1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
12 side, to decode a TLS 1.3 packet capture.
14 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
15 Previously the default library behaviour applied, sending two, each in
18 JH/03 Debug output for ACL now gives the config file name and line number for
21 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
23 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
25 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
26 buffer overrun for (non-chunking) other transports.
28 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
29 TLS1.3, means that a server rejecting a client certificate is not visible
30 to the client until the first read of encrypted data (typically the
31 response to EHLO). Add detection for that case and treat it as a failed
32 TLS connection attempt, so that the normal retry-in-clear can work (if
35 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
36 and/or domain. Found and fixed by Jason Betts.
38 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
39 configuration). If a CNAME target was not a wellformed name pattern, a
42 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
43 the OS reports them interleaved with other addresses.
45 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
46 used both for input and for a verify callout, both encrypted, SMTP
47 responses being sent by the server could be lost. This resulted in
48 dropped connections and sometimes bounces generated by a peer sending
51 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
52 string. Previously it was possible to cause undefined behaviour in a
53 library routine (usually a crash). Found by "zerons".
55 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
58 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
59 API was removed, so update to use the newer ones.
61 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
62 any timeout set, is taking a long time. Previously we would hang on to a
63 rotated logfile "forever" if the input was arriving with long gaps
64 (a previous attempt to fix addressed lack, for a long time, of initial
67 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
68 shared (NFS) environment. The length of the tempfile name is now
69 4 + 16 ("hdr.$message_exim_id") which might break on file
70 systems which restrict the file name length to lower values.
73 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
74 shared (NFS) environment.
76 HS/02 Bug 2392: exigrep does case sensitive *option* processing (as it
77 did for all versions <4.90). Notably -M, -m, --invert, -I may be
80 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
81 on some platforms for bit 31.
83 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
84 to changes apparently associated with TLS1.3 handling some of the APIs
85 previously used were either nonfunctional or inappropriate. Strings
86 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
87 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
88 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
89 This affects log line X= elements, the $tls_{in,out}_cipher variables,
90 and the use of specific cipher names in the encrypted= ACL condition.
92 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
94 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
95 verification result was not updated unless hosts_require_ocsp applied.
97 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
98 queue_list_requires_admin set to false, non-admin users were denied the
101 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
102 directory-of-certs mode. Previously they were advertised despite the
105 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
106 A single TCP connection by a client will now hold a TLS connection open
107 for multiple message deliveries, by default. Previoud the default was to
110 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
111 default. If built with the facility, DANE will be used. The facility
112 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
114 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
115 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
116 must be defined and you must still, unless you define DISABLE_TLS, manage
117 the the include-dir and library-file requirements that go with that
118 choice. Non-TLS builds are still supported.
120 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
123 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
124 default. If the platform supports and has the facility enabled, it will
125 be requested on all coneections.
127 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
128 controlled by the build-time option SUPPORT_PIPE_CONNECT.
134 JH/01 Remove code calling the customisable local_scan function, unless a new
135 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
137 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
138 non-signal-safe functions being used.
140 JH/03 Bug 2269: When presented with a received message having a stupidly large
141 number of DKIM-Signature headers, disable DKIM verification to avoid
142 a resource-consumption attack. The limit is set at twenty.
144 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
145 report of oldest_pass in ${authres } in consequence, and separate out
146 some descriptions of reasons for verification fail.
148 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
149 files in the spool were present and unlocked. A queue-runner could spot
150 them, resulting in a duplicate delivery. Fix that by doing the unlock
151 after the unlink. Investigation by Tim Stewart. Take the opportunity to
152 add more error-checking on spoolfile handling while that code is being
155 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
156 No known attacks, no CVE, this is defensive hardening.
158 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
159 a queue-runner could start a delivery while other operations were ongoing.
160 Cutthrough delivery was a common victim, resulting in duplicate delivery.
161 Found and investigated by Tim Stewart. Fix by using the open message data
162 file handle rather than opening another, and not locally closing it (which
163 releases a lock) for that case, while creating the temporary .eml format
164 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
166 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
167 $sender_verify_failure/$recipient_verify_failure to "random".
169 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
172 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
173 Previously this would segfault.
175 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
178 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
179 like zero, since the resolver should be doing this for us, But we need one
180 as a CNAME but no MX presence gets the CNAME returned; we need to check
181 that doesn't point to an MX to declare it "no MX returned" rather than
182 "error, loop". A new main option is added so the older capability of
183 following some limited number of chain links is maintained.
185 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
187 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
188 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
189 now-deprecated earlier definitions used only specified the range up to TLS
190 1.2 (in the older-version library docs).
192 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
194 JH/15 Rework TLS client-side context management. Stop using a global, and
195 explicitly pass a context around. This enables future use of TLS for
196 connections to service-daemons (eg. malware scanning) while a client smtp
197 connection is using TLS; with cutthrough connections this is quite likely.
199 JH/16 Fix ARC verification to do AS checks in reverse order.
201 JH/17 Support a "tls" option on the ${readsocket } expansion item.
203 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
204 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
205 Previously the "utf8" would be re-prepended for every additional message.
207 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
208 Previously thery were accepted, resulting in issues when attempting to
209 forward messages to a non-supporting MTA.
211 PP/02 Let -n work with printing macros too, not just options.
213 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
214 one parent address was copied, and bogus data was used at delivery-logging
215 time. Either a crash (after delivery) or bogus log data could result.
216 Discovery and analysis by Tim Stewart.
218 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
219 Previously if the string ended mid-character, we did not insert the
220 promised '?' replacement.
222 PP/04 Documentation: current string operators work on bytes, not codepoints.
224 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
225 should pack well giving a smaller memory footprint so better caching and
226 therefore performance. Group the declarations where this can't be done so
227 that the byte-sized flag variables are not interspersed among pointer
228 variables, giving a better chance of good packing by the compiler.
230 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
231 non-null, to avoid issues with sites running BATV. Previously reports were
232 sent with an empty envelope sender so looked like bounces.
234 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
235 The ignore_error flag wasn't being returned from the filter subprocess so
236 was not set for later routers. Investigation and fix by Matthias Kurz.
238 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
239 and a msg:complete for the whole, when a message is manually removed using
240 -Mrm. Developement by Matthias Kurz, hacked on by JH.
242 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
243 a "Gnu special" function, asprintf() in the DB utility binary builds; I
244 hope that is portable enough.
246 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
247 requiring a known-CA anchor certificate; make it now rely entirely on the
248 TLSA as an anchor. Checking the name on the leaf cert against the name
249 on the A-record for the host is still done for TA (but not for EE mode).
251 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
252 included in delivery lines for non-proxied connections, when compiled with
253 SUPPORT_SOCKS and running with proxy logging enabled.
255 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
256 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
257 move the existing event to fire before the normal logging of message
258 failure so that custom logging is bracketed by normal logging.
260 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
261 msg:fail:internal event. Developement by Matthias Kurz.
263 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
264 far too small for todays use of crypto signatures stored there. Go all
265 the way to the max DNS message size of 64kB, even though this might be
266 overmuch for IOT constrained device use.
268 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
269 copy a string over itself. The library routine is documented as not
270 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
272 JH/32 For main options check_spool_space and check_inode_space, where the
273 platform supports 64b integers, support more than the previous 2^31 kB
274 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
275 the previous G, M, k.
277 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
278 $authenticated_fail_id variable on authentication failure. Previously
281 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
282 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
283 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
284 GNUTLS_SEC_PARAM_MEDIUM.
286 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
287 side. Previously we would continue as if no SNI had been received.
289 JH/36 Harden the handling of string-lists. When a list consisted of a sole
290 "<" character, which should be a list-separator specification, we walked
291 off past the nul-terimation.
293 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
294 causes) even when the retry time is not yet met. Previously they were
295 not, meaning that when (say) an account was over-quota and temp-rejecting,
296 and multiple senders' messages were queued, only one sender would get
297 notified on each configured delay_warning cycle.
299 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
301 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
302 cipher-suites, an error can be left on the stack even for a succeeding
303 accept; this results in impossible error messages when a later operation
306 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
307 return error codes indicating retry. Under TLS1.3 this becomes required.
309 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
310 it only wrote the new authenticators, resulting in a lack of tracking of
311 peer changes of ESMTP extensions until the next cache flush.
313 JH/41 Fix the loop reading a message header line to check for integer overflow,
314 and more-often against header_maxsize. Previously a crafted message could
315 induce a crash of the recive process; now the message is cleanly rejected.
317 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
318 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
324 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
325 When redis_servers is set to a list of > 1 element, and the Redis servers
326 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
327 case of MOVED into a DEFER case instead, thus moving the query onto the
328 next server in the list. For a cluster of N elements, all N servers must
329 be defined in redis_servers.
331 GF/02 Catch and remove uninitialized value warning in exiqsumm
332 Check for existence of @ARGV before looking at $ARGV[0]
334 JH/01 Replace the store_release() internal interface with store_newblock(),
335 which internalises the check required to safely use the old one, plus
336 the allocate and data copy operations duplicated in both (!) of the
337 extant use locations.
339 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
340 modifier. This matches the restriction on the commandline.
342 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
343 Previously only the last row was returned.
345 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
346 we assumed that tags in the header were well-formed, and parsed the
347 element content after inspecting only the first char of the tag.
348 Assumptions at that stage could crash the receive process on malformed
351 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
352 While running the DKIM ACL we operate on the Permanent memory pool so that
353 variables created with "set" persist to the DATA ACL. Also (at any time)
354 DNS lookups that fail create cache records using the Permanent pool. But
355 expansions release any allocations made on the current pool - so a dnsdb
356 lookup expansion done in the DKIM ACL releases the memory used for the
357 DNS negative-cache, and bad things result. Solution is to switch to the
358 Main pool for expansions.
359 While we're in that code, add checks on the DNS cache during store_reset,
360 active in the testsuite.
361 Problem spotted, and debugging aided, by Wolfgang Breyha.
363 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
364 When none of the hosts presented to a transport match an already-open
365 connection, close it and proceed with the list. Previously we would
366 queue the message. Spotted by Lena with Yahoo, probably involving
369 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
370 Previously a spurious "250 OK id=" response was appended to the proper
373 JH/08 The "support for" informational output now, which built with Content
374 Scanning support, has a line for the malware scanner interfaces compiled
375 in. Interface can be individually included or not at build time.
377 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
378 by the template makefile "src/EDITME". The "STREAM" support for an older
379 ClamAV interface method is removed.
381 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
382 rows affected is given instead).
384 JH/11 The runtime Berkeley DB library version is now additionally output by
385 "exim -d -bV". Previously only the compile-time version was shown.
387 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
388 SMTP connection. Previously, when one had more receipients than the
389 first, an abortive onward connection was made. Move to full support for
390 multiple onward connections in sequence, handling cutthrough connection
391 for all multi-message initiating connections.
393 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
394 routers. Previously, a multi-recipient message would fail to match the
395 onward-connection opened for the first recipient, and cause its closure.
397 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
398 a timeout on read on a GnuTLS initiating connection, resulting in the
399 initiating connection being dropped. This mattered most when the callout
400 was marked defer_ok. Fix to keep the two timeout-detection methods
403 JH/15 Relax results from ACL control request to enable cutthrough, in
404 unsupported situations, from error to silently (except under debug)
405 ignoring. This covers use with PRDR, frozen messages, queue-only and
408 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
410 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
411 metadata, resulting in a crash in free().
413 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
414 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
415 Broken also in d185889f4, with init system revamp.
417 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
418 Previously we abruptly closed the connection after reading a malware-
419 found indication; now we go on to read the "scan ok" response line,
422 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
423 ACL. Previously, a crash would result.
425 JH/19 Speed up macro lookups during configuration file read, by skipping non-
426 macro text after a replacement (previously it was only once per line) and
427 by skipping builtin macros when searching for an uppercase lead character.
429 JH/20 DANE support moved from Experimental to mainline. The Makefile control
430 for the build is renamed.
432 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
433 was allocated for every new TLS startup, meaning one per message. Fix
434 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
436 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
437 reported the original. Fix to report (as far as possible) the ACL
438 result replacing the original.
440 JH/23 Fix memory leak during multi-message connections using STARTTLS under
441 OpenSSL. Certificate information is loaded for every new TLS startup,
442 and the resources needed to be freed.
444 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
446 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
449 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
450 DATA response info to the (existing) per-recipient response info for
451 the "C=" log element. It can have useful tracking info from the
452 destination system. Patch from Simon Arlott.
454 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
455 length value. Previously this would segfault.
457 HS/02 Support Avast multiline protoocol, this allows passing flags to
458 newer versions of the scanner.
460 JH/28 Ensure that variables possibly set during message acceptance are marked
461 dead before release of memory in the daemon loop. This stops complaints
462 about them when the debug_store option is enabled. Discovered specifically
463 for sender_rate_period, but applies to a whole set of variables.
464 Do the same for the queue-runner and queue-list loops, for variables set
465 from spool message files. Do the same for the SMTP per-message loop, for
466 certain variables indirectly set in ACL operations.
468 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
469 as a multi-recipient message from a mailinglist manager). The coding had
470 an arbitrary cutoff number of characters while checking for more input;
471 enforced by writing a NUL into the buffer. This corrupted long / fast
472 input. The problem was exposed more widely when more pipelineing of SMTP
473 responses was introduced, and one Exim system was feeding another.
474 The symptom is log complaints of SMTP syntax error (NUL chars) on the
475 receiving system, and refused recipients seen by the sending system
476 (propating to people being dropped from mailing lists).
477 Discovered and pinpointed by David Carter.
479 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
480 replaced by the ${authresults } expansion.
482 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
484 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
485 allows proper process termination in container environments.
487 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
488 Previously the "final dot" had a newline after it; ensure it is CR,LF.
490 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
491 and "err_perm", deprecated since 4.83 when the RFC-defined words
492 "temperror" and "permerror" were introduced.
494 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
495 transport-filters or DKIM-signing. The restriction was lost in the
496 consolidation of verify-callout and delivery SMTP handling.
497 Extend the restriction to also cover ARC-signing.
499 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
500 in defer=pass mode supply a 450 to the initiator. Previously the message
503 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
504 tls_require_ciphers is used as before.
506 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
507 "pass_unscanned". Only tmpfails from the scanner are written to
508 the paniclog, as they may require admin intervention (permission
509 denied, license issues). Other scanner errors (like decompression
510 bombs) do not cause a paniclog entry.
512 JH/36 Fix reinitialisation of DKIM logging variable between messages.
513 Previously it was possible to log spurious information in receive log
516 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
517 triggered odd behaviour from Outlook Express clients.
519 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
522 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
523 since the IETF WG has not yet settled on that versus the original
524 "bare" representation.
526 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
527 Previously the millisecond value corrupted the output.
528 Fix also for syslog_pid=no and log_selector +pid, for which the pid
529 corrupted the output.
535 JH/01 Rework error string handling in TLS interface so that the caller in
536 more cases is responsible for logging. This permits library-sourced
537 string to be attached to addresses during delivery, and collapses
538 pairs of long lines into single ones.
540 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
541 during configuration. Wildcards are allowed and expanded.
543 JH/02 Rework error string handling in DKIM to pass more info back to callers.
544 This permits better logging.
546 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
547 do not close it down and have the child transport start it up again on
548 the passed-on TCP connection. Instead, proxy the child (and any
549 subsequent ones) for TLS via a unix-domain socket channel. Logging is
550 affected: the continued delivery log lines do not have any DNSSEC, TLS
551 Certificate or OCSP information. TLS cipher information is still logged.
553 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
554 identical IP addresses on different listening ports. Will also affect
557 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
558 add noisy ifdef guards to special-case this sillyness.
559 Patch from Bernd Kuhls.
561 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
562 than 255 are no longer allowed.
564 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
565 Disable the session-cache too, which might reduce our load. Since we
566 currrectly use a new context for every connection, both as server and
567 client, there is no benefit for these.
568 GnuTLS appears to not support tickets server-side by default (we don't
569 call gnutls_session_ticket_enable_server()) but client side is enabled
570 by default on recent versions (3.1.3 +) unless the PFS priority string
573 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
574 <https://reproducible-builds.org/specs/source-date-epoch/>.
576 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
577 the check for any unsuccessful recipients did not notice the limit, and
578 erroneously found still-pending ones.
580 JH/08 Pipeline CHUNKING command and data together, on kernels that support
581 MSG_MORE. Only in-clear (not on TLS connections).
583 JH/09 Avoid using a temporary file during transport using dkim. Unless a
584 transport-filter is involved we can buffer the headers in memory for
585 creating the signature, and read the spool data file once for the
586 signature and again for transmission.
588 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
589 4.77 as the kernel support then wasn't solid, having issues in 64bit
590 mode. Now, it's been long enough. Add support for FreeBSD also.
592 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
593 case where the routing stage had gathered several addresses to send to
594 a host before calling the transport for the first, we previously failed
595 to close down TLS in the old transport process before passing the TCP
596 connection to the new process. The new one sent a STARTTLS command
597 which naturally failed, giving a failed delivery and bloating the retry
598 database. Investigation and fix prototype from Wolfgang Breyha.
600 JH/12 Fix check on SMTP command input synchronisation. Previously there were
601 false-negatives in the check that the sender had not preempted a response
602 or prompt from Exim (running as a server), due to that code's lack of
603 awareness of the SMTP input buffering.
605 PP/04 Add commandline_checks_require_admin option.
606 Exim drops privileges sanely, various checks such as -be aren't a
607 security problem, as long as you trust local users with access to their
608 own account. When invoked by services which pass untrusted data to
609 Exim, this might be an issue. Set this option in main configuration
610 AND make fixes to the calling application, such as using `--` to stop
613 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
614 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
615 responses to those, into a single TLS record each way (this usually means
616 a single packet). As a side issue, smtp_enforce_sync now works on TLS
619 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
620 affects you only if you're dancing at the edge of the param size limits.
621 If you are, and this message makes sense to you, then: raise the
622 configured limit or use OpenSSL 1.1. Nothing we can do for older
625 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
626 cmdline element to get the documented default one. Previously it was
629 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
630 are unsuable for non-message-specific reasons - eg. connection timeout,
631 banner-time rejection.
633 JH/16 Fix logging of delivery remote port, when specified by router, under
636 PP/06 Repair manualroute's ability to take options in any order, even if one
637 is the name of a transport.
640 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
642 JH/17 Change the list-building routines interface to use the expanding-string
643 triplet model, for better allocation and copying behaviour.
645 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
646 Previously it was constructed the first time a possibly-matching string
647 was met in the configuration file input during startup; now it is done
650 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
651 compatible one, to avoid the (poorly documented) possibility of a config
652 file in the working directory redirecting the DB files, possibly correpting
653 some existing file. CVE-2017-10140 assigned for BDB.
655 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
656 cache-hot. Previously, although the result was properly cached, the
657 initial verify call returned a defer.
659 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
660 the main verify for receipient in uncached-mode.
662 JH/22 Retire historical build files to an "unsupported" subdir. These are
663 defined as "ones for which we have no current evidence of testing".
665 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
666 if present. Previously it was ignored.
668 JH/24 Start using specified-initialisers in C structure init coding. This is
669 a C99 feature (it's 2017, so now considered safe).
671 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
672 if was a fixed-sized field and bitmask ops via macros; it is now more
675 PP/07 GitHub PR 56: Apply MariaDB build fix.
676 Patch provided by Jaroslav Škarvada.
678 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
679 during Coverity cleanups [4.87 JH/47]
680 Diagnosis and fix provided by Michael Fischer v. Mollard.
682 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
683 the right size to place the terminating semicolon on its own folded
684 line, the header hash was calculated to an incorrect value thanks to
685 the (relaxed) space the fold became.
