines that are not associated with any
6482 message, but which match the given pattern. Implemented by a patch from
6483 Martin Sluka, which also tidied up the Perl a bit.
6485 71. Recipient callout verification, like sender verification, was using <> in
6486 the MAIL FROM command. This isn't really the right thing, since the actual
6487 sender may affect whether the remote host accepts the recipient or not. I
6488 have changed it to use the actual sender in the callout; this means that
6489 the cache record is now keyed on a recipient/sender pair, not just the
6490 recipient address. There doesn't seem to be a real danger of callout loops,
6491 since a callout by the remote host to check the sender would use <>.
6492 [SEE ABOVE: changed after hitting problems.]
6494 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6495 temporary errors. However, in the case of such a code being given after
6496 the end of a data transmission (i.e. after ".") Exim was failing to write
6497 a retry record for the message. (Yes, there was some broken host that was
6498 actually sending 8xx at this point.)
6500 73. An unknown lookup type in a host list could cause Exim to panic-die when
6501 the list was checked. (An example that provoked this was putting <; in the
6502 middle of a list instead of at the start.) If this happened during a DATA
6503 ACL check, a -D file could be left lying around. This kind of configuration
6504 error no longer causes Exim to die; instead it causes a defer error. The
6505 incident is still logged to the main and panic logs.
6507 74. Buglet left over from Exim 3 conversion. The message "too many messages
6508 in one connection" was written to the rejectlog but not the mainlog, except
6509 when address rewriting (yes!) was being logged.
6511 75. Added write_rejectlog option.
6513 76. When a system filter was run not as root (that is, when system_filter_user
6514 was set), the values of the $n variables were not being returned to the
6515 main process; thus, they were not subsequently available in the $sn
6518 77. Added +return_path_on_delivery log selector.
6520 78. A connection timeout was being treated differently from recipients deferred
6521 when testing hosts_max_try with a message that was older than the host's
6522 retry timeout. (The host should not be counted, thus allowing all hosts to
6523 be tried at least once before bouncing.) This may have been the cause of an
6524 occasionally reported bug whereby a message would remain on the queue
6525 longer than the retry timeout, but would be bounced if a delivery was
6526 forced. I say "may" because I never totally pinned down the problem;
6527 setting up timeout/retry tests is difficult. See also the next item.
6529 79. The ultimate address timeout was not being applied to errors that involved
6530 a combination of host plus message (for example, a timeout on a MAIL
6531 command). When an address resolved to a number of possible hosts, and they
6532 were not all tried for each delivery (e.g. because of hosts_max_try), a
6533 message could remain on the queue longer than the retry timeout.
6535 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6538 81. Fixed an obscure SMTP outgoing bug which required at least the following
6539 conditions: (a) there was another message waiting for the same server;
6540 (b) the server returned 5xx to all RCPT commands in the first message so
6541 that the message was not completed; (c) the server dropped the connection
6542 or gave a negative response to the RSET that Exim sends to abort the
6543 transaction. The observed case was a dropped connection after DATA that had
6544 been sent in pipelining mode. That is, the server had advertised PIPELINING
6545 but was not implementing it correctly. The effect of the bug was incorrect
6546 behaviour, such as trying another host, and this could lead to a crash.
6552 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6553 and daemon.c were passed as pointers to ints; they should have been
6554 pointers to socklen_t variables (which are typically unsigned ints).
6556 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6559 3. Fixed a really odd bug that affected only the testing scheme; patching a
6560 certain fixed string in the binary changed the value of another string that
6561 happened to be identical to the end of the original first string.
6563 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6564 name", it returns that address as the IP address. On some operating
6565 systems (e.g. Solaris), it also passes back the IP address string as the
6566 "host name". However, on others (e.g. Linux), it passes back an empty
6567 string. Exim wasn't checking for this, and was changing the host name to an
6568 empty string, assuming it had been canonicalized.
6570 5. Although rare, it is permitted to have more than one PTR record for a given
6571 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6572 all the names associated with an address, because they do in Solaris.
6573 However, it seems that they do not in Linux for data that comes from the
6574 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6575 I found this out when I moved to a new Linux workstation and tried to run
6576 the Exim test suite.
6578 To get round this problem I have changed the code so that it now does its
6579 own call to the DNS to look up PTR records when searching for a host name.
6580 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6581 addresses that are only in /etc/hosts are still found.
6583 This behaviour is, however, controlled by an option called host_lookup_
6584 order, which defaults to "bydns:byaddr". If people want to use the other
6585 order, or indeed, just use one or the other means of lookup, they can
6586 specify it in this variable.