687 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
688 and confused the parent.
690 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
691 which could crash as a result. This could lead to undeliverable messages.
693 JH/28 Logging: "next input sent too soon" now shows where input was truncated
696 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
697 matters on fast-turnover and PID-randomising systems, which were getting
698 out-of-order delivery.
700 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
701 a possibly-overlapping copy. The symptom was that "Remote host closed
702 connection in response to HELO" was logged instead of the actual 4xx
705 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
706 Previously only that bufferd was discarded, resulting in SYMTP command
709 JH/32 DKIM: when a message has multiple signatures matching an identity given
710 in dkim_verify_signers, run the dkim acl once for each. Previously only
711 one run was done. Bug 2189.
713 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
714 "panic the current process" to "deliberately defer". The panic log is
715 still written with the problem list name; the mail and reject logs now
716 get a temp-reject line for the message that was being handled, saying
717 something like "domains check lookup or other defer". The SMTP 451
718 message is still "Temporary local problem".
720 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
721 A crafted sequence of BDAT commands could result in in-use memory beeing
722 freed. CVE-2017-16943.
724 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
725 from SMTP input. Previously it was always done; now only done for DATA
726 and not BDAT commands. CVE-2017-16944.
728 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
729 to the message (such as an overlong header line). Previously this was
730 not done and we did not exit BDAT mode. Followon from the previous item
731 though a different problem.
737 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
738 than -2003 did; needs libidn2 in addition to libidn.
740 JH/02 The path option on a pipe transport is now expanded before use.
742 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
743 Patch provided by "Björn", documentation fix added too.
745 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
746 missing a wire-to-host endian conversion.
748 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
749 close after a BDAT command line could be taken as a following command,
750 giving a synch failure. Fix by only checking for synch immediately
751 before acknowledging the chunk.
753 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
754 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
755 macro. Patches provided by Josh Soref.
757 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
758 Previously we did not; the RFC seems ambiguous and VRFY is not listed
759 by IANA as a service extension. However, John Klensin suggests that we
762 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
763 the dkim code may be unix-mode line endings rather than smtp wire-format
764 CRLF, so prepend a CR to any bare LF.
766 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
767 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
769 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
770 the first were themselves being wrongly included in the feed into dkim
771 processing; with most chunk sizes in use this resulted in an incorrect
772 body hash calculated value.
774 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
775 DKIM signature block, for verification. Although advised against by
776 standards it is specifically not ruled illegal.
778 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
780 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
781 missing a body hash (the bh= tag).
783 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
784 It seems that HAProxy sends the Proxy Protocol information in clear and
785 only then does a TLS startup, so do the same.
787 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
788 TCP connections (such as for Spamd) unless the daemon successfully set
789 Fast Open mode on its listening sockets. This fixes breakage seen on
790 too-old kernels or those not configured for Fast Open, at the cost of
791 requiring both directions being enabled for TFO, and TFO never being used
792 by non-daemon-related Exim processes.
794 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
795 endings, at least on the first header line. Try to canonify any that get
796 past that check, despite the cost.
798 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
799 now limited to an arbitrary five deep, while parsing addresses with the
800 strip_excess_angle_brackets option enabled.
802 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
803 instead leave the unprompted TLS handshake in socket buffer for the
804 TLS library to consume.
806 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
808 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
810 JH/16 Drop variables when they go out of scope. Memory management drops a whole
811 region in one operation, for speed, and this leaves assigned pointers
812 dangling. Add checks run only under the testsuite which checks all
813 variables at a store-reset and panics on a dangling pointer; add code
814 explicitly nulling out all the variables discovered. Fixes one known
815 bug: a transport crash, where a dangling pointer for $sending_ip_address
816 originally assigned in a verify callout, is re-used.
818 PP/06 Drop '.' from @INC in various Perl scripts.
820 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
822 PP/08 Reduce a number of compilation warnings under clang; building with
823 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
824 should be warning-free.
826 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
828 HS/01 Fix portability problems introduced by PP/08 for platforms where
829 realloc(NULL) is not equivalent to malloc() [SunOS et al].
831 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
832 chunk. This allows us to accept broken chunked messages. We need a more
833 general solution here.
835 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
836 already-broken messages in the queue.
838 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
840 JH/19 Fix reference counting bug in routing-generated-address tracking.
846 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
847 supports it and a size is available (ie. the sending peer gave us one).
849 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
850 years of being deprecated). The replacements are the ACLs
851 acl_smtp_mime and acl_not_smtp_mime.
853 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
854 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
855 or even an in-clear connection were permitted. Now, if the host lookup
856 was dnssec and dane was requested then the host is only used if the
857 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
858 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
859 if one fails this test.
860 This means that a poorly-configured remote DNS will make it incommunicado;
861 but it protects against a DNS-interception attack on it.
863 JH/04 Bug 1810: make continued-use of an open smtp transport connection
864 non-noisy when a race steals the message being considered.
866 JH/05 If main configuration option tls_certificate is unset, generate a
867 self-signed certificate for inbound TLS connections.
869 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
870 in rewrites and routers.
872 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
873 and logged a warning sing 4.83; now they are a configuration file error.
875 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
876 (lacking @domain). Apply the same qualification processing as RCPT.
878 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
880 JH/10 Support ${sha256:} applied to a string (as well as the previous
883 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
884 a cutthrough deliver is pending, as we always want to make a connection.
885 This also avoids re-routing the message when later placing the cutthrough
886 connection after a verify cache hit.
887 Do not update it with the verify result either.
889 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
890 when routing results in more than one destination address.
892 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
893 signing (which inhibits the cutthrough capability). Previously only
894 the presence of an option was tested; now an expansion evaluating as
895 empty is permissible (obviously it should depend only on data available
896 when the cutthrough connection is made).
898 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
899 the relevant preceding SMTP command did not note the pipelining mode.
901 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
902 Previously they were not counted.
904 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
905 as one having no matching records. Previously we deferred the message
906 that needed the lookup.
908 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
909 distinguished as "(=".
911 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
912 for missing MX records. Previously it only worked for missing A records.
914 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
916 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
917 after the data-go-ahead and data-ack. Patch from Jason Betts.
919 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
920 even for a "none" policy. Patch from Tony Meyer.
922 JH/22 Fix continued use of a connection for further deliveries. If a port was
923 specified by a router, it must also match for the delivery to be
926 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
927 When one of the recipients of a message was unsuitable for the connection
928 (has no matching addresses), we lost track of needing to mark it
929 deferred. As a result mail would be lost.
931 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
933 JH/25 Decoding ACL controls is now done using a binary search; the source code
934 takes up less space and should be simpler to maintain. Merge the ACL
935 condition decode tables also, with similar effect.
937 JH/26 Fix problem with one_time used on a redirect router which returned the
938 parent address unchanged. A retry would see the parent address marked as
939 delivered, so not attempt the (identical) child. As a result mail would
942 JH/27 Fix a possible security hole, wherein a process operating with the Exim
943 UID can gain a root shell. Credit to http://www.halfdog.net/ for
944 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
947 JH/28 Enable {spool,log} filesystem space and inode checks as default.
948 Main config options check_{log,spool}_{inodes,space} are now
949 100 inodes, 10MB unless set otherwise in the configuration.
951 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
952 Previously it only applied to the main-section connection policy
955 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
957 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
958 by me. Added RFC7919 DH primes as an alternative.
960 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
961 are not in the system include path.
963 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
964 GnuTLS, when a session startup failed (eg because the client disconnected)
965 Exim did stdio operations after fclose. This was exposed by a recent
966 change which nulled out the file handle after the fclose.
968 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
969 signed directly by the cert-signing cert, rather than an intermediate
970 OCSP-signing cert. This is the model used by LetsEncrypt.
972 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
974 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
975 an incoming connection.
977 HS/02 Bug 1802: Do not half-close the connection after sending a request
980 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
981 fallback to "prime256v1".
983 JH/34 SECURITY: Use proper copy of DATA command in error message.
984 Could leak key material. Remotely exploitable. CVE-2016-9963.
990 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
991 and 3.4.4 - once the server is enabled to respond to an OCSP request
992 it does even when not requested, resulting in a stapling non-aware
993 client dropping the TLS connection.
995 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
996 support variable-length bit vectors. No functional change.
998 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
999 The I= interface field on outgoing lines is now after the H= remote
1000 host field, same as incoming lines. There is a separate
1001 outgoing_interface log selector which allows you to disable the
1004 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1005 If not running log_selector +smtp_connection the mainlog would be held
1006 open indefinitely after a "too many connections" event, including to a
1007 deleted file after a log rotate. Leave the per net connection logging
1008 leaving it open for efficiency as that will be quickly detected by the
1009 check on the next write.
1011 HS/01 Bug 1671: Fix post transport crash.
1012 Processing the wait-<transport> messages could crash the delivery
1013 process if the message IDs didn't exist for some reason. When
1014 using 'split_spool_directory=yes' the construction of the spool
1015 file name failed already, exposing the same netto behaviour.
1017 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1018 mime_regex ACL conditions.
1020 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1021 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1022 to HELO, local diagnostic string.
1024 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1025 log line to debug. Even when configured with a tls authenticator many
1026 client connections are expected to not authenticate in this way, so
1027 an authenticate fail is not an error.
1029 HS/02 Add the Exim version string to the process info. This way exiwhat
1030 gives some more detail about the running daemon.
1032 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1033 matter for fast-change records such as DNSBLs.
1035 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1036 retry record, even if constant. There may be multiple transports with
1037 different interface settings and the retry behaviour needs to be kept
1040 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1042 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1044 JH/10 Bug 840: fix log_defer_output option of pipe transport
1046 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1047 hosts_randomize. This matters a lot when combined with mua_wrapper.
1049 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1050 ${quote_pgsql:<string>} operator.
1052 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1054 JH/14 Change header file name for freeradius-client. Relevant if compiling
1055 with Radius support; from the Gentoo tree and checked under Fedora.
1057 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1059 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1060 option transport_filter, meaning no filtering.
1062 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1064 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1065 defaults to "*" (all hosts). The variable is now available when not built
1066 with TLS, default unset, mainly to enable keeping the testsuite sane.
1067 If a server certificate is not supplied (via tls_certificate) an error is
1068 logged, and clients will find TLS connections fail on startup. Presumably
1069 they will retry in-clear.
1070 Packagers of Exim are strongly encouraged to create a server certificate
1071 at installation time.
1073 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1074 with the $config_file variable.
1076 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1077 in transport context, after the attempt, and per-recipient. The latter type
1078 is per host attempted. The event data is the error message, and the errno
1079 information encodes the lookup type (A vs. MX) used for the (first) host,
1080 and the trailing two digits of the smtp 4xx response.
1082 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1083 to write to mainlog (or rejectlog, paniclog) in the window between file
1084 creation and permissions/ownership being changed. Particularly affects
1085 installations where exicyclog is run as root, rather than exim user;
1086 result is that the running daemon panics and dies.
1088 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1090 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1091 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1092 "pri" and "weight". Note that the previous implicit priority given by the
1093 list order is no longer honoured.
1095 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1096 for DKIM processing.
1098 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1099 by defining SUPPORT_SOCKS.
1101 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1102 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1103 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1104 variables are renamed to proxy_{local,external}_{address,port}.
1106 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1107 for a build by defining SUPPORT_I18N
1109 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1110 of the query string, and make ${quote_redis:} do that quoting.
1112 JH/29 Move Events support from Experimental to mainline, enabled by default
1113 and removable for a build by defining DISABLE_EVENT.
1115 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1117 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1118 cached by the daemon.
1120 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1121 by defining LOOKUP_REDIS. The libhiredis library is required.
1123 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1124 keys are given for lookup.
1126 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1127 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1128 only supported when built with TLS support. The PolarSSL SHA routines
1129 are still used when the TLS library is too old for convenient support.
1131 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1132 openssl_options), for security. OpenSSL forces this from version 1.1.0
1133 server-side so match that on older versions.
1135 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1136 allocation for $value could be released as the expansion processing
1137 concluded, but leaving the global pointer active for it.
1139 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1140 and to use the domains and local_parts ACL conditions.
1142 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1143 incorrectly not doubled on cutthrough transmission, hence seen as a
1144 body-termination at the receiving system - resulting in truncated mails.
1145 Commonly the sender saw a TCP-level error, and retransmitted the message
1146 via the normal store-and-forward channel. This could result in duplicates
1147 received - but deduplicating mailstores were liable to retain only the
1148 initial truncated version.
1150 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1152 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1154 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1155 we're in there, support oversigning also; bug 1309.
1157 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1159 HS/04 Add support for keep_environment and add_environment options.
1161 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1162 either intentional arithmetic overflow during PRNG, or testing config-
1165 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1166 delivery resulted in actual delivery. Cancel cutthrough before DATA
1169 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1170 rejects a recipient: pass the reject to the originator.
1172 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1173 Many were false-positives and ignorable, but it's worth fixing the
1176 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1177 for the new environment-manipulation done at startup. Move the routines
1178 from being local to tls.c to being global via the os.c file.
1180 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1181 an extract embedded as result-arg for a map, the first arg for extract
1182 is unavailable so we cannot tell if this is a numbered or keyed
1183 extraction. Accept either.
1189 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1192 JH/02 The smtp transport option "multi_domain" is now expanded.
1194 JH/03 The smtp transport now requests PRDR by default, if the server offers
1197 JH/04 Certificate name checking on server certificates, when exim is a client,
1198 is now done by default. The transport option tls_verify_cert_hostnames
1199 can be used to disable this per-host. The build option
1200 EXPERIMENTAL_CERTNAMES is withdrawn.
1202 JH/05 The value of the tls_verify_certificates smtp transport and main options
1203 default to the word "system" to access the system default CA bundle.
1204 For GnuTLS, only version 3.0.20 or later.
1206 JH/06 Verification of the server certificate for a TLS connection is now tried
1207 (but not required) by default. The verification status is now logged by
1208 default, for both outbound TLS and client-certificate supplying inbound
1211 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1214 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1215 Status Notification (bounce) messages are now MIME format per RFC 3464.
1216 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1217 under the control of the dsn_advertise_hosts option, and routers may
1218 have a dsn_lasthop option.
1220 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1221 default, modifiable by a malware= option. The list separator for
1222 the options can now be changed in the usual way. Bug 68.
1224 JH/10 The smtp_receive_timeout main option is now expanded before use.
1226 JH/11 The incoming_interface log option now also enables logging of the
1227 local interface on delivery outgoing connections.
1229 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1230 if the interface and destination host and port all match.
1232 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1235 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1236 Patch from Andrew Lewis.
1238 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1239 now supports optional time-restrictions, weighting, and priority
1240 modifiers per server. Patch originally by <rommer@active.by>.
1242 JH/16 The spamd_address main option now supports a mixed list of local
1243 and remote servers. Remote servers can be IPv6 addresses, and
1244 specify a port-range.
1246 JH/17 Bug 68: The spamd_address main option now supports an optional
1247 timeout value per server.
1249 JH/18 Bug 1581: Router and transport options headers_add/remove can
1250 now have the list separator specified.
1252 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1255 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1258 JH/21 Support for the A6 type of dns record is withdrawn.
1260 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1261 rather than the verbs used.
1263 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1264 from 255 to 1024 chars.
1266 JH/24 Verification callouts now attempt to use TLS by default.
1268 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1269 are generic router options now. The defaults didn't change.
1271 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1272 Original patch from Alexander Shikoff, worked over by JH.
1274 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1275 files if ZCAT_COMMAND is not executable.
1277 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1279 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1281 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1282 Normally benign, it bites when the pair was led to by a CNAME;
1283 modern usage is to not canonicalize the domain to a CNAME target
1284 (and we were inconsistent anyway for A-only vs AAAA+A).
1286 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1288 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1289 when evaluating $sender_host_dnssec.
1291 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1292 $sender_helo_dnssec variable.
1294 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1296 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1298 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1300 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1301 documented as working, but never had. Support all but $spam_report.
1303 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1304 added for tls authenticator.
1306 HS/03 Add perl_taintmode main config option
1312 TL/01 When running the test suite, the README says that variables such as
1313 no_msglog_check are global and can be placed anywhere in a specific
1314 test's script, however it was observed that placement needed to be near
1315 the beginning for it to behave that way. Changed the runtest perl
1316 script to read through the entire script once to detect and set these
1317 variables, reset to the beginning of the script, and then run through
1318 the script parsing/test process like normal.
1320 TL/02 The BSD's have an arc4random API. One of the functions to induce
1321 adding randomness was arc4random_stir(), but it has been removed in
1322 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1323 function when detected.
1325 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1326 cause callback expansion.
1328 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1329 syntax errors in an expansion can be treated as a string instead of
1330 logging or causing an error, due to the internal use of bool_lax
1331 instead of bool when processing it.
1333 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1334 server certificates when making smtp deliveries.
1336 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1338 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1340 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1342 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1343 Merged patch from Sebastian Wiedenroth.
1345 JH/05 Fix results-pipe from transport process. Several recipients, combined
1346 with certificate use, exposed issues where response data items split
1347 over buffer boundaries were not parsed properly. This eventually
1348 resulted in duplicates being sent. This issue only became common enough
1349 to notice due to the introduction of connection certificate information,
1350 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1352 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1353 size buffer was used, resulting in syntax errors when an expansion
1356 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1357 version 3.3.6 or later.
1359 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1360 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1361 both become "event_action", the variables become $event_name, $event_data
1362 and $event_defer_errno. There is a new variable $verify_mode, usable in
1363 routers, transports and related events. The tls:cert event is now also
1364 raised for inbound connections, if the main configuration event_action
1367 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1368 early OCSP support, but no stapling (appears to be less than 1.0.0).
1370 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1371 server certificate names available under the smtp transport option
1372 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1375 JH/10 Time-related extraction expansions from certificates now use the main
1376 option "timezone" setting for output formatting, and are consistent
1377 between OpenSSL and GnuTLS compilations. Bug 1541.
1379 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1380 encoded parameter in the incoming message. Bug 1558.
1382 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1383 include certificate info, eximon was claiming there were spoolfile
1386 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1388 JH/14 Log delivery-related information more consistently, using the sequence
1389 "H=<name> [<ip>]" wherever possible.
1391 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1392 are problematic for Debian distribution, omit them from the release
1395 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1397 JH/16 Fix string representation of time values on 64bit time_t architectures.
1400 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1401 output list separator was used.
1406 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1407 checkers that were complaining about end of non-void function with no
1410 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1411 This was a regression introduced in 4.83 by another bugfix.
1413 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1415 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1416 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1422 TF/01 Correctly close the server side of TLS when forking for delivery.
1424 When a message was received over SMTP with TLS, Exim failed to clear up
1425 the incoming connection properly after forking off the child process to
1426 deliver the message. In some situations the subsequent outgoing
1427 delivery connection happened to have the same fd number as the incoming
1428 connection previously had. Exim would try to use TLS and fail, logging
1429 a "Bad file descriptor" error.
1431 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1432 utilities have not been installed.
1434 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1435 temporary space as the ACL may create new global variables.
1437 TL/01 LDAP support uses per connection or global context settings, depending
1438 upon the detected version of the libraries at build time.
1440 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1441 to extract and use the src ip:port in logging and expansions as if it
1442 were a direct connection from the outside internet. PPv2 support was
1443 updated based on HAProxy spec change in May 2014.
1445 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1447 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1448 Properly escape header and check for NULL return.
1450 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1453 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1455 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1456 characters in header names, implemented as a verify condition.