6588 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6589 some operating systems, this comes back from gethostbyaddr() as an empty
6590 string, and this is what Exim used to test for. However, it seems that in
6591 other systems, "." is yielded. Exim now tests for this case too.
6593 7. The values of check_spool_space and check_log_space are now held internally
6594 as a number of kilobytes instead of an absolute number of bytes. If a
6595 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6596 kilobyte. This means that much larger values can be stored.
6598 8. Exim monitor: an attempt to get the action menu when not actually pointing
6599 at a message produces an empty menu entitled "No message selected". This
6600 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6601 no entries in it ("Shell widget menu has zero width and/or height"). So I
6602 have added a single, blank menu entry in this case.
6604 9. Added ${quote_local_part.
6606 10. MIME decoding is now applied to the contents of Subject: header lines when
6609 11. Now that a reference to $sender_host_address automatically causes a reverse
6610 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6611 host lookup before query-style lookups in lists that might use this
6612 variable. This has therefore been abolished, and the "net-" prefix is no
6613 longer necessary for query-style lookups.
6615 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6616 have been a typo for LFLAGS, so it has been changed.
6618 13. The install script calls Exim with "-C /dev/null" in order to find the
6619 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6620 to be output. However, since Exim outputs its version number before the
6621 error, it didn't break the script. It just looked ugly. I fixed this by
6622 always allowing "-C /dev/null" if the caller is root.
6624 14. Ignore overlarge ACL variable number when reading spool file - insurance
6625 against a later release with more variables having written the file.
6627 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6628 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6629 this, as well as the form without the "IPv6" on the front.
6631 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6632 OS/Makefile-Darwin file.
6634 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6635 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6637 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6638 variable. However, some OS use ints for inodes. I've added cast to long int
6639 to get rid of the compiler warning.
6641 19. I had forgotten to lock out "/../" in configuration file names when
6642 ALT_CONFIG_PREFIX was set.
6644 20. Routers used for verification do not need to specify transports. However,
6645 if such a router generated a host list, and callout was configured, Exim
6646 crashed, because it could not find a port number from the (non-existent)
6647 transport. It now assumes port 25 in this circumstance.
6649 21. Added the -t option to exigrep.
6651 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6652 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6653 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6654 them is set, LOOKUP_LSEARCH is forced.
6656 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6658 24. Added sender and host information to the "rejected by local_scan()" log
6659 line; previously there was no indication of these.
6661 25. Added .include_if_exists.
6663 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6664 Linux. It turns out that not all file systems support this. Apparently some
6665 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6666 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6667 not supported on the file descriptor, is now ignored when Exim is trying to
6668 sync a directory. This applies only to Linux.
6670 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6672 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6673 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6676 29. Added support for maildirsize files from supplied patch (modified a bit).
6678 30. The use of :fail: followed by an empty string could lead Exim to respond to
6679 sender verification failures with (e.g.):
6681 550 Verification failed for <xxx>
6682 550 Sender verify failed
6684 where the first response line was missing the '-' that indicates it is not
6685 the final line of the response.
6687 31. The loop for finding the name of the user that called Exim had a hardwired
6688 limit of 10; it now uses the value of finduser_retries, which is used for
6689 all other user lookups.
6691 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6694 33. Exim was neglecting to zero errno before one call of strtol() when
6695 expanding a string and expecting an integer value. On some systems this
6696 resulted in spurious "integer overflow" errors. Also, it was casting the
6697 result into an int without checking.
6699 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6700 did not work. The code looks as if it has *never* worked, though it appears
6701 to have been documented since at least release 1.62. I have made it work.
6703 35. The "timeout_DNS" error in retry rules, also documented since at least
6704 1.62, also never worked. As it isn't clear exactly what this means, and
6705 clearly it isn't a major issue, I have abolished the feature by treating it
6706 as "timeout", and writing a warning to the main and panic logs.
6708 36. The display of retry rules for -brt wasn't always showing the error code
6711 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6712 timeout_connect_A, timeout_connect_MX.
6714 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6715 to the empty sender.
6717 39. The daemon was not analysing the content of -oX till after it had closed
6718 stderr and disconnected from the controlling terminal. This meant that any
6719 syntax errors were only noted on the panic log, and the return code from
6720 the command was 0. By re-arranging the code a little, I've made the
6721 decoding happen first, so such errors now appear on stderr, and the return
6722 code is 1. However, the actual setting up of the sockets still happens in
6723 the disconnected process, so errors there are still only recorded on the
6726 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6727 connections (as happens on some IP stacks) was logged at start up time as
6728 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6729 it from "IPv6 and IPv4", which means that two separate sockets are being
6732 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6733 says whether AF_INET or AF_INET6 was passed as an argument.
6735 42. Exiwhat output was messed up when time zones were included in log
6738 43. Exiwhat now gives more information about the daemon's listening ports,
6739 and whether -tls-on-connect was used.
6741 44. The "port" option of the smtp transport is now expanded.
6743 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6744 silently ignored. Now an error message is written to the main and panic
6747 46. There's a new ACL modifier called "logwrite" which writes to a log file
6748 as soon as it is encountered.
6750 47. Added $local_user_uid and $local_user_gid at routing time.
6752 48. Exim crashed when trying to verify a sender address that was being
6755 49. Exim was recognizing only a space character after ".include". It now also
6756 recognizes a tab character.
6758 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6759 extracting the relevant information from the specification. The man page no
6760 longer contains scrambled data for the -d option, and I've added a section
6761 at the front about calling Exim under different names.
6763 51. Added "extra_headers" argument to the "mail" command in filter files.
6765 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6768 53. Installed eximstats 1.29.
6770 54. Added transport_filter_timeout as a generic transport option.
6772 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6773 Cc: header lines. This was required by RFC 822, but it not required by RFC
6776 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6777 incoming messages that did not have them. Now it does so only if the
6778 message originates locally, that is, if there is no associated remote host
6779 address. When Resent- header lines are present, this applies to the Resent-
6780 lines rather than the non-Resent- lines.
6782 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6783 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6785 58. Messages for configuration errors now include the name of the main
6786 configuration file - useful now that there may be more than one file in a
6787 list (.included file names were always shown).
6789 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6790 for those rare installations that do not start the daemon as root or run it
6791 setuid root. I've cut out the call to initgroups() if the daemon is not
6794 60. The Exim user and group can now be bound into the binary as text strings
6795 that are looked up at the start of Exim's processing.
6797 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6799 62. Added $mailstore_basename variable.
6801 63. Installed patch to sieve.c from Michael Haardt.
6803 64. When Exim failed to open the panic log after failing to open the main log,
6804 the original message it was trying to log was written to stderr and debug
6805 output, but if they were not available (the usual case in production), it
6806 was lost. Now it is written to syslog before the two lines that record the
6807 failures to open the logs.
6809 65. Users' Exim filters run in subprocesses under the user's uid. It is
6810 possible for a "deliver" command or an alias in a "personal" command to
6811 provoke an address rewrite. If logging of address rewriting is configured,
6812 this fails because the process is not running as root or exim. There may be
6813 a better way of dealing with this, but for the moment (because 4.30 needs
6814 to be released), I have disabled address rewrite logging when running a
6815 filter in a non-root, non-exim process.
6821 1. The buildconfig auxiliary program wasn't quoting the value set for
6822 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6823 not defined. This bug was masked in 4.22 by the effect that was fixed in
6826 2. Some messages that were rejected after a message id was allocated were
6827 shown as "incomplete" by exigrep. It no longer does this for messages that
6828 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6830 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6831 have allow_domain_literals set, the ID did not get logged in the <= line.
6832 Domain literals are now always recognized in Message-ID: header lines.
6834 4. The first argument for a ${extract expansion item is the key name or field
6835 number. Leading and trailing spaces in this item were not being ignored,
6836 causing some misleading effects.
6838 5. When deliver_drop_privilege was set, single queue runner processes started
6839 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6840 same command in the process it spins off) were not dropping privilege.
6842 6. When the daemon running as "exim" started a queue runner, it always
6843 re-executed Exim in the spun-off process. This is a waste of effort when
6844 deliver_drop_privilege is set. The new process now just calls the
6845 queue-runner function directly.
6851 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6854 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6855 ignored. Though the use of -oP was forcing the writing of a pid file, it
6856 was always written to the default place.
6858 3. If the message "no IP address found for host xxxx" is generated during
6859 incoming verification, it is now followed by identification of the incoming
6860 connection (so you can more easily find what provoked it).
6862 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6864 5. Added some features to "harden" Exim a bit more against certain attacks:
6866 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6867 be put in Local/Makefile. This is like the never_users runtime option,
6868 but it cannot be overridden. The default setting is "root".