1457 Contributed by Michael Fischer v. Mollard.
1459 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1460 results permerror and temperror. Previous values are deprecated but
1461 still accepted. In a future release, err_perm and err_temp will be
1462 completely removed, which will be a backward incompatibility if the
1463 ACL tests for either of these two old results. Patch contributed by
1464 user bes-internal on the mailing list.
1466 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1468 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1469 selectors, in both main and reject logs.
1471 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1474 JH/07 Add malware type "sock" for talking to simple daemon.
1476 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1478 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1479 routers/transports under cutthrough routing.
1481 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1482 numbers. Touch up "bool" conditional to keep the same definition.
1484 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1486 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1488 JH/12 Expand items in router/transport headers_add or headers_remove lists
1489 individually rather than the list as a whole. Bug 1452.
1491 Required for reasonable handling of multiple headers_ options when
1492 they may be empty; requires that headers_remove items with embedded
1493 colons must have them doubled (or the list-separator changed).
1495 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1496 view the policy declared in the DMARC record. Currently, $dmarc_status
1497 is a combined value of both the record presence and the result of the
1500 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1502 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1503 dnslookup router and the smtp transport (applying to the forward
1506 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1507 of ldap servers used for a specific lookup. Patch provided by Heiko
1510 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1511 New variable $lookup_dnssec_authenticated for observability.
1513 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1514 Patch submitted by Lars Timman.
1516 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1518 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1519 Requires trusted mode and valid format message id, aborts otherwise.
1520 Patch contributed by Heiko Schlichting.
1522 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1523 certextract with support for various fields. Bug 1358.
1525 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1526 is requested by default, modifiable by smtp transport option
1529 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1530 operate on certificate variables to give certificate fingerprints
1531 Also new ${sha256:cert_variable}.
1533 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1535 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1536 Christian Aistleitner.
1538 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1540 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1541 file. Patch from Wolfgang Breyha.
1543 JH/25 Expand the coverage of the delivery $host and $host_address to
1544 client authenticators run in verify callout. Bug 1476.
1546 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1547 align with daemon_smtp_ports. Bug 72.
1549 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1550 support and error reporting did not work properly.
1552 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1553 and is readable. Patch from Andrew Colin Kissa.
1555 TL/14 Enhance documentation of ${run expansion and how it parses the
1556 commandline after expansion, particularly in the case when an
1557 unquoted variable expansion results in an empty value.
1559 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1561 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1562 by renaming away from it. Addresses GH issue 15, reported by
1565 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1567 TL/15 SECURITY: prevent double expansion in math comparison functions
1568 (can expand unsanitized data). Not remotely exploitable.
1575 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1577 PP/02 Make -n do something, by making it not do something.
1578 When combined with -bP, the name of an option is not output.
1580 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1583 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1584 $sender_host_name and config options to manage this, and basic check
1587 PP/05 DSCP support for outbound connections and control modifier for inbound.
1589 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1590 (Only plugin which currently uses this is kerberos4, which nobody should
1591 be using, but we should make it available and other future plugins might
1592 conceivably use it, even though it would break NAT; stuff *should* be
1593 using channel bindings instead).
1595 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1596 name; added for Sendmail compatibility; requires admin caller.
1597 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1598 just ignore it); requires trusted caller.
1599 Also parse but ignore: -Ac -Am -X<logfile>
1602 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1604 TL/02 Add +smtp_confirmation as a default logging option.
1606 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1607 Patch by Magnus Holmgren from 2007-02-20.
1609 TL/04 Bugzilla 1281 - Spec typo.
1610 Bugzilla 1283 - Spec typo.
1611 Bugzilla 1290 - Spec grammar fixes.
1613 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1615 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1617 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1618 mailing list by by Dmitry Isaikin.
1620 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1622 JH/02 Support "G" suffix to numbers in ${if comparisons.
1624 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1626 NM/01 Bugzilla 1197 - Spec typo
1627 Bugzilla 1196 - Spec examples corrections
1629 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1631 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1632 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1635 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1636 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1638 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1639 "acl {{name}{arg}...}", and optional args on acl condition
1642 JH/05 Permit multiple router/transport headers_add/remove lines.
1644 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1646 JH/07 Avoid using a waiting database for a single-message-only transport.
1647 Performance patch from Paul Fisher. Bugzilla 1262.
1649 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1652 JH/09 Add $headers_added variable, with content from use of ACL modifier
1653 add_header (but not yet added to the message). Bugzilla 199.
1655 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1656 Pulled from Bugzilla 817 by Wolfgang Breyha.
1658 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1660 (nb: this is the same fix as in Exim 4.80.1)
1662 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1665 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1668 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1670 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1671 advertise SMTP AUTH mechanism to us, instead of a generic
1672 protocol violation error. Also, make Exim more robust to bad
1673 data from the Dovecot auth socket.
1675 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1677 When a queue runner is handling a message, Exim first routes the
1678 recipient addresses, during which it prunes them based on the retry
1679 hints database. After that it attempts to deliver the message to
1680 any remaining recipients. It then updates the hints database using
1683 So if a recipient address works intermittently, it can get repeatedly
1684 deferred at routing time. The retry hints record remains fresh so the
1685 address never reaches the final cutoff time.
1687 This is a fairly common occurrence when a user is bumping up against
1688 their storage quota. Exim had some logic in its local delivery code
1689 to deal with this. However it did not apply to per-recipient defers
1690 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1692 This change adds a proper retry rule check during routing so that the
1693 final cutoff time is checked against the message's age. We only do
1694 this check if there is an address retry record and there is not a
1695 domain retry record; this implies that previous attempts to handle
1696 the address had the retry_use_local_parts option turned on. We use
1697 this as an approximation for the destination being like a local
1698 delivery, as in LMTP.
1700 I suspect this new check makes the old local delivery cutoff check
1701 redundant, but I have not verified this so I left the code in place.
1703 TF/02 Correct gecos expansion when From: is a prefix of the username.
1705 Test 0254 submits a message to Exim with the header
1709 When I ran the test suite under the user fanf2, Exim expanded
1710 the header to contain my full name, whereas it should have added
1711 a Resent-Sender: header. It erroneously treats any prefix of the
1712 username as equal to the username.
1714 This change corrects that bug.
1716 GF/01 DCC debug and logging tidyup
1717 Error conditions log to paniclog rather than rejectlog.
1718 Debug lines prefixed by "DCC: " to remove any ambiguity.
1720 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1722 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1723 Bug spotted by Jeremy Harris; was flawed since initial commit.
1724 Would have resulted in OCSP responses post-SNI triggering an Exim
1725 NULL dereference and crash.
1727 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1729 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1730 Bug detection, analysis and fix by Samuel Thibault.
1731 Bugzilla 1331, Debian bug #698092.
1733 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1735 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1736 Server implementation by Todd Lyons, client by JH.
1737 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1738 config variable "prdr_enable" controls whether the server
1739 advertises the facility. If the client requests PRDR a new
1740 acl_data_smtp_prdr ACL is called once for each recipient, after
1741 the body content is received and before the acl_smtp_data ACL.
1742 The client is controlled by both of: a hosts_try_prdr option
1743 on the smtp transport, and the server advertisement.
1744 Default client logging of deliveries and rejections involving
1745 PRDR are flagged with the string "PRDR".
1747 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1748 fclose(). Diagnosis by Todd Lyons.
1750 PP/17 Update configure.default to handle IPv6 localhost better.
1751 Patch by Alain Williams (plus minor tweaks).
1754 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1755 This is now consistent with GnuTLS, and is now documented: the
1756 previous undocumented portable approach to treating the option as
1757 unset was to force an expansion failure. That still works, and
1758 an empty string is now equivalent.
1760 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1761 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1762 not performing validation itself.
1764 PP/20 Added force_command boolean option to pipe transport.
1765 Patch from Nick Koston, of cPanel Inc.
1767 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1770 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1772 PP/21 Fix eximon continuous updating with timestamped log-files.
1773 Broken in a format-string cleanup in 4.80, missed when I repaired the
1774 other false fix of the same issue.
1775 Report and fix from Heiko Schlichting.
1778 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1779 Report from Prashanth Katuri.
1781 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1782 It's SecureTransport, so affects any MacOS clients which use the
1783 system-integrated TLS libraries, including email clients.
1785 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1786 using a MIME ACL for non-SMTP local injection.
1787 Report and assistance in diagnosis by Warren Baker.
1789 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1791 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1793 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1794 last id that failed so it may be referenced in subsequent ACL's.
1796 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1799 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1800 ldap library initialization, allowing self-signed CA's to be
1801 used. Also properly sets require_cert option later in code by
1802 using NULL (global ldap config) instead of ldap handle (per
1803 session). Bug diagnosis and testing by alxgomz.
1805 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1806 the src/util/ subdirectory.
1808 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1809 renamed to Transport Post Delivery Action by Jeremy Harris, as
1812 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1813 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1814 redis_servers = needs to be configured which will be used by the redis
1815 lookup. Patch from Warren Baker, of The Packet Hub.
1817 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1819 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1820 hostname or reverse DNS when processing a host list. Used suggestions
1821 from multiple comments on this bug.
1823 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1825 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1826 Missed a few lines, added it to make the runtest require no keyboard
1829 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1830 contains upper case chars. Make router use caseful_local_part.
1832 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1833 support when GnuTLS has been built with p11-kit.
1839 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1841 This, or similar/improved, will also be change PP/11 of 4.82.
1847 PP/01 Handle short writes when writing local log-files.
1848 In practice, only affects FreeBSD (8 onwards).
1849 Bugzilla 1053, with thanks to Dmitry Isaikin.
1851 NM/01 Bugzilla 949 - Documentation tweak
1853 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1856 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1858 PP/02 Implemented gsasl authenticator.
1860 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1862 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1863 `pkg-config foo` for cflags/libs.
1865 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1866 with rest of GSASL and with heimdal_gssapi.
1868 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1869 `pkg-config foo` for cflags/libs for the TLS implementation.
1871 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1872 properties get this fed in as external SSF. A number of robustness
1873 and debugging improvements to the cyrus_sasl authenticator.
1875 PP/08 cyrus_sasl server now expands the server_realm option.
1877 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1878 Patch by Jeremy Harris.
1880 PP/10 Added dbmjz lookup type.
1882 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1884 PP/12 MAIL args handles TAB as well as SP, for better interop with
1885 non-compliant senders.
1886 Analysis and variant patch by Todd Lyons.
1888 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1889 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1890 Patch from Dirk Mueller <dmueller@suse.com>
1892 PP/13 tls_peerdn now print-escaped for spool files.
1893 Observed some $tls_peerdn in wild which contained \n, which resulted
1894 in spool file corruption.
1896 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1897 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1898 or write after TLS renegotiation, which otherwise led to messages
1901 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1902 as a tracking header (ie: a signed header comes before the signature).
1903 Patch from Wolfgang Breyha.
1905 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1906 comma-sep list; embedded commas doubled.
1908 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1910 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1912 Report and patch from Dmitry Banschikov.
1914 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1915 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1916 failures. We appear to now support TLS1.1+ with Exim.
1918 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1919 lets Exim select keys and certificates based upon TLS SNI from client.
1920 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1921 before an outbound SMTP session. New log_selector, +tls_sni.
1923 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1924 NULL dereference. Report and patch from Alun Jones.
1926 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1927 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1928 Not seeing resolver debug output on NetBSD, but suspect this is a
1929 resolver implementation change.
1931 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1932 Left warnings. Added "eximon gdb" invocation mode.
1934 PP/21 Defaulting "accept_8bitmime" to true, not false.
1936 PP/22 Added -bw for inetd wait mode support.
1938 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1939 locate the relevant includes and libraries. Made this the default.
1941 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1942 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1944 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1945 This may cause build issues on older platforms.
1947 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1948 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1949 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1950 Added SNI support via GnuTLS too.
1951 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1953 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1955 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1956 Applied second patch from Janne, implementing suggestion to default
1957 multiple-strings-in-record handling to match SPF spec.
1959 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1961 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1962 read-only, out of scope).
1963 Patch from Wolfgang Breyha, report from Stuart Northfield.
1965 PP/29 Fix three issues highlighted by clang analyser static analysis.
1966 Only crash-plausible issue would require the Cambridge-specific
1967 iplookup router and a misconfiguration.
1968 Report from Marcin Mirosław.
1970 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1972 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1973 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1974 As part of this, removing so much warning spew let me fix some minor
1975 real issues in debug logging.
1977 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1978 assignment on my part. Fixed.
1980 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1981 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1982 Janne Snabb (who went above and beyond: thank you).
1984 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1985 string otherwise requires a connection and a bunch more work and it's
1986 relatively easy to get wrong. Should also expose TLS library linkage
1989 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1990 64-bit ${eval} (JH/03).
1992 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1993 GNU libc to support some of the 64-bit stuff, should not lead to
1994 conflicts. Defined before os.h is pulled in, so if a given platform
1995 needs to override this, it can.
1997 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1998 protection layer was required, which is not implemented.
1999 Bugzilla 1254, patch from Wolfgang Breyha.
2001 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2002 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2003 tls_dhparam take prime identifiers. Also unbreak combination of
2004 OpenSSL+DH_params+TLSSNI.
2006 PP/39 Disable SSLv2 by default in OpenSSL support.
2012 PP/01 Solaris build fix for Oracle's LDAP libraries.
2013 Bugzilla 1109, patch from Stephen Usher.
2015 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2017 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2020 TF/02 Fix a couple more cases where we did not log the error message
2021 when unlink() failed. See also change 4.74-TF/03.
2023 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2024 lock up or crash if it happened to be inside a call to libc when it
2025 got a SIGUSR1 from exiwhat.
2027 The SIGUSR1 handler appends the current process status to the process
2028 log which is later printed by exiwhat. It used to use the general
2029 purpose logging code to do this, but several functions it calls are
2030 not safe for signals.
2032 The new output code in the SIGUSR1 handler is specific to the process
2033 log, and simple enough that it's easy to inspect for signal safety.
2034 Removing some special cases also simplifies the general logging code.
2035 Removing the spurious timestamps from the process log simplifies
2038 TF/04 Improved ratelimit ACL condition.
2040 The /noupdate option has been deprecated in favour of /readonly which
2041 has clearer semantics. The /leaky, /strict, and /readonly update modes
2042 are mutually exclusive. The update mode is no longer included in the
2043 database key; it just determines when the database is updated. (This
2044 means that when you upgrade Exim will forget old rate measurements.)
2046 Exim now checks that the per_* options are used with an update mode that
2047 makes sense for the current ACL. For example, when Exim is processing a
2048 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2049 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2050 must specify per_mail/readonly. If you omit the update mode it defaults to
2051 /leaky where that makes sense (as before) or /readonly where required.
2053 The /noupdate option is now undocumented but still supported for
2054 backwards compatibility. It is equivalent to /readonly except that in
2055 ACLs where /readonly is required you may specify /leaky/noupdate or
2056 /strict/noupdate which are treated the same as /readonly.
2058 A useful new feature is the /count= option. This is a generalization
2059 of the per_byte option, so that you can measure the throughput of other
2060 aggregate values. For example, the per_byte option is now equivalent
2061 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2063 The per_rcpt option has been generalized using the /count= mechanism
2064 (though it's more complicated than the per_byte equivalence). When it is
2065 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2066 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2067 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2068 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2069 Note that using per_rcpt with a non-readonly update mode in more than
2070 one ACL will cause the recipients to be double-counted. (The per_mail
2071 and per_byte options don't have this problem.)
2073 The handling of very low rates has changed slightly. If the computed rate
2074 is less than the event's count (usually one) then this event is the first
2075 after a long gap. In this case the rate is set to the same as this event's
2076 count, so that the first message of a spam run is counted properly.
2078 The major new feature is a mechanism for counting the rate of unique
2079 events. The new per_addr option counts the number of different
2080 recipients that someone has sent messages to in the last time period. It
2081 behaves like per_rcpt if all the recipient addresses are different, but
2082 duplicate recipient addresses do not increase the measured rate. Like
2083 the /count= option this is a general mechanism, so the per_addr option
2084 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2085 example, measure the rate that a client uses different sender addresses
2086 with the options per_mail/unique=$sender_address. There are further
2087 details in the main documentation.
2089 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2091 TF/06 Removed a few PCRE remnants.
2093 TF/07 Automatically extract Exim's version number from tags in the git
2094 repository when doing development or release builds.
2096 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2097 Bugzilla 879. Patch from Paul Fisher.
2099 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2100 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2103 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2105 PP/05 Variable $av_failed, true if the AV scanner deferred.
2106 Bugzilla 1078. Patch from John Horne.
2108 PP/06 Stop make process more reliably on build failure.
2109 Bugzilla 1087. Patch from Heiko Schlittermann.
2111 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2112 Bugzilla 1089. Patch from Heiko Schlittermann.
2114 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2115 Bugzilla 1131. Patch from Holger Weiß.
2117 PP/09 Handle IPv6 addresses with SPF.
2118 Bugzilla 860. Patch from Wolfgang Breyha.
2120 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2122 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2125 PP/11 match_* no longer expand right-hand-side by default.
2126 New compile-time build option, EXPAND_LISTMATCH_RHS.
2127 New expansion conditions, "inlist", "inlisti".
2129 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2131 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2133 PP/14 fix log_write() format string regression from TF/03.
2134 Bugzilla 1152. Patch from Dmitry Isaikin.
2140 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2142 PP/02 Harmonised TLS library version reporting; only show if debugging.
2143 Layout now matches that introduced for other libraries in 4.74 PP/03.
2145 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2147 PP/04 New "dns_use_edns0" global option.
2149 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2152 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2153 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2155 TK/01 Updated PolarSSL code to 0.14.2.
2156 Bugzilla 1097. Patch from Andreas Metzler.
2158 PP/07 Catch divide-by-zero in ${eval:...}.
2159 Fixes bugzilla 1102.
2161 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2164 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2165 format-string attack -- SECURITY: remote arbitrary code execution.
2167 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2168 time unintentionally subject to list matching rules, letting the header
2169 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2170 arbitrary string expansion). This allowed for information disclosure.
2172 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2173 INT_MIN/-1 -- value coerced to INT_MAX.
2179 NM/01 Workaround for PCRE version dependency in version reporting
2182 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2183 This fixes portability to compilers other than gcc, notably
2184 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2186 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2187 makefiles for portability to HP-UX and POSIX correctness.
2189 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2190 Also via indented variable definition in the Makefile.
2191 (Debugging by Oliver Heesakkers).
2193 PP/02 Restore caching of spamd results with expanded spamd_address.
2194 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2196 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2197 Improves build reliability. Fix from: Frank Elsner
2199 NM/02 Fix wide character breakage in the rfc2047 coding
2200 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2202 NM/03 Allow underscore in dnslist lookups
2203 Fixes bug 1026. Patch from Graeme Fowler
2205 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2206 Code patches from Adam Ciarcinski of NetBSD.
2208 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2211 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2212 is logged, to avoid truncation. Patch from John Horne.
2214 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2215 Patch from Jakob Hirsch.
2217 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2218 SQL string expansion failure details.
2219 Patch from Andrey Oktyabrski.
2221 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2222 Patch from Simon Arlott.
2224 PP/09 New lookups functionality failed to compile on old gcc which rejects
2225 extern declarations in function scope.
2226 Patch from Oliver Fleischmann
2228 PP/10 Use sig_atomic_t for flags set from signal handlers.
2229 Check getgroups() return and improve debugging.
2230 Fixed developed for diagnosis in bug 927 (which turned out to be
2233 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2234 Patch from Mark Zealey.