6870 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6871 prefix string with which any file named in a -C command line option
6874 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6875 is retained for -C and -D only if the caller of Exim is root. Without
6876 it, the exim user may also use -C and -D and retain privilege.
6878 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6879 command line option is disabled.
6881 6. Macro names set by the -D option must start with an upper case letter, just
6882 like macro names defined in the configuration file.
6884 7. Added "dereference=" facility to LDAP.
6886 8. Two instances of the typo "uknown" in the source files are fixed.
6888 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6889 the Configure-Makefile script screwed up while processing it.
6891 10. Incorporated PCRE 4.4.
6893 11. The SMTP synchronization check was not operating right at the start of an
6894 SMTP session. For example, it could not catch a HELO sent before the client
6895 waited for the greeting. There is now a check for outstanding input at the
6896 point when the greeting is written. Because of the duplex, asynchronous
6897 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6898 way, but not yet received, when the check is performed.
6900 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6901 on TCP/IP sockets, because this apparently causes some broken clients to
6904 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6905 unchanged) from the Cygwin maintainer.
6907 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6908 was supported instead of testing for mbx. Effectively a typo.
6910 15. The spa authenticator server code was not checking that the input it
6911 received was valid base64.
6913 16. The debug output line for the "set" modifier in ACLs was not showing the
6914 name of the variable that was being set.
6916 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6918 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6919 lookup on its own. Something else was needed to trigger the lookup. For
6920 example, a match in host_lookup or the need for a host name in a host list.
6921 Now, if $sender_host_name is referenced and the host name has not yet been
6922 looked up, a lookup is performed. If the lookup fails, the variable remains
6923 empty, and $host_lookup_failed is set to "1".
6925 19. Added "eqi" as a case-independent comparison operator.
6927 20. The saslauthd authentication condition could segfault if neither service
6928 nor realm was specified.
6930 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6931 error message that was logged was misleading, and incoming SMTP
6932 connections were dropped. The message is now more accurate, and temporary
6933 errors are given to SMTP connections.
6935 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6936 (except RSET) with a 421 error, until QUIT is received. However, it was
6937 failing to send a response to QUIT.
6939 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6940 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6941 ACL is now run after the helo_try_verify_hosts code.
6943 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6944 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6945 case-independent, so other case variants are also recognized). Apparently
6946 some systems use these upper case variants.
6948 25. If more than two messages were waiting for the same host, and a transport
6949 filter was specified for the transport, Exim sent two messages over the
6950 same TCP/IP connection, and then failed with "socket operation on non-
6951 socket" when it tried to send the third.
6953 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6955 27. The extern definition of crypt16() in expand.c was not being excluded when
6956 the OS had its own crypt16() function.
6958 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6959 as a preferred synonym for return_size_limit, both as an option and as an
6962 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6964 30. Changed the default configuration ACL to relax the local part checking rule
6965 for addresses that are not in any local domains. For these addresses,
6966 slashes and pipe symbols are allowed within local parts, but the sequence
6967 /../ is explicitly forbidden.
6969 31. SPA server authentication was not clearing the challenge buffer before
6972 32. log_message in a "warn" ACL statement was writing to the reject log as
6973 well as to the main log, which contradicts the documentation and doesn't
6974 seem right (because no rejection is happening). So I have stopped it.
6976 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6977 However, I am unable to do any testing of this.
6979 34. Fixed an infelicity in the appendfile transport. When checking directories
6980 for a mailbox, to see if any needed to be created, it was accidentally
6981 using path names with one or more superfluous leading slashes; tracing
6982 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6984 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6985 contents are added to the log line that is written for every discarded
6986 recipient. (Previously a log_message setting was ignored.)
6988 36. The ${quote: operator now quotes the string if it is empty.
6990 37. The install script runs exim in order to find its version number. If for
6991 some reason other than non-existence or emptiness, which it checks, it
6992 could not run './exim', it was installing it with an empty version number,
6993 i.e. as "exim-". This error state is now caught, and the installation is
6996 38. An argument was missing from the function that creates an error message
6997 when Exim fails to connect to the socket for saslauthd authentication.
6998 This could cause Exim to crash, or give a corrupted message.
7000 39. Added isip, isip4, and isip6 to ${if conditions.
7002 40. The ACL variables $acl_xx are now saved with the message, and can be
7003 accessed later in routers, transports, and filters.
7005 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7006 strings in the file are not string-expanded.
7008 42. If a MAIL command specified a SIZE value that was too large to fit into an
7009 int variable, the check against message_size_limit failed. Such values are
7010 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7011 day this will have to be increased, but I don't think I want to be around
7012 when emails are that large.