2236 PP/12 Bugzilla 1056: Improved spamd server selection.
2237 Patch from Mark Zealey.
2239 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2240 Based on patch from Heiko Schlittermann.
2242 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2243 Patch from Uwe Doering, sign-off by Michael Haardt.
2245 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2246 variable declaration deep within a block. Bug and patch from
2249 PP/15 lookups-Makefile IRIX compatibility coercion.
2251 PP/16 Make DISABLE_DKIM build knob functional.
2253 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2254 Patch by Simon Arlott
2256 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2257 variable argument macros. Our copy now differs from upstream.
2263 TF/01 Failure to get a lock on a hints database can have serious
2264 consequences so log it to the panic log.
2266 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2267 controlled using the smtp_confirmation log selector.
2269 TF/03 Include the error message when we fail to unlink a spool file.
2271 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2272 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2273 for maintaining out-of-tree patches for some time.
2275 PP/01 Bugzilla 139: Documentation and portability issues.
2276 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2277 Handle per-OS dynamic-module compilation flags.
2279 PP/02 Let /dev/null have normal permissions.
2280 The 4.73 fixes were a little too stringent and complained about the
2281 permissions on /dev/null. Exempt it from some checks.
2282 Reported by Andreas M. Kirchwitz.
2284 PP/03 Report version information for many libraries, including
2285 Exim version information for dynamically loaded libraries. Created
2286 version.h, now support a version extension string for distributors
2287 who patch heavily. Dynamic module ABI change.
2289 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2290 privilege escalation vulnerability whereby the Exim run-time user
2291 can cause root to append content of the attacker's choosing to
2294 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2297 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2298 If dropping privileges for untrusted macros, we disabled normal logging
2299 on the basis that it would fail; for the Exim run-time user, this is not
2300 the case, and it resulted in successful deliveries going unlogged.
2301 Fixed. Reported by Andreas Metzler.
2307 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2308 only prepend for the Resent-* case. Fixes regression introduced in
2309 Exim 4.70 by NM/22 for Bugzilla 607.
2311 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2312 increasing numbers of administrators be bitten by this.
2314 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2316 PP/03 Bugzilla 994: added openssl_options main configuration option.
2318 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2320 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2322 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2324 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2325 without a peer certificate, leading to a segfault because of an
2326 assumption that peers always have certificates. Be a little more
2327 paranoid. Problem reported by Martin Tscholak.
2329 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2330 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2331 NB: ClamAV planning to remove STREAM in "middle of 2010".
2332 CL also introduces -bmalware, various -d+acl logging additions and
2333 more caution in buffer sizes.
2335 PP/09 Implemented reverse_ip expansion operator.
2337 PP/10 Bugzilla 937: provide a "debug" ACL control.
2339 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2341 PP/12 Bugzilla 973: Implement --version.
2343 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2345 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2347 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2349 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2350 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2351 ignore trailing whitespace.
2353 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2355 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2358 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2359 Notification from Dr Andrew Aitchison.
2361 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2362 ExtendedDetectionInfo response format.
2363 Notification from John Horne.
2365 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2368 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2369 XSL and documented dependency on system catalogs, with examples of how
2372 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2375 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2376 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2377 configuration file which is writeable by the Exim user or group.
2379 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2380 of configuration files to cover files specified with the -C option if
2381 they are going to be used with root privileges, not just the default
2384 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2385 option (effectively making it always true).
2387 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2388 files to be used while preserving root privileges.
2390 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2391 that rogue child processes cannot use them.
2393 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2394 run-time user, instead of root.
2396 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2397 Exim run-time user without dropping privileges.
2399 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2400 result string, instead of calling string_vformat() twice with the same
2403 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2404 for other users. Others should always drop root privileges if they use
2405 -C on the command line, even for a whitelisted configure file.
2407 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2409 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2415 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2416 $data_path, and $header_path variables; fixed documentation bugs and
2419 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2420 exipick to access non-standard spools, including the "frozen" queue
2423 NM/01 Bugzilla 965: Support mysql stored procedures.
2424 Patch from Alain Williams
2426 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2428 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2429 Patch from Andreas Metzler
2431 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2432 Patch from Kirill Miazine
2434 NM/05 Bugzilla 671: Added umask to procmail example.
2436 JJ/03 installed exipick 20100323.0, fixing doc bug
2438 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2439 directory. Notification and patch from Dan Rosenberg.
2441 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2443 TK/02 Improve log output when DKIM signing operation fails.
2445 MH/01 Treat the transport option dkim_domain as a colon separated
2446 list, not as a single string, and sign the message with each element,
2447 omitting multiple occurences of the same signer.
2449 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2450 Bugzilla 985, 986. Patch by Simon Arlott
2452 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2453 Patch by Simon Arlott
2455 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2456 MBX locking. Notification from Dan Rosenberg.
2462 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2464 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2466 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2468 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2470 NM/04 Bugzilla 915: Flag broken perl installation during build.
2476 TK/01 Added patch by Johannes Berg that expands the main option
2477 "spamd_address" if it starts with a dollar sign.
2479 TK/02 Write list of recipients to X-Envelope-Sender header when building
2480 the mbox-format spool file for content scanning (suggested by Jakob
2483 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2484 (http://www.dcc-servers.net/) support via dccifd. Activated by
2485 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2487 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2488 by Mark Daniel Reidel <mr@df.eu>.
2490 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2491 When building exim an external PCRE library is now needed -
2492 PCRE is a system library on the majority of modern systems.
2493 See entry on PCRE_LIBS in EDITME file.
2495 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2496 conversation. Added nologin parameter to request.
2497 Patch contributed by Kirill Miazine.
2499 TF/01 Do not log submission mode rewrites if they do not change the address.
2501 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2503 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2504 log files in place. Contributed by Roberto Lima.
2506 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2508 TF/03 Bugzilla 615: When checking the local_parts router precondition
2509 after a local_part_suffix or local_part_prefix option, Exim now
2510 does not use the address's named list lookup cache, since this
2511 contains cached lookups for the whole local part.
2513 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2514 Robert Millan. Documentation is in experimental-spec.txt.
2516 TF/04 Bugzilla 668: Fix parallel build (make -j).
2518 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2520 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2521 Patch provided by Jan Srzednicki.
2523 TF/05 Leading white space used to be stripped from $spam_report which
2524 wrecked the formatting. Now it is preserved.
2526 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2527 that they are available at delivery time.
2529 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2531 TF/08 TLS error reporting now respects the incoming_interface and
2532 incoming_port log selectors.
2534 TF/09 Produce a more useful error message if an SMTP transport's hosts
2535 setting expands to an empty string.
2537 NM/06 Bugzilla 744: EXPN did not work under TLS.
2538 Patch provided by Phil Pennock.
2540 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2541 Patch provided by Richard Godbee.
2543 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2544 acl_smtp_notquit, added index entry.
2546 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2547 Patch provided by Eugene Bujak.
2549 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2550 accept(). Patch provided by Maxim Dounin.
2552 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2553 Patch provided by Phil Pennock.
2555 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2557 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2558 Patch provided by Brad "anomie" Jorsch.
2560 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2561 Patch provided by Dean Brooks.
2563 TK/05 Add native DKIM support (does not depend on external libraries).
2565 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2566 Patch provided by Graeme Fowler.
2568 NM/16 Bugzilla 851: Documentation example syntax fix.
2570 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2572 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2575 NM/19 Bugzilla 745: TLS version reporting.
2576 Patch provided by Phil Pennock.
2578 NM/20 Bugzilla 167: bool: condition support.
2579 Patch provided by Phil Pennock.
2581 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2582 clients. Patch provided by Phil Pennock.
2584 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2585 Patch provided by Brad "anomie" Jorsch.
2587 NM/23 Bugzilla 687: Fix misparses in eximstats.
2588 Patch provided by Heiko Schlittermann.
2590 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2591 Patch provided by Heiko Schlittermann.
2593 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2594 plus update to original patch.
2596 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2598 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2599 Patch provided by David Brownlee.
2601 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2603 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2605 NM/30 Bugzilla 888: TLS documentation bugfixes.
2607 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2609 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2610 Unlike the original bugzilla I have changed all shell scripts in src tree.
2612 NM/33 Bugzilla 898: Transport filter timeout fix.
2613 Patch by Todd Rinaldo.
2615 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2616 Patch by Serge Demonchaux.
2618 NM/35 Bugzilla 39: Base64 decode bug fixes.
2619 Patch by Jakob Hirsch.
2621 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2623 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2625 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2627 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2633 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2634 ALT-N's libdkim that I have put here:
2635 http://duncanthrax.net/exim-experimental/
2637 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2638 were called 'true' and it seems that C99 defines that as a reserved
2639 keyword to be used with 'bool' variable types. That means you could
2640 not include C99-style headers which use bools without triggering
2641 build errors in sieve.c.
2643 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2644 as mailq or other aliases. Changed the --help handling significantly
2645 to do whats expected. exim_usage() emits usage/help information.
2647 SC/01 Added the -bylocaldomain option to eximstats.
2649 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2651 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2653 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2659 PH/01 Another patch from the Sieve maintainer.
2661 PH/02 When an IPv6 address is converted to a string for single-key lookup
2662 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2663 dots are used instead of colons so that keys in lsearch files need not
2664 contain colons. This was done some time before quoting was made available
2665 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2666 (notated using the quote facility) so as to distinguish them from IPv4
2667 keys. This meant that lookups for IP addresses in host lists did not work
2668 for iplsearch lookups.
2670 This has been fixed by arranging for IPv6 addresses to be expressed with
2671 colons if the lookup type is iplsearch. This is not incompatible, because
2672 previously such lookups could never work.
2674 The situation is now rather anomalous, since one *can* have colons in
2675 ordinary lsearch keys. However, making the change in all cases is
2676 incompatible and would probably break a number of configurations.
2678 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2681 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2682 conversion specification without a maximum field width, thereby enabling
2683 a rogue spamd server to cause a buffer overflow. While nobody in their
2684 right mind would setup Exim to query an untrusted spamd server, an
2685 attacker that gains access to a server running spamd could potentially
2686 exploit this vulnerability to run arbitrary code as the Exim user.
2688 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2689 $primary_hostname instead of what libspf2 thinks the hosts name is.
2691 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2692 a directory entry by the name of the lookup key. Previously, if a
2693 symlink pointed to a non-existing file or a file in a directory that
2694 Exim lacked permissions to read, a lookup for a key matching that
2695 symlink would fail. Now it is enough that a matching directory entry
2696 exists, symlink or not. (Bugzilla 503.)
2698 PH/03 The body_linecount and body_zerocount variables are now exported in the
2701 PH/04 Added the $dnslist_matched variable.
2703 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2704 This means they are set thereafter only if the connection becomes
2707 PH/06 Added the client_condition to authenticators so that some can be skipped
2708 by clients under certain conditions.
2710 PH/07 The error message for a badly-placed control=no_multiline_responses left
2711 "_responses" off the end of the name.
2713 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2715 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2716 (without spaces) instead of just copying the configuration text.
2718 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2720 PH/11 Added $max_received_linelength.
2722 PH/12 Added +ignore_defer and +include_defer to host lists.
2724 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2725 way in which PCRE > 7.0 is built.
2727 PH/14 Implemented queue_only_load_latch.
2729 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2730 MAIL command. The effect was to mangle the value on 64-bit systems.
2732 PH/16 Another patch from the Sieve maintainer.
2734 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2736 PH/18 If a system quota error occurred while trying to create the file for
2737 a maildir delivery, the message "Mailbox is full" was not appended to the
2738 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2739 only to a quota excession during the actual writing of the file.
2741 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2742 characters?) which causes problems in log lines. The DN values are now
2743 passed through string_printing() before being added to log lines.
2745 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2746 and InterBase are left for another time.)
2748 PH/21 Added message_body_newlines option.
2750 PH/22 Guard against possible overflow in moan_check_errorcopy().
2752 PH/23 POSIX allows open() to be a macro; guard against that.
2754 PH/24 If the recipient of an error message contained an @ in the local part
2755 (suitably quoted, of course), incorrect values were put in $domain and
2756 $local_part during the evaluation of errors_copy.
2762 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2763 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2766 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2767 issue a MAIL command.
2769 PH/02 In an ACL statement such as
2771 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2773 if a client was not listed at all, or was listed with a value other than
2774 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2775 the condition was not true (as it should be), so access was not denied.
2776 The bug was that the ! inversion was incorrectly passed on to the second
2777 item. This has been fixed.
2779 PH/03 Added additional dnslists conditions == and =& which are different from
2780 = and & when the dns lookup returns more than one IP address.
2782 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2783 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2785 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2786 FSYNC, which compiles an option called disable_fsync that allows for
2787 bypassing fsync(). The documentation is heavily laced with warnings.
2789 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2791 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2792 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2793 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2794 including adding "make clean"; (3) Added -fPIC when compiling the test
2795 dynamically loaded module, to get rid of a warning.
2797 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2798 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2799 = 0s. The bug is otherwise harmless.
2801 PH/07 There was a bug in the dovecot authenticator such that the value of
2802 $auth1 could be overwritten, and so not correctly preserved, after a
2803 successful authentication. This usually meant that the value preserved by
2804 the server_setid option was incorrect.
2806 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2808 PH/09 Installed PCRE release 7.0.
2810 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2811 run for batched SMTP input. It is now run at the start of every message
2812 in the batch. While fixing this I discovered that the process information
2813 (output by running exiwhat) was not always getting set for -bs and -bS
2814 input. This is fixed, and it now also says "batched" for BSMTP.
2816 PH/11 Added control=no_pipelining.
2818 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2819 patch, slightly modified), and move the expansion of helo_data till after
2820 the connection is made in the smtp transport (so it can use these
2823 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2825 PH/14 Added log_selector = +pid.
2827 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2829 PH/16 Add ${if forany and ${if forall.
2831 PH/17 Added dsn_from option to vary the From: line in DSNs.
2833 PH/18 Flush SMTP output before performing a callout, unless control =
2834 no_callout_flush is set.
2836 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2837 was true (the default) a successful delivery failed to delete the retry
2838 item, thus causing premature timeout of the address. The bug is now
2841 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2843 PH/21 Long custom messages for fakedefer and fakereject are now split up
2844 into multiline responses in the same way that messages for "deny" and
2845 other ACL rejections are.
2847 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2848 with slight modification.
2850 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2851 draft, changing the syntax and factoring some duplicate code".
2853 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2854 for deliveries of the second and subsequent messages over the same SMTP
2857 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2858 ${reduce, with only minor "tidies".
2860 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2862 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2863 expansion side effects.
2865 PH/27 When a message times out after an over-quota error from an Exim-imposed
2866 quota, the bounce message says "mailbox is full". This message was not
2867 being given when it was a system quota that was exceeded. It now should
2870 MH/03 Made $recipients available in local_scan(). local_scan() already has
2871 better access to the recipient list through recipients_list[], but
2872 $recipients can be useful in postmaster-provided expansion strings.
2874 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2875 in the case of a MAIL command with additional options following the
2876 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2877 were accidentally chopped off.
2879 PH/29 SMTP synchronization checks are implemented when a command is read -
2880 there is a check that no more input is waiting when there shouldn't be
2881 any. However, for some commands, a delay in an ACL can mean that it is
2882 some time before the response is written. In this time, more input might
2883 arrive, invalidly. So now there are extra checks after an ACL has run for
2884 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2885 pipelining has not been advertised.
2887 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2889 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2890 returned in the SMTP response when smtp_return_error_details was false.
2891 This has been fixed.
2893 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2894 instead of the C I/O that was originally supplied, because problems were
2895 reported on Solaris.
2897 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2898 Exim which did not show up earlier: it was assuming that a call to
2899 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2900 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2901 was a macro that became an assignment, so it seemed to work. This has
2902 changed to a proper function call with a void return, hence the compile
2903 error. Exim's code has been fixed.
2905 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2908 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2910 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2912 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2913 because it turns out to be too big or there is a timeout) there is no
2914 "Completed" line in the log. When some messages of this type were
2915 selected by exigrep, they were listed as "not completed". Others were
2916 picked up by some special patterns. I have improved the selection
2917 criteria to be more general.
2919 PH/38 The host_find_failed option in the manualroute router can now be set
2920 to "ignore", to completely ignore a host whose IP address cannot be
2921 found. If all hosts are ignored, the behaviour is controlled by the new
2922 host_all_ignored option.
2924 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2925 homing or because of multiple MX records with /mx) generated more than
2926 one IP address, and the following item turned out to be the local host,
2927 all the secondary addresses of the first item were incorrectly removed
2928 from the list, along with the local host and any following hosts (which
2929 is what is supposed to happen).
2931 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2932 whoever called Exim into the -H file. In the case of the daemon it was
2933 behaving confusingly. When first started, it used values for whoever
2934 started the daemon, but after a SIGHUP it used the Exim user (because it
2935 calls itself on a restart). I have changed the code so that it now always
2938 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2939 message are rejected with the same error (e.g. no authentication or bad
2940 sender address), and a DATA command is nevertheless sent (as can happen
2941 with PIPELINING or a stupid MUA), the error message that was given to the
2942 RCPT commands is included in the rejection of the DATA command. This is
2943 intended to be helpful for MUAs that show only the final error to their
2946 PH/42 Another patch from the Sieve maintainer.
2948 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2949 Eximstats - Fixed some broken HTML links and added missing column headers
2951 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2952 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2954 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2956 PH/43 Yet another patch from the Sieve maintainer.
2958 PH/44 I found a way to check for a TCP/IP connection going away before sending
2959 the response to the final '.' that terminates a message, but only in the
2960 case where the client has not sent further data following the '.'
2961 (unfortunately, this is allowed). However, in many cases there won't be
2962 any further data because there won't be any more messages to send. A call
2963 to select() can be used: if it shows that the input is "ready", there is
2964 either input waiting, or the socket has been closed. An attempt to read
2965 the next input character can distinguish the two cases. Previously, Exim
2966 would have sent an OK response which the client would never have see.
2967 This could lead to message repetition. This fix should cure that, at
2968 least in a lot of common cases.
2970 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2971 advertised in response to EHLO.
2977 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2978 fixed by 4.65/MH/01 (is this a record?) are fixed:
2980 (i) An empty string was always treated as zero by the numeric comparison
2981 operators. This behaviour has been restored.
2983 (ii) It is documented that the numeric comparison operators always treat
2984 their arguments as decimal numbers. This was broken in that numbers
2985 starting with 0 were being interpreted as octal.
2987 While fixing these problems I realized that there was another issue that
2988 hadn't been noticed. Values of message_size_limit (both the global option
2989 and the transport option) were treated as octal if they started with 0.
2990 The documentation was vague. These values are now always treated as
2991 decimal, and I will make that clear in the documentation.
2997 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2998 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3001 MH/01 Don't check that the operands of numeric comparison operators are
3002 integers when their expansion is in "skipping" mode (fixes bug
3003 introduced by 4.64-PH/07).
3005 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3006 child addresses, Exim now panics and dies. Previously, because the count
3007 is held in a short int, deliveries were likely to be lost. As such a
3008 large number of recipients for a single message is ridiculous
3009 (performance will be very, very poor), I have chosen to impose a limit
3010 rather than extend the field.
3016 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3017 leftover -K file (the existence of which was triggered by #402).
3018 While we were at it, introduced process PID as part of the -K
3019 filename. This should rule out race conditions when creating
3022 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3023 processing considerably. Previous code took too long for large mails,
3024 triggering a timeout which in turn triggers #401.