7019 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7020 iconv() is not standard in FreeBSD.
7022 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7023 IPv6 enabled. The observed symptom was a segmentation fault on return from
7024 the function os_common_find_running_interfaces() in src/os.c.
7026 3. In the check_special_case() function in daemon.c I had used "errno" as an
7027 argument name, which causes warnings on some systems. This was basically a
7028 typo, since it was named "eno" in the comments!
7030 4. The code that waits for the clock to tick (at a resolution of some fraction
7031 of a second) so as to ensure message-id uniqueness was always waiting for
7032 at least one whole tick, when it could have waited for less. [This is
7033 almost certainly not relevant at current processor speeds, where it is
7034 unlikely to ever wait at all. But we try to future-proof.]
7036 5. The function that sleeps for a time interval that includes fractions of a
7037 second contained a race. It did not block SIGALRM between setting the
7038 timer, and suspending (a couple of lines later). If the interval was short
7039 and the sigsuspend() was delayed until after it had expired, the suspension
7040 never ended. On busy systems this could lead to processes getting stuck for
7043 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7044 process, before it forks any delivery processes. The open lookup caching
7045 mechanism meant that the open file or database connection was passed into
7046 the delivery process. The problem was that delivery processes always tidy
7047 up cached lookup data. This could cause a problem for the next delivery
7048 process started by the queue runner, because the external queue runner
7049 process does not know about the closure. So the next delivery process
7050 still has data in the lookup cache. In the case of a file lookup, there was
7051 no problem because closing a file descriptor in a subprocess doesn't affect
7052 the parent. However, if the lookup was caching a connection to a database,
7053 the connection was closed, and the second delivery process was likely to
7054 see errors such as "PGSQL: query failed: server closed the connection
7055 unexpectedly". The problem has been fixed by closing all cached lookups
7056 in a queue runner before running a delivery process.
7058 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7059 seem to have the "const" qualifier which it has on other OS. I've
7062 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7063 *of the same type* (client or server) with the same public name that an
7064 error should be diagnosed.
7066 9. When Exim looked up a host name for an IP address, but failed to find the
7067 original IP address when looking up the host name (a safety check), it
7068 output the message "<ip address> does not match any IP for NULL", which was
7069 confusing, to say the least. The bug was that the host name should have
7070 appeared instead of "NULL".
7072 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7073 user that is built into the binary, and the -C or -D options is used, root
7074 privilege is dropped before the configuration file is read. In addition,
7075 logging is switched to stderr instead of the normal log files. If the
7076 configuration then re-defines the Exim user, the unprivileged environment
7077 is probably not what is expected, so Exim logs a panic warning message (but
7080 However, if deliver_drop_privilege is set, the unprivileged state may well
7081 be exactly what is intended, so the warning has been cut out in that case,
7082 and Exim is allowed to try to write to its normal log files.
7088 1. smtp_return_error_details was not giving details for temporary sender
7089 or receiver verification errors.
7091 2. Diagnose a configuration error if two authenticators have the same public
7094 3. Exim used not to create the message log file for a message until the first
7095 delivery attempt. This could be confusing when incoming messages were held
7096 for policy or load reasons. The message log file is now created at the time
7097 the message is received, and an initial "Received" line is written to it.
7099 4. The automatically generated man page for command line options had a minor
7100 bug that caused no ill effects; however, a more serious problem was that
7101 the procedure for building the man page automatically didn't always
7102 operate. Consequently, release 4.20 contains an out-of-date version. This
7103 shouldn't happen again.
7105 5. When building Exim with embedded Perl support, the script that builds the
7106 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7107 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7109 6. The freeze_tell option was not being used for messages that were frozen on
7110 arrival, either by an ACL or by local_scan().
7112 7. Added the smtp_incomplete_transaction log selector.
7114 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7115 was accepting AUTH without a new EHLO.
7117 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7118 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7121 10. Logging of TCP/IP connections (when configured) now happens in the main
7122 daemon process instead of the child process, so that the TCP/IP connection
7123 count is more accurate (but it can never be perfect).
7125 11. The use of "drop" in a nested ACL was not being handled correctly in the
7126 outer ACL. Now, if condition failure induced by the nested "drop" causes
7127 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7128 or "require"), the connection is dropped.
7130 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7131 that yield "discard" can now be used with an "accept" or a "discard" verb,
7132 but an error is generated for any others (because I can't see a useful way
7133 to define what should happen).