3026 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3027 in the DK code in transports.c. sendfile() is not really portable,
3028 hence the _LINUX specificness.
3030 TF/01 In the add_headers option to the mail command in an Exim filter,
3031 there was a bug that Exim would claim a syntax error in any
3032 header after the first one which had an odd number of characters
3035 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3036 callout verification, Exim cached a "reject" for the entire domain. This
3037 is correct for most verifications, but it is not correct for a recipient
3038 verification with use_sender or use_postmaster set, because in that case
3039 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3040 case of MAIL FROM:<> rejection from other early rejections (e.g.
3041 rejection of HELO). When verifying a recipient using a non-null MAIL
3042 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3043 Whatever the result of the callout, the value of the domain cache is
3044 left unchanged (for any other kind of callout, getting as far as trying
3045 RCPT means that the domain itself is ok).
3047 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3050 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3051 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3052 session, the daemon ignores ECONNECT errors and logs others; it now
3053 ignores EPIPE as well.
3055 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3056 (quoted-printable decoding).
3058 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3059 later the small subsequent patch to fix an introduced bug.
3061 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3063 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3065 PH/08 An error is now given if message_size_limit is specified negative.
3067 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3068 to be given (somewhat) arbitrary names.
3070 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3073 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3074 miscellaneous code fixes
3076 PH/10 Added the log_reject_target ACL modifier to specify where to log
3079 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3080 hostname. This is wrong, because it relates to the incoming message (and
3081 probably the interface on which it is arriving) and not to the outgoing
3082 callout (which could be using a different interface). This has been
3083 changed to use the value of the helo_data option from the smtp transport
3084 instead - this is what is used when a message is actually being sent. If
3085 there is no remote transport (possible with a router that sets up host
3086 addresses), $smtp_active_hostname is used.
3088 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3089 tweaks were necessary in order to get it to work (see also 21 below):
3090 (a) The code assumed that strncpy() returns a negative number on buffer
3091 overflow, which isn't the case. Replaced with Exim's string_format()
3093 (b) There were several signed/unsigned issues. I just did the minimum
3094 hacking in of casts. There is scope for a larger refactoring.
3095 (c) The code used strcasecmp() which is not a standard C function.
3096 Replaced with Exim's strcmpic() function.
3097 (d) The code set only $1; it now sets $auth1 as well.
3098 (e) A simple test gave the error "authentication client didn't specify
3099 service in request". It would seem that Dovecot has changed its
3100 interface. Fortunately there's a specification; I followed it and
3101 changed what the client sends and it appears to be working now.
3103 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3106 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3107 address A is aliased to B and C, where B exists and C does not. Without
3108 -v the output is "A verified" because verification stops after a
3109 successful redirection if more than one address is generated. However,
3110 with -v the child addresses are also verified. Exim was outputting "A
3111 failed to verify" and then showing the successful verification for C,
3112 with its parentage. It now outputs "B failed to verify", showing B's
3113 parentage before showing the successful verification of C.
3115 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3116 look up a TXT record in a specific list after matching in a combined
3119 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3120 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3121 they consult the DNS. I had assumed they would set it the way they
3122 wanted; and indeed my experiments on Linux seem to show that in some
3123 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3124 To be on the safe side, however, I have now made the interface to
3125 host_find_byname() similar to host_find_bydns(), with an argument
3126 containing the DNS resolver options. The host_find_byname() function now
3127 sets these options at its start, just as host_find_bydns() does. The smtp
3128 transport options dns_qualify_single and dns_search_parents are passed to
3129 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3130 of host_find_byname() use the default settings of RES_DEFNAMES
3131 (qualify_single) but not RES_DNSRCH (search_parents).
3133 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3134 spool_read_header() do less string testing, by means of a preliminary
3135 switch on the second character of optional "-foo" lines. (This is
3136 overdue, caused by the large number of possibilities that now exist.
3137 Originally there were few.) While I was there, I also converted the
3138 str(n)cmp tests so they don't re-test the leading "-" and the first
3139 character, in the hope this might squeeze out yet more improvement.
3141 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3142 flag allowing group syntax was set by the header_syntax check but not
3143 turned off, possible causing trouble later; (2) The flag was not being
3144 set at all for the header_verify test, causing "group"-style headers to
3145 be rejected. I have now set it in this case, and also caused header_
3146 verify to ignore an empty address taken from a group. While doing this, I
3147 came across some other cases where the code for allowing group syntax
3148 while scanning a header line wasn't quite right (mostly, not resetting
3149 the flag correctly in the right place). These bugs could have caused
3150 trouble for malformed header lines. I hope it is now all correct.
3152 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3153 with the "reply" argument non-NULL. The code, however (which originally
3154 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3155 but it didn't always do it. This confused somebody who was copying the
3156 code for some other use. I have removed all the tests.
3158 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3159 feature that was used to support insecure browsers during the U.S. crypto
3160 embargo. It requires special client support, and Exim is probably the
3161 only MTA that supported it -- and would never use it because real RSA is
3162 always available. This code has been removed, because it had the bad
3163 effect of slowing Exim down by computing (never used) parameters for the
3164 RSA_EXPORT functionality.
3166 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3167 authenticator to fail if there's a tab character in the incoming data
3168 (there should never be unless someone is messing about, as it's supposed
3169 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3170 if the connection is using TLS or if the remote IP is the same as the
3171 local IP, and the "valid-client-cert option" if a client certificate has
3174 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3175 authenticators. This can be used for authorization after authentication
3176 succeeds. (In the case of plaintext, it servers for both authentication
3179 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3180 if any retry times were supplied.
3182 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3183 connection, where there is no client IP address to check. In this
3184 situation, the verify now always succeeds.
3186 PH/25 Applied John Jetmore's -Mset patch.
3188 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3190 PH/27 In a string expansion for a processed (not raw) header when multiple
3191 headers of the same name were present, leading whitespace was being
3192 removed from all of them, but trailing whitespace was being removed only
3193 from the last one. Now trailing whitespace is removed from each header
3194 before concatenation. Completely empty headers in a concatenation (as
3195 before) are ignored.
3197 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3198 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3200 PH/29 [Removed. This was a change that I later backed out, and forgot to
3201 correct the ChangeLog entry (that I had efficiently created) before
3202 committing the later change.]
3204 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3205 address errors (4xx response to RCPT) over the same connection as other
3206 messages routed to the same hosts. Such deliveries are always "forced",
3207 so retry times are not inspected. This resulted in far too many retries
3208 for the affected addresses. The effect occurred only when there were more
3209 hosts than the hosts_max_try setting in the smtp transport when it had
3210 the 4xx errors. Those hosts that it had tried were not added to the list
3211 of hosts for which the message was waiting, so if all were tried, there
3212 was no problem. Two fixes have been applied:
3214 (i) If there are any address or message errors in an SMTP delivery, none
3215 of the hosts (tried or untried) are now added to the list of hosts
3216 for which the message is waiting, so the message should not be a
3217 candidate for sending over the same connection that was used for a
3218 successful delivery of some other message. This seems entirely
3219 reasonable: after all the message is NOT "waiting for some host".
3220 This is so "obvious" that I'm not sure why it wasn't done
3221 previously. Hope I haven't missed anything, but it can't do any
3222 harm, as the worst effect is to miss an optimization.
3224 (ii) If, despite (i), such a delivery is accidentally attempted, the
3225 routing retry time is respected, so at least it doesn't keep
3226 hammering the server.
3228 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3229 in ${readsocket because some servers need this prod.
3231 PH/32 Added some extra debug output when updating a wait-xxx database.
3233 PH/33 The hint "could be header name not terminated by colon", which has been
3234 given for certain expansion errors for a long time, was not being given
3235 for the ${if def:h_colon_omitted{... case.
3237 PH/34 The spec says: "With one important exception, whenever a domain list is
3238 being scanned, $domain contains the subject domain." There was at least
3239 one case where this was not true.
3241 PH/35 The error "getsockname() failed: connection reset by peer" was being
3242 written to the panic log as well as the main log, but it isn't really
3243 panic-worthy as it just means the connection died rather early on. I have
3244 removed the panic log writing for the ECONNRESET error when getsockname()
3247 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3248 runs only) independently of the message's sender address. This meant
3249 that, if the 4xx error was in fact related to the sender, a different
3250 message to the same recipient with a different sender could confuse
3251 things. In particular, this can happen when sending to a greylisting
3252 server, but other circumstances could also provoke similar problems.
3253 I have changed the default so that the retry time for these errors is now
3254 based a combination of the sender and recipient addresses. This change
3255 can be overridden by setting address_retry_include_sender=false in the
3258 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3259 remote server are returned as part of bounce messages. This was not
3260 happening for LMTP over a pipe (the lmtp transport), but now it is the
3261 same for both kinds of LMTP.
3263 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3264 in header lines that were in fact CNAMEs. This is no longer the case.
3266 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3267 and queue runs started by the daemon processed all messages. This has
3268 been fixed so that -R and -S can now usefully be given with -q<time>.
3270 PH/40 Import PCRE release 6.7 (fixes some bugs).
3272 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3274 PH/42 Give an error if -q is specified more than once.
3276 PH/43 Renamed the variables $interface_address and $interface_port as
3277 $received_ip_address and $received_port, to make it clear that these
3278 values apply to message reception, and not to the outgoing interface when
3279 a message is delivered. (The old names remain recognized, of course.)
3281 PH/44 There was no timeout on the connect() call when using a Unix domain
3282 socket in the ${readsocket expansion. There now is.
3284 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3285 be meaningful with "accept".
3287 SC/01 Eximstats V1.43
3288 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3290 SC/02 Eximstats V1.44
3291 Use a glob alias rather than an array ref in the generated
3292 parser. This improves both readability and performance.
3294 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3295 Collect SpamAssassin and rejection statistics.
3296 Don't display local sender or destination tables unless
3297 there is data to show.
3298 Added average volumes into the top table text output.
3300 SC/04 Eximstats V1.46
3301 Collect data on the number of addresses (recipients)
3302 as well as the number of messages.
3304 SC/05 Eximstats V1.47
3305 Added 'Message too big' to the list of mail rejection
3306 reasons (thanks to Marco Gaiarin).
3308 SC/06 Eximstats V1.48
3309 Mainlog lines which have GMT offsets and are too short to
3310 have a flag are now skipped.
3312 SC/07 Eximstats V1.49 (Alain Williams)
3313 Added the -emptyok flag.
3315 SC/08 Eximstats V1.50
3316 Fixes for obtaining the IP address from reject messages.
3318 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3319 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3320 whitespace changes from 4.64-PH/27
3322 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3325 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3326 are found, allow negative numbers in numeric criteria)
3328 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3330 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3331 to match changes made in 4.64-PH/43
3333 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3335 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3336 "domain" is actually an IP address, and force a failure. This locks out
3337 those revolvers/nameservers that support "A-for-A" lookups, in
3338 contravention of the specifications.
3340 PH/48 When a host name was looked up from an IP address, and the subsequent
3341 forward lookup of the name timed out, the host name was left in
3342 $sender_host_name, contrary to the specification.
3344 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3345 restricted to single-key lookups, Exim was not diagnosing an error if
3346 * or *@ was used with a query-style lookup.
3348 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3350 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3351 long ago, but noone interested enough thought of it. Let's just say that
3352 the "1.1" means that there are some new functions that weren't there at
3353 some point in the past.
3355 PH/51 Error processing for expansion failure of helo_data from an smtp
3356 transport during callout processing was broken.
3358 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3359 tested/used via the -bh/-bhc/-bs options.
3361 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3362 bug, fixed in subsequent PCRE releases).
3364 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3365 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3367 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3373 SC/01 Use a glob alias rather than an array ref in eximstats generated
3374 parser. This improves both readability and performance.
3376 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3377 Don't display local sender or destination tables in eximstats unless
3378 there is data to show.
3379 Added average volumes into the eximstats top table text output.
3381 SC/03 Collect data on the number of addresses (recipients) as well
3382 as the number of messages in eximstats.
3384 TF/01 Correct an error in the documentation for the redirect router. Exim
3385 does (usually) call initgroups() when daemonizing.
3387 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3388 with consistent privilege compared to when running as a daemon.
3390 TF/03 Note in the spec that $authenticated_id is not set for local
3391 submissions from trusted users.
3393 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3394 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3396 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3397 by adding some example configuration directives to the default
3398 configuration file. A little bit of work is required to uncomment the
3399 directives and define how usernames and passwords are checked, but
3400 there is now a framework to start from.
3402 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3403 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3404 without this. I don't know how relevant this is to other LDAP libraries.
3406 PH/02 Add the verb name to the "unknown ACL verb" error.
3408 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3410 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3412 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3413 directory not expanded when it should be if an expanded home directory
3414 was set for the address (which is overridden by the transport).
3416 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3419 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3420 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3421 because it is too late at that time, and has no effect.
3423 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3424 security issue with \' (bugzilla #107). I could not use the
3425 PQescapeStringConn() function, because it needs a PGconn value as one of
3428 PH/08 When testing addresses using -bt, indicate those final addresses that
3429 are duplicates that would not cause an additional delivery. At least one
3430 person was confused, thinking that -bt output corresponded to deliveries.
3431 (Suppressing duplicates isn't a good idea as you lose the information
3432 about possibly different redirections that led to the duplicates.)
3434 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3435 systems where poll() doesn't work, in particular OS X.
3437 PH/10 Added more information to debugging output for retry time not reached.
3439 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3440 operations in malware.c.
3442 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3445 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3446 syslog_duplication set false, log lines that would normally be written
3447 both the the main log and to the reject log were not written to syslog at
3450 PH/14 In the default configuration, change the use of "message" in ACL warn
3451 statements to "add_header".
3453 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3454 not followed by a command (e.g. "seen endif").
3456 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3457 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3460 PH/17 Added extra conditions to the default value of delay_warning_condition
3464 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3465 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3466 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3469 The Auto-Submitted: and various List- headers are standardised, whereas I
3470 don't think Precedence: ever was.
3472 PH/18 Refactored debugging code in route_finduser() to show more information,
3473 in particular, the error code if getpwnam() issues one.
3475 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3476 This is apparently needed in addition to the PH/07 change above to avoid
3477 any possible encoding problems.
3479 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3480 but not after initializing Perl.
3482 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3483 output them only if debugging. By default they are written stderr,
3484 apparently, which is not desirable.
3486 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3489 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3492 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3494 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3495 authenticated or an ident call has been made. Suppress the default
3496 values for $authenticated_id and $authenticated_sender (but permit -oMai
3497 and -oMas) when testing with -bh.
3499 PH/24 Re-jigged the order of the tests in the default configuration so that the
3500 tests for valid domains and recipients precede the DNS black list and CSA
3501 tests, on the grounds that those ones are more expensive.
3503 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3504 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3505 This bug exists in every version of Exim that I still have, right back to
3508 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3509 However, an attempt to turn on case-sensitivity in a regex key by
3510 including (?-i) didn't work because the subject string was already
3511 lowercased, and the effects were non-intuitive. It turns out that a
3512 one-line patch can be used to allow (?-i) to work as expected.
3518 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3519 other effects) broke the use of negated acl sub-conditions.
3521 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3524 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3525 "Deny" causes Exim to reject the incoming connection with a 554 error.
3526 Unfortunately, if there is a major crisis, such as a disk failure,
3527 tcp-wrappers gives "deny", whereas what one would like would be some
3528 kind of temporary error. A kludge has been added to help with this.
3529 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3530 554 error is used if errno is still zero or contains ENOENT (which occurs
3531 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3534 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3536 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3537 errors. Otherwise a message that provokes a temporary error (when other
3538 messages do not) can cause a whole host to time out.
3540 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3541 addresses were routed directly to files or pipes from a redirect router.
3542 File deliveries just didn't batch; pipe deliveries might have suffered
3545 PH/06 A failure to get a lock for a hints database would erroneously always say
3546 "Failed to get write lock", even when it was really a read lock.
3548 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3549 of 0600. This has been changed to use the value of the lockfile_mode
3550 option (which defaults to 0600).
3552 PH/08 Applied small patch from the Sieve maintainer.
3554 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3555 folder from quota calculations, a direct delivery into this folder messed
3556 up the contents of the maildirsize file. This was because the regex was
3557 used only to exclude .Trash (or whatever) when the size of the mailbox
3558 was calculated. There was no check that a delivery was happening into an
3559 excluded directory. This bug has been fixed by ignoring all quota
3560 processing for deliveries into excluded directories.
3562 PH/10 Added the maildirfolder_create_regex option to appendfile.
3568 PH/01 The code for finding all the local interface addresses on a FreeBSD
3569 system running IPv6 was broken. This may well have applied to all BSD
3570 systems, as well as to others that have similar system calls. The broken
3571 code found IPv4 interfaces correctly, but gave incorrect values for the
3572 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3573 that it would not match correctly against @[] and not recognize the IPv6
3576 PH/02 The ipliteral router was not recognizing addresses of the form user@
3577 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3579 PH/03 Added disable_ipv6.
3581 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3582 decoded form, because it is most often used to construct To: headers
3583 lines in autoreplies, and the decoded form may well be syntactically
3584 invalid. However, $reply_address has leading white space removed, and all
3585 newlines turned into spaces so that the autoreply transport does not
3588 PH/05 If group was specified without a user on a router, and no group or user
3589 was specified on a transport, the group from the router was ignored.
3591 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3592 for visible compile-time settings that can be used to change these
3593 numbers, for those that want even more. Backwards compatibility with old
3594 spool files has been maintained. However, going back to a previous Exim
3595 release will lost any variables that are in spool files.
3597 PH/07 Two small changes when running in the test harness: increase delay when
3598 passing a TCP/IP connection to a new process, in case the original
3599 process has to generate a bounce, and remove special handling of
3600 127.0.0.2 (sic), which is no longer necessary.
3602 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3603 be the same on different OS.
3605 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3608 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3609 whether --show-vars was specified or not
3611 JJ/02 exipick: Added support for new ACL variable spool format introduced
3614 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3615 syntactically invalid From: or Reply-to: line, and a filter used this to
3616 generate an autoreply, and therefore failed to obtain an address for the
3617 autoreply, Exim could try to deliver to a non-existent relative file
3618 name, causing unrelated and misleading errors. What now happens is that
3619 it logs this as a hard delivery error, but does not attempt to create a
3622 PH/11 The exinext utility has a -C option for testing purposes, but although
3623 the given file was scanned by exinext itself; it wasn't being passed on
3624 when Exim was called.
3626 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3627 an end-of-file indication when reading a command response.
3629 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3630 compiled. In many other places in Exim, IPv6 addresses are always
3631 recognized, so I have changed this. It also means that IPv4 domain
3632 literals of the form [IPV4:n.n.n.n] are now always recognized.
3634 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3635 used if the router is not running as root, for example, when verifying at
3636 ACL time, or when using -bh. The debugging output from this situation was
3637 non-existent - all you got was a failure to exec. I have made two
3640 (a) Failures to set uid/gid, the current directory, or a process leader
3641 in a subprocess such as that created by queryprogram now generate
3642 suitable debugging output when -d is set.
3644 (b) The queryprogram router detects when it is not running as root,
3645 outputs suitable debugging information if -d is set, and then runs
3646 the subprocess without attempting to change uid/gid.
3648 PH/15 Minor change to Makefile for building test_host (undocumented testing
3651 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3652 additional section of a DNS packet that returns MX or SRV records.
3653 Instead, it always explicitly searches for A/AAAA records. This avoids
3654 major problems that occur when a DNS server includes only records of one
3655 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3656 fixed another bug: if SRV records were looked up and the corresponding
3657 address records were *not* found in the additional section, the port
3658 values from the SRV records were lost.
3660 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3661 using the correct key (the original address) when searching the retry
3662 rules in order to find which one to use for generating the retry hint.
3664 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3665 adding the default one. Similarly, if it contains a Reply-To: header, the
3666 errors_reply_to option, if set, is not used.
3668 PH/19 When calculating a retry time, Exim used to measure the "time since
3669 failure" by looking at the "first failed" field in the retry record. Now
3670 it does not use this if it is later than than the arrival time of the
3671 message. Instead it uses the arrival time. This makes for better
3672 behaviour in cases where some deliveries succeed, thus re-setting the
3673 "first failed" field. An example is a quota failure for a huge message
3674 when small messages continue to be delivered. Without this change, the
3675 "time since failure" will always be short, possible causing more frequent
3676 delivery attempts for the huge message than are intended.
3677 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3679 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3680 $1, $2, $3) because the numerical variables can be reset during some
3681 expansion items (e.g. "match"), thereby losing the authentication data.
3683 PH/21 Make -bV show the size of off_t variables so that the test suite can
3684 decide whether to run tests for quotas > 2G.
3686 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3687 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3688 filecount value is greater than 2G or if a quota value is greater than 2G
3689 on a system where the size of off_t is not greater than 4, a panic error
3692 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3693 never match. The debug and -bh output now contains an explicit error
3694 message indicating a malformed IPv4 address or mask.
3696 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3697 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3698 PH/23 above applies.
3700 PH/25 Do not write to syslog when running in the test harness. The only
3701 occasion when this arises is a failure to open the main or panic logs
3702 (for which there is an explicit test).
3704 PH/26 Added the /no_tell option to "control=freeze".
3706 PH/27 If a host name lookup failed very early in a connection, for example, if
3707 the IP address matched host_lookup and the reverse lookup yielded a name
3708 that did not have a forward lookup, an error message of the form "no IP
3709 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3710 could be logged. Now it outputs the IP address instead of "NULL".
3712 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3713 allows the sender and the authenticated sender to be set when
3714 submitting a message from within Exim. Since child_open_exim() is
3715 documented for local_scan(), the new function should be too.
3717 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3718 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3719 results in an empty string is now treated as unset.
3721 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3723 PH/31 Added sender_verify_fail logging option.
3725 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3726 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3727 tidied the source and removed it altogether.
3729 PH/33 When a queue run was abandoned because the load average was too high, a
3730 log line was always written; now it is written only if the queue_run log
3731 selector is set. In addition, the log line for abandonment now contains
3732 information about the queue run such as the pid. This is always present
3733 in "start" and "stop" lines but was omitted from the "abandon" line.
3735 PH/34 Omit spaces between a header name and the colon in the error message that
3736 is given when verify = headers_syntax fails (if there are lots of them,
3737 the message gets confusing).
3739 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3740 names, as there are now some PTR records that contain slashes. This check
3741 is only to protect against broken name servers that fall over on strange
3742 characters, so the fact that it applies to all lookups doesn't matter.
3744 PH/36 Now that the new test suite is complete, we can remove some of the
3745 special code in Exim that was needed for the old test suite. For example,
3746 sorting DNS records because real resolvers return them in an arbitrary
3747 order. The new test suite's fake resolver always returns records in the
3750 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3751 bounces) except when queue_only is set, to avoid logging races between
3752 the different processes.
3754 PH/38 Panic-die if .include specifies a non-absolute path.
3756 PH/39 A tweak to the "H" retry rule from its user.
3758 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3759 a label. They prevented compilation on older perls.
3761 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3762 a warning to be raised on newish perls.
3764 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3765 on queue. Changes to match documented behaviour of showing count of
3766 messages matching specified criteria.
3768 PH/40 Changed the default ident timeout from 30s to 5s.
3770 PH/41 Added support for the use of login_cap features, on those BSD systems
3771 that have them, for controlling the resources used by pipe deliveries.
3773 PH/42 The content-scanning code uses fopen() to create files in which to put
3774 message data. Previously it was not paying any attention to the mode of
3775 the files. Exim runs with umask(0) because the rest of the code creates
3776 files with open(), and sets the required mode explicitly. Thus, these
3777 files were ending up world-writeable. This was not a big issue, because,
3778 being within the spool directory, they were not world-accessible. I have
3779 created a function called modefopen, which takes an additional mode
3780 argument. It sets umask(777), creates the file, chmods it to the required
3781 mode, then resets the umask. All the relevant calls to fopen() in the
3782 content scanning code have been changed to use this function.
3784 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3785 to 24 hours. This avoids potential overflow problems when processing G
3786 and H retry rules. I suspect nobody ever tinkers with this value.
3788 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3790 PH/45 When the plaintext authenticator is running as a client, the server's
3791 challenges are checked to ensure they are valid base64 strings. By
3792 default, the authentication attempt is cancelled if an invalid string is
3793 received. Setting client_ignore_invalid_base64 true ignores these errors.
3794 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3795 they are received. Thus, the responses can be made to depend on the
3796 challenges. If an invalid string is ignored, an empty string is placed in
3799 PH/46 Messages that are created by the autoreply transport now contains a
3800 References: header, in accordance with RFCs 2822 and 3834.
3802 PH/47 Added authenticated_sender_force to the smtp transport.
3804 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3806 PH/49 Installed latest patch from the Sieve maintainer.
3808 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3809 was also set, the appendfile transport was unnecessarily scanning a
3810 directory of message files (e.g. for maildir delivery) to find the count
3811 of files (along with the size), even though it did not need this
3812 information. It now does the scan only if it needs to find either the
3813 size of the count of files.
3815 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3817 PH/52 Two bugs concerned with error handling when the smtp transport is
3820 (i) Exim was not creating retry information for temporary errors given
3821 for individual recipients after the DATA command when the smtp transport
3822 was used in LMTP mode. This meant that they could be retried too
3823 frequently, and not timed out correctly.
3825 (ii) Exim was setting the flag that allows error details to be returned
3826 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3827 recipients that were returned after the DATA command.
3829 PH/53 This is related to PH/52, but is more general: for any failing address,
3830 when detailed error information was permitted to be returned to the
3831 sender, but the error was temporary, then after the final timeout, only
3832 "retry timeout exceeded" was returned. Now it returns the full error as
3833 well as "retry timeout exceeded".
3835 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3836 do this, and (what is worse) MTAs that accept it.
3838 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3839 will now be deprecated.
3841 PH/56 New os.c-cygwin from the Cygwin maintainer.
3843 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3844 formats (previously only available in exim formats via -bpr, -bpru,
3845 and -bpra. Now also available in native and exiqgrep formats)
3847 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3848 with very large, slow to parse queues
3850 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3852 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3854 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3855 responses to authentication challenges, though it was showing the
3856 challenges; (ii) I've removed the CR characters from the debug output for
3859 PH/58 Allow for the insertion of a newline as well as a space when a string
3860 is turned into more than one encoded-word during RFC 2047 encoding. The
3861 Sieve code now uses this.
3863 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3864 data_4xx, lost_connection, tls_required.
3866 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3867 message was being sent as an SMTP response.
3869 PH/61 Add -l and -k options to exicyclog.
3871 PH/62 When verifying, if an address was redirected to one new address, so that
3872 verification continued, and the new address failed or deferred after
3873 having set something in $address_data, the value of $address_data was not
3874 passed back to the ACL. This was different to the case when no
3875 redirection occurred. The value is now passed back in both cases.
3877 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3878 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3879 use login_cap.h, so on its own it isn't the distinguishing feature. The
3880 new name refers directly to the setclassresources() function.
3882 PH/65 Added configuration files for NetBSD3.
3884 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3886 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3887 is preferred over IPv4.
3889 PH/68 The bounce_return_message and bounce_return_body options were not being
3890 honoured for bounces generated during the reception of non-SMTP messages.
3891 In particular, this applied to messages rejected by the ACL. This bug has
3892 been fixed. However, if bounce_return_message is true and bounce_return_
3893 body is false, the headers that are returned for a non-SMTP message
3894 include only those that have been read before the error was detected.
3895 (In the case of an ACL rejection, they have all been read.)
3897 PH/69 The HTML version of the specification is now built in a directory called
3898 spec_html instead of spec.html, because the latter looks like a path with
3899 a MIME-type, and this confuses some software.
3901 PH/70 Catch two compiler warnings in sieve.c.
3903 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3904 function verify_get_ident() calls ip_connect() to connect a socket, but
3905 if the "connect()" function timed out, ip_connect() used to close the
3906 socket. However, verify_get_ident() also closes the socket later, and in
3907 between Exim writes to the log, which may get opened at this point. When
3908 the socket was closed in ip_connect(), the log could get the same file
3909 descriptor number as the socket. This naturally causes chaos. The fix is
3910 not to close the socket in ip_connect(); the socket should be closed by
3911 the function that creates it. There was only one place in the code where
3912 this was missing, in the iplookup router, which I don't think anybody now
3913 uses, but I've fixed it anyway.
3915 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3916 well as to direct DNS lookups. Otherwise the handling of names in host
3917 lists is inconsistent and therefore confusing.
3923 PH/01 Two changes to the default runtime configuration:
3925 (1) Move the checks for relay_from_hosts and authenticated clients from
3926 after to before the (commented out) DNS black list checks.
3928 (2) Add control=submission to the relay_from_hosts and authenticated
3929 clients checks, on the grounds that messages accepted by these
3930 statements are most likely to be submissions.
3932 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3934 (1) Generate an error if the third argument for the ${prvs expansion is
3937 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3940 (3) Reset the variables that are obtained from the first argument of
3941 ${prvscheck and used in the second argument before leaving the code,
3942 because their memory is reclaimed, so using them afterwards may do
3945 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3946 one (it's much easier than Tom thought :-).
3948 (5) Because of (4), we can now allow for the use of $prvscheck_result
3949 inside the third argument.
3951 PH/03 For some reason, the default setting of PATH when running a command from
3952 a pipe transport was just "/usr/bin". I have changed it to
3955 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3956 anything to be listed in the output from -bV.
3958 PH/05 When a filter generated an autoreply, the entire To: header line was
3959 quoted in the delivery log line, like this:
3961 => >A.N.Other <ano@some.domain> <original@ddress> ...
3963 This has been changed so that it extracts the operative address. There
3964 may be more than one such address. If so, they are comma-separated, like
3967 => >ano@some.domain,ona@other.domain <original@ddress> ...
3969 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3970 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3971 not being looked up in the rDNS to get a host name, Exim was showing the
3972 IP address twice in Received: lines, even though the IP addresses were
3973 identical. For example:
3975 Received: from [1.2.3.4] (helo=[1.2.3.4])
3977 However, if the real host name was known, it was omitting the HELO data
3978 if it matched the actual IP address. This has been tidied up so that it
3979 doesn't show the same IP address twice.
3981 PH/07 When both +timestamp and +memory debugging was on, the value given by
3982 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3983 called by the debug printing, thereby overwriting the timestamp buffer.
3984 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3986 PH/08 When the original message was included in an autoreply transport, it
3987 always said "this is a copy of the message, including all the headers",
3988 even if body_only or headers_only was set. It now gives an appropriate
3991 PH/09 Applied a patch from the Sieve maintainer which:
3993 o fixes some comments
3994 o adds the (disabled) notify extension core
3995 o adds some debug output for the result of if/elsif tests
3996 o points to the current vacation draft in the documentation
3997 and documents the missing references header update
4001 o fixes a bug in processing the envelope test (when testing
4002 multiple envelope elements, the last element determined the
4005 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4006 Electronic Mail") by including:
4008 Auto-submitted: auto-generated
4010 in the messages that it generates (bounce messages and others, such as
4011 warnings). In the case of bounce messages for non-SMTP messages, there was
4012 also a typo: it was using "Auto_submitted" (underscore instead of
4013 hyphen). Since every message generated by Exim is necessarily in response
4014 to another message, thes have all been changed to:
4016 Auto-Submitted: auto-replied
4018 in accordance with these statements in the RFC:
4020 The auto-replied keyword:
4022 - SHOULD be used on messages sent in direct response to another
4023 message by an automatic process,
4025 - MUST NOT be used on manually-generated messages,
4027 - MAY be used on Delivery Status Notifications (DSNs) and Message
4028 Disposition Notifications (MDNs),
4030 - MUST NOT be used on messages generated by automatic or periodic
4031 processes, except for messages which are automatic responses to
4034 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4035 to the default Received: header definition.
4037 PH/12 Added log selector acl_warn_skipped (default on).
4039 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4040 variables because (a) they are in the wrong storage pool and (b) even if
4041 they were copied, it wouldn't work properly because of the caching.
4043 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4044 checking when decoding. Apparently there are clients that generate
4045 overlong encoded strings. Why am I not surprised?
4047 PH/15 If the first argument of "${if match_address" was not empty, but did not
4048 contain an "@" character, Exim crashed. Now it writes a panic log message
4049 and treats the condition as false.
4051 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4053 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4054 extension "envelope-auth". The code is finished and in agreement with
4055 other implementations, but there is no documentation so far and in fact,
4056 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4057 not changing the active code.
4059 Print executed "if" and "elsif" statements when debugging is used. This
4060 helps a great deal to understand what a filter does.
4062 Document more things not specified clearly in RFC3028. I had all this
4063 sorted out, when out of a sudden new issues came to my mind. Oops."
4065 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4068 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4069 canonical form (as documented). However, after a host name lookup from
4070 the IP address, check_host() was doing a simple string comparison with
4071 addresses acquired from the DNS when checking that the found name did
4072 have the original IP as one of its addresses. Since any found IPv6
4073 addresses are likely to be in abbreviated form, the comparison could
4074 fail. Luckily, there already exists a function for doing the comparison
4075 by converting both addresses to binary, so now that is used instead of
4076 the text comparison.
4078 PH/20 There was another similar case to PH/19, when a complete host name was
4079 given in a host list; looking up its IP address could give an abbreviated
4080 form, whereas the current host's name might or might not be abbreviated.
4081 The same fix has been applied.
4087 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4088 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4091 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4092 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4094 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4096 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4097 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4098 Also, for "Auto-Submitted", treat anything other than "no" as
4099 non-personal, in accordance with RFC 3834. (Previously it treated
4100 anything starting "auto-" as non-personal.)
4102 TF/01 The control=submission/name=... option had a problem with syntax
4103 errors if the name included a slash character. The /name= option
4104 now slurps the rest of the string, so it can include any characters
4105 but it must come last in the list of options (after /sender_retain
4108 PH/05 Some modifications to the interface to the fake nameserver for the new
4116 TK/01 Added the "success_on_redirect" address verification option. See
4117 NewStuff for rationale and an example.
4119 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4121 PH/02 Patch to exigrep to allow it to work on syslog lines.
4123 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4124 fread() to skip over the body file's header line, because in Cygwin the
4125 header line is locked and is inaccessible.
4127 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4128 co-exist for some time) to make it clear that it is the Exim ID that is
4129 referenced, not the Message-ID: header line.
4131 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4132 string_format() function, because snprintf() does not exist on all
4135 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4136 ${stat: expansion item.
4138 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4139 protocol synchronization error", to keep the pedants happy.
4141 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4142 well as for IRIX systems, when gcc is being used. See the host.c source
4145 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4147 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4150 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4151 transport and to the smtp transport in LMTP mode.
4153 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4155 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4156 run a filter in a subprocess. This could lead to confusion in subsequent
4157 lookups in the parent process. There should also be a search_tidyup() at
4158 the end of the subprocess.
4160 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4161 only if the host matched helo_try_verify_hosts, which caused the
4162 verification to occur when the EHLO/HELO command was issued. The ACL just
4163 tested the remembered result. Now, if a previous verification attempt has
4164 not happened, "verify = helo" does it there and then.
4166 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4168 TK/03 Fix log output including CR from clamd.
4170 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4171 exist provoked a memory error which could cause a segfault.
4173 PH/15 Installed PCRE 6.2
4175 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4177 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4178 of the problem. Specifically, suggested +O2 rather than +O1 for the
4181 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4183 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4184 record, the port was not correctly logged when the outgoing_port log
4185 selector was set (it logged the transort's default port).
4187 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4188 fallback_hosts, and "hosts" in the smtp transport.
4190 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4191 host errors such as "Connection refused".
4193 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4194 authentication with radiusclient 0.4.9:
4196 - Error returned from rc_read_config was caught wrongly
4197 - Username/password not passed on to radius server due to wrong length.
4199 The presumption is that some radiusclient API changes for 4.51/PH/17
4200 were not taken care of correctly. The code is still untested by me (my
4201 Linux distribution still has 0.3.2 of radiusclient), but it was
4202 contributed by a Radius user.
4204 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4205 expanding the "port" option of the smtp transport.
4207 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4208 while reading a MIME header. Thanks to Tom Hughes for a patch.
4210 PH/24 Include config.h inside local_scan.h so that configuration settings are
4213 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4214 that in an ACL for RCPT (for example), you can examine exactly what was
4217 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4218 commands, but it was not correctly comparing the address with the actual
4219 client host address. Thus, it would show the EHLO address in Received:
4220 header lines when this was not necessary.
4222 PH/27 Added the % operator to ${eval:}.
4224 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4225 it should be ignoring failures (because with -C, for example, it has lost
4226 privilege). It wasn't ignoring creation failures other than "already
4229 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4232 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4233 because an input line was too long, either on its own, or by virtue of
4234 too many continuations, the temporary file was not being removed, and the
4235 return code was incorrect.
4237 PH/31 Missing "BOOL" in function definition in filtertest.c.
4239 PH/32 Applied Sieve patches from the maintainer.
4241 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4243 PH/33 Added "verify = not_blind".
4245 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4246 Local/Makefile (with some defaults set). These are used in built scripts
4247 such as exicyclog, but they have never been used in the exim_install
4248 script (though there are many overriding facilities there). I have
4249 arranged that the exim_install script now takes note of these two
4252 PH/35 Installed configuration files for Dragonfly.
4254 PH/36 When a locally submitted message by a trusted user did not contain a
4255 From: header, and the sender address was obtained from -f or from an SMTP
4256 MAIL command, and the trusted user did not use -F to supply a sender
4257 name, $originator_name was incorrectly used when constructing a From:
4258 header. Furthermore, $originator_name was used for submission mode
4259 messages from external hosts without From: headers in a similar way,
4260 which is clearly wrong.
4262 PH/37 Added control=suppress_local_fixups.
4264 PH/38 When log_selector = +received_sender was set, and the addition of the
4265 sender made the log line's construction buffer exactly full, or one byte
4266 less than full, an overflow happened when the terminating "\n" was
4269 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4270 when the result of a list match is failure because a DNS lookup failed.
4272 PH/40 RM_COMMAND is now used in the building process.
4274 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4275 the "build-* directories that it finds.
4277 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4278 address in a domain literal was a prefix of an interface address.
4280 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4281 when verifying a sender address, unless rewrite_headers is false.
4283 PH/44 Wrote a long comment about why errors_to addresses are verified as
4284 recipients, not senders.
4286 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4287 the ratelimit ACL was added.
4289 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4291 PH/46 Added extra information about PostgreSQL errors to the error string.
4293 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4294 avoiding the need to install special zones in a real server. This is
4295 backwards compatible; if it can't find the fake resolver, it drops back.
4296 Thus, both old and new test suites can be run.
4298 TF/02 Added util/ratelimit.pl
4300 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4301 clock is set back in time.
4303 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4304 Candler <B.Candler@pobox.com>.
4306 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4307 OK for addresses that are the result of redirections.
4309 PH/48 A number of further additions for the benefit of the new test suite,
4310 including a fake gethostbyname() that interfaces to the fake DNS resolver
4313 TF/06 The fix for widen_domains has also been applied to qualify_single and
4314 search_parents which are the other dnslookup options that can cause
4317 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4320 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4322 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4323 service" if it encountered an MX record with an empty target hostname.
4324 The message is now "an MX or SRV record indicated no SMTP service".
4326 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4327 if the DNS of the sending site is misconfigured. This is quite a
4328 common situation. This change restores the behaviour of treating a
4329 helo verification defer as a failure.
4331 PH/51 If self=fail was set on a router, the bounce message did not include the
4332 actual error message.
4338 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4340 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4341 command itself ended in error, the underlying message about the transport
4342 filter timeout was being overwritten with the pipe command error. Now the
4343 underlying error message should be appended to the second error message.
4345 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4347 PH/02 Reduce the amount of output that "make" produces by default. Full output
4348 can still be requested.
4350 PH/03 The warning log line about a condition test deferring for a "warn" verb
4351 was being output only once per connection, rather than after each
4352 occurrence (because it was using the same function as for successful
4353 "warn" verbs). This seems wrong, so I have changed it.
4355 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4356 it should not have, which might have caused a crash in the right
4357 circumstances, but probably never did.
4359 PH/04 Installed a modified version of Tony Finch's patch to make submission
4360 mode fix the return path as well as the Sender: header line, and to
4361 add a /name= option so that you can make the user's friendly name appear
4364 TF/03 Added the control = fakedefer ACL modifier.
4366 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4367 Mark Lowes for thorough testing.
4369 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4371 TK/03 Merged latest SRS patch from Miles Wilton.
4373 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4374 with the definition in sysexits.h (which is #included earlier).
4375 Fortunately, Exim does not actually use EX_OK. The code used to try to
4376 preserve the sysexits.h value, by assuming that macro definitions were
4377 scanned for macro replacements. I have been disabused of this notion,
4378 so now the code just undefines EX_OK before #including unistd.h.
4380 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4381 in the smtp transport. When a block could not be written in a single
4382 write() function, the timeout was being re-applied to each part-write.
4383 This seems wrong - if the receiver was accepting one byte at a time it
4384 would take for ever. The timeout is now adjusted when this happens. It
4385 doesn't have to be particularly precise.
4387 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4388 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4390 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4391 without a domain if the check to <postmaster@domain> fails.
4393 SC/01 Eximstats: added -xls and the ability to specify output files
4394 (patch written by Frank Heydlauf).
4396 SC/02 Eximstats: use FileHandles for outputting results.
4398 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4400 SC/04 Eximstats: fixed display of large numbers with -nvr option
4402 SC/05 Eximstats: fixed merging of reports with empty tables.
4404 SC/06 Eximstats: added the -include_original_destination flag
4406 SC/07 Eximstats: removed tabs and trailing whitespace.
4408 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4410 TK/06 MBOX spool code: Add real "From " MBOX separator line
4411 so the .eml file is really in mbox format (even though
4412 most programs do not really care). Patch from Alex Miller.
4414 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4415 The latter is generated from $received_to and is only set if the
4416 message has one envelope recipient. SA can use these headers,
4417 obviously out-of-the-box. Patch from Alex Miller.
4419 PH/08 The ${def test on a variable was returning false if the variable's
4420 value was "0", contrary to what the specification has always said!
4421 The result should be true unless the variable is empty.
4423 PH/09 The syntax error of a character other than { following "${if
4424 def:variable_name" (after optional whitespace) was not being diagnosed.
4425 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4426 accidental colon was present, for example, could give incorrect results.
4428 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4429 result is used (not including appendfile, where other changes are about
4432 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4433 This involved changing a lot of size variables from int to off_t. It
4434 should work with maildirs and everything.
4436 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4437 spamd dying while we are connected to it.
4439 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4442 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4443 function for BDB 4.3.
4445 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4447 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4448 the sources for PCRE have been re-arranged and more files are now
4451 PH/15 The code I had for printing potentially long long variables in PH/11
4452 above was not the best (it lost precision). The length of off_t variables
4453 is now inspected at build time, and an appropriate printing format (%ld
4454 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4455 to be "long long int" or "long int". This is needed for the internal
4456 formatting function string_vformat().
4458 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4459 the configuration file to be ":syslog", then the script "guesses" where
4460 the logs files are, rather than using the compiled in default. In our
4461 case the guess is not the same as the compiled default, so the script
4462 suddenly stopped working when I started to use syslog. The patch checks
4463 to see if log_file_path is "". If so, it attempts to read it from exim
4464 with no configuration file to get the compiled in version, before it
4465 falls back to the previous guessing code."
4467 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4468 implementing BATV in an Exim configuration. See NewStuff for the gory
4471 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4472 Makefile that are specific to HP-UX.
4474 PH/18 If the "use_postmaster" option was set for a recipient callout together
4475 with the "random" option, the postmaster address was used as the MAIL
4476 FROM address for the random test, but not for the subsequent recipient
4477 test. It is now used for both.
4479 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4480 patch removes a few documentation additions to RFC 3028, because the
4481 latest draft now contains them. It adds the new en;ascii-case comparator
4482 and a new error check for 8bit text in MIME parts. Comparator and
4483 require names are now matched exactly. I enabled the subaddress
4484 extension, but it is not well tested yet (read: it works for me)."
4486 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4487 rework some of the code of TK/09 above to avoid the hardwired use of
4488 "%lld" and "long long". Replaced the call to snprintf() with a call to
4491 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4492 records point to non-existent hosts", "retry timeout exceeded", and
4493 "retry time not reached for any host after a long failure period".
4495 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4496 experimental DomainKeys support:
4498 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4499 (2) On an error such as an illegally used "control", the wrong name for
4500 the control was given.
4502 These problems did NOT occur unless DomainKeys support was compiled.
4504 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4506 PH/24 Added ${if match_ip condition.
4508 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4509 databases so that it will be absolutely obvious if a crash occurs in the
4510 DB library. This is a regular occurrence (often caused by mis-matched
4513 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4514 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4515 picked up on a user's system that detects such things. There doesn't seem
4516 to be a gcc warning option for this - only an attribute that has to be
4517 put on the function's prototype. It seems that in Fedora Core 4 they have
4518 set this on a number of new functions. No doubt there will be more in due
4521 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4522 specify a transport. However, if an address that was verified by such a
4523 router was the subject of a callout, Exim crashed because it tried to
4524 read the rcpt_include_affixes from the non-existent transport. Now it
4525 just assumes that the setting of that option is false. This bug was
4526 introduced by 4.51/PH/31.
4528 PH/28 Changed -d+all to exclude +memory, because that information is very
4529 rarely of interest, but it makes the output a lot bigger. People tend to
4530 do -d+all out of habit.
4532 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4533 code in os-type was giving problems when libc.so lives in lib64, like on
4536 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4537 aren't the modern standard, and it seems that some systems' include files
4538 don't always have them. Exim was already checking for some of the newer
4539 ones like T_AAAA, and defining it itself. I've added checks for all the
4540 record types that Exim uses.
4542 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4543 not automatically generating a new one, as it is supposed to. This
4544 prevented TLS from working. If the file did exist, but contained invalid
4545 data, a new version was generated, as expected. It was only the case of a
4546 non-existent file that was broken.
4548 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4549 with a change in libdomainkeys > 0.64.
4551 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4552 from DNS. If the selector record carries the flag, it now has
4553 precedence over the domain-wide flag.
4555 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4557 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4558 the use of a port name that isn't defined in /etc/services) occurred, the
4559 message was deferred as in a normal delivery, and thus remained on the
4560 spool, instead of being failed because of the mua_wrapper setting. This
4561 is now fixed, and I tidied up some of the mua_wrapper messages at the
4564 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4565 the messages in a hash of arrays rather than using individual hashes.
4566 This is a bit cleaner and results in dramatic memory savings, albeit
4567 at a slight CPU cost.
4569 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4570 as requested by Marc Sherman.
4572 SC/10 Eximstats: added histograms for user specified patterns as requested
4575 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4577 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4578 fopen() in the content-scanning modules that did not already have it.
4584 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4585 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4587 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4589 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4591 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4592 to test Sieve filters that use "vacation".
4594 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4595 that changes the way the GnuTLS parameters are stored in the cache file.
4596 The new format can be generated externally. For backward compatibility,
4597 if the data in the cache doesn't make sense, Exim assumes it has read an
4598 old-format file, and it generates new data and writes a new file. This
4599 means that you can't go back to an older release without removing the
4602 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4603 work if there are any delivery delays because "one_time" forces the
4604 parent to be marked "delivered", so its unseen clone is never tried
4605 again. For this reason, Exim now forbids the simultaneous setting of
4608 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4609 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4610 entry if you want to know the details. The fix, however, neglected to
4611 consider the case when local delivery batching is involved. The test for
4612 "previously delivered" was not happening when checking to see if an
4613 address could be batched with a previous (undelivered) one; under
4614 certain circumstances this could lead to multiple deliveries to the same
4617 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4618 in its include files, and this causes problems building Exim.
4620 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4621 header_syntax) but Exim was just ignoring anything given after a slash.
4622 In particular, this caused confusion with an attempt to use "verify =
4623 reverse_host_lookup/defer_ok". An error is now given when options are
4624 supplied for verify items that do not have them. (Maybe reverse_host_
4625 lookup should have a defer_ok option, but that's a different point.)
4627 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4628 defined by RFC 821) to 2048, because there were problems with some AUTH
4629 commands, and RFC 1869 says the size should be increased for extended
4630 SMTP commands that take arguments.
4632 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4635 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4636 "unknown" error; now it says that the functionality isn't in the binary.
4638 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4639 an address' error message when a string expansion fails (syntax or
4640 whatever). Otherwise the password may appear in the log. Following change
4641 PH/42 below, there is no longer a chance of it appearing in a bounce
4644 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4646 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4647 of its addresses was ever tried. (Bugzilla bug #2.)
4649 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4650 the result incorrectly in the debug output. (It correctly added a newline
4651 to what was transported.)
4653 TF/01 Added $received_time.
4655 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4656 commented out examples of how to interface to a virus scanner and to
4657 SpamAssassin. Also added commented examples of av_scanner and
4658 spamd_address settings.
4660 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4661 and controls are allowed in which ACLs. There were a couple of minor
4662 errors. Some of the entries in the conditions table (which is a table of
4663 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4664 negation of where the condition IS allowed.
4666 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4668 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4669 header file does not have a version number, so I've had to invent a new
4670 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4671 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4672 radiusclient), but it was contributed by a Radius user.
4674 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4675 files or directories, for OpenSSL.
4677 PH/19 When an Exim process that is running as root has to create an Exim log
4678 file, it does so in a subprocess that runs as exim:exim so as to get the
4679 ownership right at creation (otherwise, other Exim processes might see
4680 the file with the wrong ownership). There was no test for failure of this
4681 fork() call, which would lead to the process getting stuck as it waited
4682 for a non-existent subprocess. Forks do occasionally fail when resources
4683 run out. I reviewed all the other calls to fork(); they all seem to check
4686 PH/20 When checking for unexpected SMTP input at connect time (before writing
4687 the banner), Exim was not dealing correctly with a non-positive return
4688 from the read() function. If the client had disconnected by this time,
4689 the result was a log entry for a synchronization error with an empty
4690 string after "input=" when read() returned zero. If read() returned -1
4691 (an event I could not check), uninitialized data bytes were printed.
4692 There were reports of junk text (parts of files, etc) appearing after
4695 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4697 PH/22 Added support for macro redefinition, and (re)definition in between
4698 driver and ACL definitions.
4700 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4701 forgetting to use the resulting value; it was using the unexpanded value.
4703 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4704 hadn't been configured. The fix is from Juergen Kreileder, who
4705 understands it better than I do:
4707 "Here's what I see happening with three configured cyrus_sasl
4708 authenticators configured (plain, login, cram-md5):
4710 On startup auth_cyrus_sasl_init() gets called for each of these.
4711 This means three calls to sasl_listmech() without a specified mech_list.
4712 => SASL tests which mechs of all available mechs actually work
4713 => three warnings about OTP not working
4714 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4716 With the patch, sasl_listmech() also gets called three times. But now
4717 SASL's mech_list option is set to the server_mech specified in the the
4718 authenticator. Or in other words, the answer from sasl_listmech()
4719 gets limited to just the mech you're testing for (which is different
4721 => the return list contains just 'plain' or 'login', 'cram-md5' or
4722 nothing depending on the value of ob->server_mech.
4724 I've just tested the patch: Authentication still works fine,
4725 unavailable mechs specified in the exim configuration are still
4726 caught, and the auth.log warnings about OTP are gone."
4728 PH/25 When debugging is enabled, the contents of the command line are added
4729 to the debugging output, even when log_selector=+arguments is not
4732 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4733 answer is "GNU", and only if the return is "GNU/something" is the answer
4736 PH/27 $acl_verify_message is now set immediately after the failure of a
4737 verification in an ACL, and so is available in subsequent modifiers. In
4738 particular, the message can be preserved by coding like this:
4740 warn !verify = sender
4741 set acl_m0 = $acl_verify_message
4743 Previously, $acl_verify_message was set only while expanding "message"
4744 and "log_message" when a very denied access.
4746 PH/28 Modified OS/os.c-Linux with
4748 -#ifndef OS_LOAD_AVERAGE
4749 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4751 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4752 nomenclature these days.)
4754 PH/29 Installed patch from the Sieve maintainer that adds the options
4755 sieve_useraddress and sieve_subaddress to the redirect router.
4757 PH/30 In these circumstances:
4758 . Two addresses routed to the same list of hosts;
4759 . First host does not offer TLS;
4760 . First host accepts first address;
4761 . First host gives temporary error to second address;
4762 . Second host offers TLS and a TLS session is established;
4763 . Second host accepts second address.
4764 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4765 and peerdn, if requested) that were in fact used only for the second
4768 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4769 attention to any local part prefix or suffix that was matched by the
4770 router that accepted the address. It now behaves in the same way as it
4771 does for delivery: the affixes are removed from the local part unless
4772 rcpt_include_affixes is set on the transport.
4774 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4775 timeout during the DATA phase of an incoming message.
4777 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4778 applied a patch sanctioned by the Sieve maintainer.
4780 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4781 the uid or gid is negative. A case of a negative gid caused this to be
4782 noticed. The fix allows for either to be negative.
4784 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4785 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4788 PH/36 The change PH/12 above was broken. Fixed it.
4790 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4791 the grounds that routing the same address twice would always produce the
4792 same answer. This might have been true once, but it is certainly no
4793 longer true now. Routing a child address may depend on the previous
4794 routing that produced that child. Some complicated redirection strategies
4795 went wrong when messages had multiple recipients, and made Exim's
4796 behaviour dependent on the order in which the addresses were given.
4798 I have moved the duplicate checking until after the routing is complete.
4799 Exim scans the addresses that are assigned to local and remote
4800 transports, and removes any duplicates. This means that more work will be
4801 done, as duplicates will always all be routed, but duplicates are
4802 presumably rare, so I don't expect this is of any significance.
4804 For deliveries to pipes, files, and autoreplies, the duplicate checking
4805 still happens during the routing process, since they are not going to be
4808 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4809 It corrects a timeout issue with spamd. This is Ian's comment: "The
4810 background is that sometimes spamd either never reads data from a
4811 connection it has accepted, or it never writes response data. The exiscan
4812 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4813 blindly assumes that writes won't block so it may never time out."
4815 PH/39 Allow G after quota size as well as K and M.
4817 PH/40 The value set for $authenticated_id in an authenticator may not contain
4818 binary zeroes or newlines because the value is written to log lines and
4819 to spool files. There was no check on this. Now the value is run through
4820 the string_printing() function so that such characters are converted to
4821 printable escape sequences.
4823 PH/41 $message_linecount is a new variable that contains the total number of
4824 lines in the message. Compare $body_linecount, which is the count for the
4827 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4828 bounce and delay warning messages, except in certain special cases, which
4831 (a) An SMTP error message from a remote host;
4832 (b) A message specified in a :fail: redirection;
4833 (c) A message specified in a "fail" command in a system filter;
4834 (d) A message specified in a FAIL return from the queryprogram router;
4835 (e) A message specified by the cannot_route_message router option.
4837 In these cases only, Exim does include the error details in bounce and
4838 warning messages. There are also a few cases where bland messages such
4839 as "unrouteable address" or "local delivery error" are given.
4841 PH/43 $value is now also set for the "else" part of a ${run expansion.
4843 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4844 being worked on, but at least Exim now implements the latest version to
4847 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4848 process to complete was treated as a delivery failure, a timeout while
4849 writing the message to the pipe was logged, but erroneously treated as a
4850 successful delivery. Such timeouts include transport filter timeouts. For
4851 consistency with the overall process timeout, these timeouts are now
4852 treated as errors, giving rise to delivery failures by default. However,
4853 there is now a new Boolean option for the pipe transport called
4854 timeout_defer, which, if set TRUE, converts the failures into defers for
4855 both kinds of timeout. A transport filter timeout is now identified in
4858 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4859 systems where "make" and "gmake" are different, calling "gmake" at top
4860 level broke things. I've arranged for the value of $(MAKE) to be passed
4861 from the Makefile to this script so that it can call the same version of
4865 A note about Exim versions 4.44 and 4.50
4866 ----------------------------------------
4868 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4869 changes of various kinds. As a consequence, a big documentation update was
4870 needed. This delayed the release for rather longer than seemed good, especially
4871 in the light of a couple of (minor) security issues. Therefore, the changes
4872 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4873 release. So 4.44 and 4.50 are in effect two different branches that both start
4876 I have left the 4.50 change log unchanged; it contains all the changes since
4877 4.43. The change log for 4.44 is below; many of its items are identical to
4878 those for 4.50. This seems to be the most sensible way to preserve the
4879 historical information.
4885 1. Minor wording change to the doc/README.SIEVE file.
4887 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4888 computation of the current number of files was incorrect.
4890 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4891 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4894 4. Give more explanation in the error message when the command for a transport
4895 filter fails to execute.
4897 5. There are several places where Exim runs a non-Exim command in a
4898 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4899 was being done only for the command run by the queryprogram router. It is
4900 now done for all such subprocesses. The other cases are: ${run, transport
4901 filters, and the commands run by the lmtp and pipe transports.
4903 6. Added CONFIGURE_GROUP build-time option.
4905 7. Some older OS have a limit of 256 on the maximum number of file
4906 descriptors. Exim was using setrlimit() to set 1000 as a large value
4907 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4908 systems. I've change it so that if it can't get 1000, it tries for 256.
4910 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4911 was an oversight, and furthermore, ever since the addition of extra
4912 controls (e.g. 4.43/32), the checks on when to allow different forms of
4913 "control" were broken. There should now be diagnostics for all cases when a
4914 control that does not make sense is encountered.
4916 9. Added the /retain_sender option to "control=submission".
4918 10. $recipients is now available in the predata ACL (oversight).
4920 11. Tidy the search cache before the fork to do a delivery from a message
4921 received from the command line. Otherwise the child will trigger a lookup
4922 failure and thereby defer the delivery if it tries to use (for example) a
4923 cached ldap connection that the parent has called unbind on.
4925 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4926 of $address_data from the recipient verification was clobbered by the
4927 sender verification.
4929 13. The value of address_data from a sender verification is now available in
4930 $sender_address_data in subsequent conditions in the ACL statement.