7135 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7136 are now processed in the same way as lines in the Exim configuration file.
7137 In particular, continuation lines are supported.
7139 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7141 15. Added -ti meaning -t -i.
7143 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7144 domains, and warn by logging if others are found.
7146 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7147 returned from GIFCONF when Exim is trying to find the list of interfaces on
7148 a host. The code in os.c has been modified to copy each ifreq to an aligned
7149 structure in all cases.
7151 Also, in some cases, the returned ifreq's were being copied to a 'struct
7152 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7153 means the last couple of bytes of an IPv6 address could be chopped if the
7154 ifreq contained only a normal sockaddr (14 bytes storage).
7156 18. Named domain lists were not supported in the hosts_treat_as_local option.
7157 An entry such as +xxxx was not recognized, and was treated as a literal
7160 19. Ensure that header lines added by a DATA ACL are included in the reject log
7161 if the ACL subsequently rejects the message.
7163 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7164 MD5 (which is deprecated).
7166 21. When testing a filter file using -bf, Exim was writing a message when it
7167 took the sender from a "From " line in the message, but it was not doing so
7168 when it took $return_path from a Return-Path: header line. It now does.
7170 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7171 with a valid header line field name (a series of printing characters
7172 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7174 23. Changed "disc" in the source to "disk" to conform to the documentation and
7175 the book and for uniformity.
7177 24. Ignore Sendmail's -Ooption=value command line item.
7179 25. When execve() failed while trying to run a command in a pipe transport,
7180 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7181 could be confused with a return value of 69 from the command itself. This
7182 has been changed to 127, the value the shell returns if it is asked to run
7183 a non-existent command. The wording for the related log line suggests a
7184 non-existent command as the problem.
7186 26. If received_header_text expands to an empty string, do not add a Received:
7187 header line to the message. (Well, it adds a token one on the spool, but
7188 marks it "old" so that it doesn't get used or transmitted.)
7190 27. Installed eximstats 1.28 (addition of -nt option).
7192 28. There was no check for failure on the call to getsockname() in the daemon
7193 code. This can fail if there is a shortage of resources on the system, with
7194 ENOMEM, for example. A temporary error is now given on failure.
7196 29. Contrary to the C standard, it seems that in some environments, the
7197 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7198 program. Exim now does this explicitly; it affects the formatting of
7199 timestamps using strftime().
7201 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7202 complaints. I've now initialized all the variables, to avoid this.
7204 32. Header lines added by a system filter were not being "seen" during
7205 transport-time rewrites.
7207 33. The info_callback() function passed to OpenSSL is set up with type void
7208 (*)(SSL *, int, int), as described somewhere. However, when calling the
7209 function (actually a macro) that sets it up, the type void(*)() is
7210 expected. I've put in a cast to prevent warnings from picky compilers.
7212 34. If a DNS black list lookup found a CNAME record, but there were no A
7213 records associated with the domain it pointed at, Exim crashed.
7215 35. If a DNS black list lookup returned more than one A record, Exim ignored
7216 all but the first. It now scans all returned addresses if a particular IP
7217 value is being sought. In this situation, the contents of the
7218 $dnslist_value variable are a list of all the addresses, separated by a
7221 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7222 to accept a host name and all its aliases if the forward lookup for any of
7223 them yielded the IP address of the incoming connection. Now it accepts only
7224 those names whose forward lookup yields the correct IP address. Any other
7225 names are discarded. This closes a loophole whereby a rogue DNS
7226 administrator could create reverse DNS records to break through a
7227 wildcarded host restriction in an ACL.
7229 37. If a user filter or a system filter that ran in a subprocess used any of
7230 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7231 the wrong values were passed to the pipe command ($thisaddress had the
7232 value of $0, $0 had the value of $1, etc). This bug was introduced by
7233 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7235 38. Improved the line breaking for long SMTP error messages from ACLs.
7236 Previously, if there was no break point between 40 and 75 characters, Exim
7237 left the rest of the message alone. Two changes have been made: (a) I've
7238 reduced the minimum length to 35 characters; (b) if it can't find a break
7239 point between 35 and 75 characters, it looks ahead and uses the first one
7240 that it finds. This may give the occasional overlong line, but at least the
7241 remaining text gets split now.
7243 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7244 file descriptors might be low, and that setting 1000 would always raise it.
7245 It turns out that in some environments, the limit is already over 1000 and
7246 that lowering it causes trouble. So now Exim takes care not to decrease it.
7248 40. When delivering a message, the value of $return_path is set to $sender_
7249 address at the start of routing (routers may change the value). By an
7250 oversight, this default was not being set up when an address was tested by
7251 -bt or -bv, which affected the outcome if any router or filter referred to
7254 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7255 log only when "message" or "log_message" are set. However, if one of the
7256 conditions was an address verification, or a call to a nested ACL, the
7257 messages generated by the underlying test were being passed through. This
7258 no longer happens. The underlying message is available in $acl_verify_
7259 message for both "message" and "log_message" expansions, so it can be
7260 passed through if needed.
7262 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7263 new expansion $bh_ to give the encoded byte string without charset
7264 translation. Translation happens only if iconv() is available; HAVE_ICONV
7265 indicates this at build time. HEADERS_CHARSET gives the charset to
7266 translate to; headers_charset can change it in the configuration, and
7267 "headers charset" can change it in an individual filter file.
7269 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7270 that creates RFC 2047 encoded "words" labels them as that charset instead
7271 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7272 expansion operator; (ii) when Exim creates a From: line for a local
7273 message; (iii) when a header line is rewritten to include a "phrase" part.
7275 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7276 buggy, causing it to skip the first lines of messages whose message ID
7277 ended in 'D'. This would not have bitten before Exim release 4.14, because
7278 message IDs were unlikely to end in 'D' before then. The effect was to have
7279 incorrect size information for certain domains.
7281 45. #include "config.h" was missing at the start of the crypt16.c module. This
7282 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7285 46. If there was a timeout during a "random" callout check, Exim treated it as
7286 a failure of the random address, and carried on sending RSET and the real
7287 address. If the delay was just some slowness somewhere, the response to the
7288 original RCPT would be taken as a response to RSET and so on, causing
7289 mayhem of various kinds.
7291 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7292 when I implemented it. It didn't allow for the fact that some option values
7293 may legitimately be negative (e.g. size_addition), and it didn't even do
7294 the right test for positive values.
7296 48. Domain names in DNS records are case-independent. Exim always looks them up
7297 in lower case. Some resolvers return domain names in exactly the case they
7298 appear in the zone file, that is, they may contain uppercase letters. Not
7299 all resolvers do this - some return always lower case. Exim was treating a
7300 change of case by a resolver as a change of domain, similar to a widening
7301 of a domain abbreviation. This triggered its re-routing code and so it was
7302 trying to route what was effectively the same domain again. This normally
7303 caused routing to fail (because the router wouldn't handle the domain
7304 twice). Now Exim checks for this case specially, and just changes the
7305 casing of the domain that it ultimately uses when it transmits the message
7308 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7311 50. If a filter generated a file delivery with a non-absolute name (possible if
7312 no home directory exists for the router), the forbid_file option was not
7315 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7316 the existing equality matching.
7318 52. Exim was using ints instead of ino_t variables in some places where it was
7319 dealing with inode numbers.
7321 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7322 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7323 called TMPDIR, and if it finds it is different, it changes its value.
7325 54. The smtp_printf() function is now made available to local_scan() so
7326 additional output lines can be written before returning. There is also an
7327 smtp_fflush() function to enable the detection of a dropped connection.
7328 The variables smtp_input and smtp_batched_input are exported to
7331 55. Changed the default runtime configuration: the message "Unknown user"
7332 has been removed from the ACL, and instead placed on the localuser router,
7333 using the cannot_route_message feature. This means that any verification
7334 failures that generate their own messages won't get overridden. Similarly,
7335 the "Unrouteable address" message that was in the ACL for unverifiable
7336 relay addresses has also been removed.
7338 56. Added hosts_avoid_esmtp to the smtp transport.
7340 57. The exicyclog script was not checking for the esoteric option
7341 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7342 will work only if exicyclog is run under the appropriate euid.
7344 58. Following a discussion on the list, the rules by which Exim recognises line
7345 endings on incoming messages have been changed. The -dropcr and drop_cr
7346 options are now no-ops, retained only for backwards compatibility. The
7347 following line terminators are recognized: LF CRLF CR. However, special
7348 processing applies to CR:
7350 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7351 nor a local message in the state where . is a terminator.
7353 (ii) If a bare CR is encountered in a header line, an extra space is added
7354 after the line terminator so as not to end the header. The reasoning
7355 behind this is that bare CRs in header lines are most likely either
7356 to be mistakes, or people trying to play silly games.
7358 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7359 was being incorrectly given as 18 bytes larger than it should have been.