4932 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4934 15. Added a new option "connect=<time>" to callout options, to set a different
4937 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4938 was its contents. (It was OK if the option was not defined at all.)
4940 17. A "Completed" log line is now written for messages that are removed from
4941 the spool by the -Mrm option.
4943 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4944 information about exactly what failed.
4946 19. Added -dd to debug only the daemon process.
4948 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4949 handles timeouts, both on the server side and network timeouts. Renamed the
4950 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4952 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4953 It is now set to "smtps".
4955 22. $host_address is now set to the target address during the checking of
4956 ignore_target_hosts.
4958 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4959 being passed; this would have caused $sender_host_name to have been used if
4960 matching the list had actually called for a host name (not very likely,
4961 since this list is usually IP addresses). A host name is now passed as
4964 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4965 code that specifies a non-restarting handler (typically sigaction() in
4966 modern systems) in an attempt to fix a rare and obscure crash bug.
4968 25. Narrowed the window for a race in the daemon that could cause it to ignore
4969 SIGCHLD signals. This is not a major problem, because they are used only to
4970 wake it up if nothing else does.
4972 26. A malformed maildirsize file could cause Exim to calculate negative values
4973 for the mailbox size or file count. Odd effects could occur as a result.
4974 The maildirsize information is now recalculated if the size or filecount
4977 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4978 support for a long time. Removed HAVE_SYS_VFS_H.
4980 28. Installed the latest version of exipick from John Jetmore.
4982 29. In an address list, if the pattern was not a regular expression, an empty
4983 subject address (from a bounce message) matched only if the pattern was an
4984 empty string. Non-empty patterns were not even tested. This was the wrong
4985 because it is perfectly reasonable to use an empty address as part of a
4986 database query. An empty address is now tested by patterns that are
4987 lookups. However, all the other forms of pattern expect the subject to
4988 contain a local part and a domain, and therefore, for them, an empty
4989 address still always fails if the pattern is not itself empty.
4991 30. Exim went into a mad DNS loop when attempting to do a callout where the
4992 host was specified on an smtp transport, and looking it up yielded more
4993 than one IP address.
4995 31. Re-factored the code for checking spool and log partition space into a
4996 function that finds that data and another that does the check. The former
4997 is then used to implement four new variables: $spool_space, $log_space,
4998 $spool_inodes, and $log_inodes.
5000 32. The RFC2047 encoding function was originally intended for short strings
5001 such as real names; it was not keeping to the 75-character limit for
5002 encoded words that the RFC imposes. It now respects the limit, and
5003 generates multiple encoded words if necessary. To be on the safe side, I
5004 have increased the buffer size for the ${rfc2047: expansion operator from
5007 33. It is now permitted to omit both strings after an "if" condition; if the
5008 condition is true, the result is "true". As before, when the second string
5009 is omitted, a false condition yields an empty string. This makes it less
5010 cumbersome to write custom ACL and router conditions.
5012 34. Failure to deliver a bounce message always caused it to be frozen, even if
5013 there was an errors_to setting on the router. The errors_to setting is now
5016 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5017 canonical form (fully expanded) before being placed in
5018 $sender_host_address.
5020 36. The table in the code that translates DNS record types into text (T_A to
5021 "A" for instance) was missing entries for NS and CNAME. It is just possible
5022 that this could have caused confusion if both these types were looked up
5023 for the same domain, because the text type is used as part of Exim's
5024 per-process caching. But the chance of anyone hitting this buglet seems
5027 37. The dnsdb lookup has been extended in a number of ways.
5029 (1) There is a new type, "zns", which walks up the domain tree until it
5030 finds some nameserver records. It should be used with care.
5032 (2) There is a new type, "mxh", which is like "mx" except that it returns
5033 just the host names, not the priorities.
5035 (3) It is now possible to give a list of domains (or IP addresses) to be
5036 looked up. The behaviour when one of the lookups defers can be
5037 controlled by a keyword.
5039 (4) It is now possible to specify the separator character for use when
5040 multiple records are returned.
5042 38. The dnslists ACL condition has been extended: it is now possible to supply
5043 a list of IP addresses and/or domains to be looked up in a particular DNS
5046 39. Added log_selector=+queue_time_overall.
5048 40. When running the queue in the test harness, wait just a tad after forking a
5049 delivery process, to get repeatability of debugging output.
5051 41. Include certificate and key file names in error message when GnuTLS fails
5052 to set them up, because the GnuTLS error message doesn't include the name
5053 of the failing file when there is a problem reading it.
5055 42. Allow both -bf and -bF in the same test run.
5057 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5059 44. The "Exiscan patch" is now merged into the mainline Exim source.
5061 45. Sometimes the final signoff response after QUIT could fail to get
5062 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5063 < 0 before doing a fflush(). This bug looks as though it goes back to the
5064 introduction of TLS in release 3.20, but "sometimes" must have been rare
5065 because the tests only now provoked it.
5067 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5068 (this can affect the format of dates).
5070 47. exim_tidydb, when checking for the continued existence of a message for
5071 which it has found a message-specific retry record, was not finding
5072 messages that were in split spool directories. Consequently, it was
5073 deleting retry records that should have stayed in existence.
5075 48. Steve fixed some bugs in eximstats.
5077 49. The SPA authentication driver was not abandoning authentication and moving
5078 on to the next authenticator when an expansion was forced to fail,
5079 contradicting the general specification for all authenticators. Instead it
5080 was generating a temporary error. It now behaves as specified.
5082 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5083 (the order specifies the preference for clients). The order is now AES256,
5084 AES128, 3DES, ARCFOUR128.
5086 51. Small patch to Sieve code - explicitly set From: when generating an
5089 52. Exim crashed if a remote delivery caused a very long error message to be
5090 recorded - for instance if somebody sent an entire SpamAssassin report back
5091 as a large number of 550 error lines. This bug was coincidentally fixed by
5092 increasing the size of one of Exim's internal buffers (big_buffer) that
5093 happened as part of the Exiscan merge. However, to be on the safe side, I
5094 have made the code more robust (and fixed the comments that describe what
5097 53. Now that there can be additional text after "Completed" in log lines (if
5098 the queue_time_overall log selector is set), a one-byte patch to exigrep
5099 was needed to allow it to recognize "Completed" as not the last thing in
5102 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5103 patch that reportedly fixes this has been added. I am not expert enough to
5104 create a test for it. This is what the patch creator wrote:
5106 "I found a little strange behaviour of ldap code when working with
5107 Windows 2003 AD Domain, where users was placed in more than one
5108 Organization Units. When I tried to give exim partial DN, the exit code
5109 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5110 But simultaneously result of request was absolutely normal ldap result,
5111 so I produce this patch..."
5113 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5114 REFERENCE, so I have modified the code to exclude the patch when that macro
5117 55. Some experimental protocols are using DNS PTR records for new purposes. The
5118 keys for these records are domain names, not reversed IP addresses. The
5119 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5120 leaves it alone. Component reversal etc. now happens only for IP addresses.
5123 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5125 57. Double the size of the debug message buffer (to 2048) so that more of very
5126 long debug lines gets shown.
5128 58. The exicyclog utility now does better if the number of log files to keep
5129 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5131 59. Two changes related to the smtp_active_hostname option:
5133 (1) $smtp_active_hostname is now available as a variable.
5134 (2) The default for smtp_banner uses $smtp_active_hostname instead
5135 of $primary_hostname.
5137 60. The host_aton() function is supposed to be passed a string that is known
5138 to be a valid IP address. However, in the case of IPv6 addresses, it was
5139 not checking this. This is a hostage to fortune. Exim now panics and dies
5140 if the condition is not met. A case was found where this could be provoked
5141 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5142 components; fortuitously, this particular loophole had already been fixed
5143 by change 4.50/55 above.
5145 If there are any other similar loopholes, the new check in host_aton()
5146 itself should stop them being exploited. The report I received stated that
5147 data on the command line could provoke the exploit when Exim was running as
5148 exim, but did not say which command line option was involved. All I could
5149 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5150 running as the user.
5153 61. There was a buffer overflow vulnerability in the SPA authentication code
5154 (which came originally from the Samba project). I have added a test to the
5155 spa_base64_to_bits() function which I hope fixes it.
5158 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5159 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5161 63. The daemon start-up calls getloadavg() while still root for those OS that
5162 need the first call to be done as root, but it missed one case: when
5163 deliver_queue_load_max is set with deliver_drop_privilege. This is
5164 necessary for the benefit of the queue runner, because there is no re-exec
5165 when deliver_drop_privilege is set.
5167 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5168 This has been fixed.
5170 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5171 was in use, was not putting the data itself into the right store pool;
5172 consequently, it could be overwritten for a subsequent message in the same
5173 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5176 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5178 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5179 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5180 the function were treating the return as a boolean value, which happened to
5181 work because 0=false and not-0=true, but is not correct code.
5183 68. The host_aton() function was not handling scoped IPv6 addresses (those
5184 with, for example, "%eth0" on the end) correctly.
5186 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5187 negated items (that is, ~something) in unsigned ints. Some compilers
5188 apparently mutter when there is no cast.
5190 70. If an address verification called from an ACL failed, and did not produce a
5191 user-specific message (i.e. there was only a "system" message), nothing was
5192 put in $acl_verify_message. In this situation, it now puts the system
5195 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5196 session; change 4.31/43 added the unwanted input to the log line - except
5197 that it did not do this in the start of session case. It now does.
5199 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5200 This is wrong and can cause the other end to generate a synchronization
5201 error if it is another Exim or anything else that does the synchronization
5202 check. A QUIT command is no longer sent after a timeout.
5204 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5205 during host lookups.
5207 74. The defer_ok option of callout verification was not working if it was used
5208 when verifying addresses in header lines, that is, for this case:
5210 verify = header_sender/callout=defer_ok
5212 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5213 those file descriptors could be used for SMTP connections. If anything
5214 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5215 could be sent to the SMTP client, causing chaos. The daemon now opens
5216 stdin, stdout, and stderr to /dev/null when it puts itself into the
5219 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5220 log by default. The user can override this with suitable Perl magic.
5222 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5223 the log message when discard triggers, was not working for the DATA ACL or
5224 for the non-SMTP ACL.
5226 78. Error message wording change in sieve.c.
5228 79. If smtp_accept_max_per_host was set, the number of connections could be
5229 restricted to fewer than expected, because the daemon was trying to set up
5230 a new connection before checking whether the processes handling previous
5231 connections had finished. The check for completed processes is now done
5232 earlier. On busy systems, this bug wouldn't be noticed because something
5233 else would have woken the daemon, and it would have reaped the completed
5236 80. If a message was submitted locally by a user whose login name contained one
5237 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5238 It caused a spool format error. I have fixed the spool reading code. A
5239 related problem was that the "from" clause in the Received: line became
5240 illegal because of the space(s). It is now covered by ${quote_local_part.
5242 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5245 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5247 83. Patch from Sieve maintainer to fix unterminated string problem in
5248 "vacation" handling.
5250 84. Some minor changes to the Linux configuration files to help with other
5251 OS variants using glibc.
5253 85. One more patch for Sieve to update vacation handling to latest spec.
5256 ----------------------------------------------------
5257 See the note above about the 4.44 and 4.50 releases.
5258 ----------------------------------------------------
5264 1. Change 4.43/35 introduced a bug that caused file counts to be
5265 incorrectly computed when quota_filecount was set in an appendfile
5268 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5269 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5272 3. Give more explanation in the error message when the command for a transport
5273 filter fails to execute.
5275 4. There are several places where Exim runs a non-Exim command in a
5276 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5277 was being done only for the command run by the queryprogram router. It is
5278 now done for all such subprocesses. The other cases are: ${run, transport
5279 filters, and the commands run by the lmtp and pipe transports.
5281 5. Some older OS have a limit of 256 on the maximum number of file
5282 descriptors. Exim was using setrlimit() to set 1000 as a large value
5283 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5284 systems. I've change it so that if it can't get 1000, it tries for 256.
5286 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5287 was an oversight, and furthermore, ever since the addition of extra
5288 controls (e.g. 4.43/32), the checks on when to allow different forms of
5289 "control" were broken. There should now be diagnostics for all cases when a
5290 control that does not make sense is encountered.
5292 7. $recipients is now available in the predata ACL (oversight).
5294 8. Tidy the search cache before the fork to do a delivery from a message
5295 received from the command line. Otherwise the child will trigger a lookup
5296 failure and thereby defer the delivery if it tries to use (for example) a
5297 cached ldap connection that the parent has called unbind on.
5299 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5300 of $address_data from the recipient verification was clobbered by the
5301 sender verification.
5303 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5304 was its contents. (It was OK if the option was not defined at all.)
5306 11. A "Completed" log line is now written for messages that are removed from
5307 the spool by the -Mrm option.
5309 12. $host_address is now set to the target address during the checking of
5310 ignore_target_hosts.
5312 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5313 being passed; this would have caused $sender_host_name to have been used if
5314 matching the list had actually called for a host name (not very likely,
5315 since this list is usually IP addresses). A host name is now passed as
5318 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5319 code that specifies a non-restarting handler (typically sigaction() in
5320 modern systems) in an attempt to fix a rare and obscure crash bug.
5322 15. Narrowed the window for a race in the daemon that could cause it to ignore
5323 SIGCHLD signals. This is not a major problem, because they are used only to
5324 wake it up if nothing else does.
5326 16. A malformed maildirsize file could cause Exim to calculate negative values
5327 for the mailbox size or file count. Odd effects could occur as a result.
5328 The maildirsize information is now recalculated if the size or filecount
5331 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5332 support for a long time. Removed HAVE_SYS_VFS_H.
5334 18. Updated exipick to current release from John Jetmore.
5336 19. Allow an empty sender to be matched against a lookup in an address list.
5337 Previously the only cases considered were a regular expression, or an
5340 20. Exim went into a mad DNS lookup loop when doing a callout where the
5341 host was specified on the transport, if the DNS lookup yielded more than
5344 21. The RFC2047 encoding function was originally intended for short strings
5345 such as real names; it was not keeping to the 75-character limit for
5346 encoded words that the RFC imposes. It now respects the limit, and
5347 generates multiple encoded words if necessary. To be on the safe side, I
5348 have increased the buffer size for the ${rfc2047: expansion operator from
5351 22. Failure to deliver a bounce message always caused it to be frozen, even if
5352 there was an errors_to setting on the router. The errors_to setting is now
5355 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5356 canonical form (fully expanded) before being placed in
5357 $sender_host_address.
5359 24. Updated eximstats to version 1.33
5361 25. Include certificate and key file names in error message when GnuTLS fails
5362 to set them up, because the GnuTLS error message doesn't include the name
5363 of the failing file when there is a problem reading it.
5365 26. Expand error message when OpenSSL has problems setting up cert/key files.
5368 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5369 (this can affect the format of dates).
5371 28. exim_tidydb, when checking for the continued existence of a message for
5372 which it has found a message-specific retry record, was not finding
5373 messages that were in split spool directories. Consequently, it was
5374 deleting retry records that should have stayed in existence.
5376 29. eximstats updated to version 1.35
5377 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5378 1.35 - bugfix such that pie charts by volume are generated correctly
5380 30. The SPA authentication driver was not abandoning authentication and moving
5381 on to the next authenticator when an expansion was forced to fail,
5382 contradicting the general specification for all authenticators. Instead it
5383 was generating a temporary error. It now behaves as specified.
5385 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5386 (the order specifies the preference for clients). The order is now AES256,
5387 AES128, 3DES, ARCFOUR128.
5389 31. Small patch to Sieve code - explicitly set From: when generating an
5392 32. Exim crashed if a remote delivery caused a very long error message to be
5393 recorded - for instance if somebody sent an entire SpamAssassin report back
5394 as a large number of 550 error lines. This bug was coincidentally fixed by
5395 increasing the size of one of Exim's internal buffers (big_buffer) that
5396 happened as part of the Exiscan merge. However, to be on the safe side, I
5397 have made the code more robust (and fixed the comments that describe what
5400 33. Some experimental protocols are using DNS PTR records for new purposes. The
5401 keys for these records are domain names, not reversed IP addresses. The
5402 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5403 leaves it alone. Component reversal etc. now happens only for IP addresses.
5406 34. The host_aton() function is supposed to be passed a string that is known
5407 to be a valid IP address. However, in the case of IPv6 addresses, it was
5408 not checking this. This is a hostage to fortune. Exim now panics and dies
5409 if the condition is not met. A case was found where this could be provoked
5410 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5411 components; fortuitously, this particular loophole had already been fixed
5412 by change 4.50/55 or 4.44/33 above.
5414 If there are any other similar loopholes, the new check in host_aton()
5415 itself should stop them being exploited. The report I received stated that
5416 data on the command line could provoke the exploit when Exim was running as
5417 exim, but did not say which command line option was involved. All I could
5418 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5419 running as the user.
5422 35. There was a buffer overflow vulnerability in the SPA authentication code
5423 (which came originally from the Samba project). I have added a test to the
5424 spa_base64_to_bits() function which I hope fixes it.
5427 36. The daemon start-up calls getloadavg() while still root for those OS that
5428 need the first call to be done as root, but it missed one case: when
5429 deliver_queue_load_max is set with deliver_drop_privilege. This is
5430 necessary for the benefit of the queue runner, because there is no re-exec
5431 when deliver_drop_privilege is set.
5433 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5434 was in use, was not putting the data itself into the right store pool;
5435 consequently, it could be overwritten for a subsequent message in the same
5436 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5439 38. Sometimes the final signoff response after QUIT could fail to get
5440 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5441 < 0 before doing a fflush(). This bug looks as though it goes back to the
5442 introduction of TLS in release 3.20, but "sometimes" must have been rare
5443 because the tests only now provoked it.
5449 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5450 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5451 Now it returns an integer. A number of calls were still expecting a T/F
5452 return. Fortuitously, in all cases, the tests worked in OK situations,
5453 which is the norm. However, things would have gone wrong on any write
5454 failures on the smtp file descriptor. This function is used when sending
5455 messages over SMTP and also when doing verify callouts.
5457 2. When Exim is called to do synchronous delivery of a locally submitted
5458 message (the -odf or -odi options), it no longer closes stderr before doing
5461 3. Implemented the mua_wrapper option.
5463 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5465 5. Implemented the functions header_remove(), header_testname(),
5466 header_add_at_position(), and receive_remove_recipient(), and exported them
5469 6. If an ACL "warn" statement specified the addition of headers, Exim already
5470 inserted X-ACL-Warn: at the start if there was no header name. However, it
5471 was not making this test for the second and subsequent header lines if
5472 there were newlines in the string. This meant that an invalid header could
5473 be inserted if Exim was badly configured.
5475 7. Allow an ACL "warn" statement to add header lines at the start or after all
5476 the Received: headers, as well as at the end.
5478 8. Added the rcpt_4xx retry error code.
5480 9. Added postmaster_mailfrom=xxx to callout verification option.
5482 10. Added mailfrom=xxxx to the callout verification option, for verify=
5485 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5486 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5488 12. Inserted some casts to stop certain compilers warning when using pointer
5489 differences as field lengths or precisions in printf-type calls (mostly
5490 affecting debugging statements).
5492 13. Added optional readline() support for -be (dynamically loaded).
5494 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5495 same clock tick as a message's arrival, so that its received time was the
5496 same as the "first fail" time on the retry record, and that message
5497 remained on the queue past the ultimate address timeout, every queue runner
5498 would try a delivery (because it was past the ultimate address timeout) but
5499 after another failure, the ultimate address timeout, which should have then
5500 bounced the address, did not kick in. This was a "< instead of <=" error;
5501 in most cases the first