7360 This is a VOB (very old bug).
7362 60. This may never have affected anything current, but just in case it has:
7363 When the local host is found other than at the start of a list of hosts,
7364 the local host, those with the same MX, and any that follow, are discarded.
7365 When the list in question was part of a longer list of hosts, the following
7366 hosts (not currently being processed) were also being discarded. This no
7367 longer happens. I'm not sure if this situation could ever has previously
7370 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7373 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7374 header. This is something that is recommended in a new Internet Draft, and
7375 is something that is documented as being done by Sendmail. There are two
7376 possible values. For messages generated by the autoreply transport, Exim
7379 Auto-Submitted: auto-replied
7381 whereas for all other generated messages (e.g. bounces) it adds
7383 Auto-Submitted: auto-generated
7385 63. The "personal" condition in filters now includes a test for the
7386 Auto-Submitted: header. If it contains the string "auto-" the message it
7387 not considered personal.
7389 64. Added rcpt_include_affixes as a generic transport option.
7391 65. Added queue_only_override (default true).
7393 66. Added the syslog_duplication option.
7395 67. If what should have been the first header line of a message consisted of
7396 a space followed by a colon, Exim was mis-interpreting it as a header line.
7397 It isn't of course - it is syntactically invalid and should therefore be
7398 treated as the start of the message body. The misbehaviour could have
7399 caused a number of strange effects, including loss of data in subsequent
7400 header lines, and spool format errors.
7402 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7403 client host had authenticated. This control can now be exercised by an ACL
7404 for more flexibility.
7406 69. By default, callouts do not happen when testing with -bh. There is now a
7407 variant, -bhc, which does actually run the callout code, including
7408 consulting and updating the callout cache.
7410 70. Added support for saslauthd authentication, courtesy of Alexander
7413 71. If statvfs() failed on the spool or log directories while checking their
7414 size for availability, Exim confusingly gave the error "space shortage".
7415 Furthermore, in debugging mode it crashed with a floating point exception.
7416 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7417 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7418 serious problem, Exim now writes to the main and panic logs when this
7419 happens, with details of the failure. It then refuses to accept the
7420 incoming message, giving the message "spool directory problem" or "log
7421 directory problem" with a 421 code for SMTP messages.
7423 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7424 0, 1, and 2 exist, because some OS complain for execs without them (see
7425 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7426 descriptors. However, the code omitted to check that the open succeeded,
7427 causing mysterious errors if for some reason the permissions on /dev/null
7428 got screwed. Now Exim writes a message to the main and panic logs, and
7429 bombs out if it can't open /dev/null.
7431 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7432 interact so that it is all more flexible. It is supposed to remain
7433 backwards compatible. Also added extra_local_interfaces.
7435 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7436 to bomb out with an assertion failure - to the client this appears as a
7437 connection drop. This problem occurs in the part of the code that was taken
7438 from the Samba project. Fortunately, the assertion is in a very simple
7439 function, so I have fixed this by reproducing the function inline in the
7440 one place where it is called, and arranging for authentication to fail
7441 instead of killing the process with assert().
7443 75. The SPA client code was not working when the server requested OEM rather
7444 than Unicode encoding.
7446 76. Added code to make require_files with a specific uid setting more usable in
7447 the case where statting the file as root fails - usually a non-root-mounted
7448 NFS file system. When this happens and the failure is EACCES, Exim now
7449 forks a subprocess and does the per-uid checking as the relevant uid.
7451 77. Added process_log_path.
7453 78. If log_file_path was not explicitly set, a setting of check_log_space or
7454 check_log_inodes was ignored.
7456 79. If a space check for the spool or log partitions fails, the incident is now
7457 logged. Of course, in the latter case the data may get lost...
7459 80. Added the %p formatting code to string_format() so that it can be used to
7460 print addresses in debug_print(). Adjusted all the address printing in the
7461 debugging in store.c to use %p rather than %d.
7463 81. There was a concern that a line of code in smtp_in.c could overflow a
7464 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7465 initially expressed, the concern was not well-founded, because trailing
7466 spaces are removed early. However, if the trailing spaces were followed by
7467 a NULL, they did not get removed, so the overflow was possible. Two fixes
7470 (a) I re-wrote the offending code in a cleaner fashion.
7471 (b) If an incoming SMTP command contains a NULL character, it is rejected
7474 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7475 runs initgroups(), so that if the Exim user is in any additional groups,
7476 they will be used during message reception.
7482 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob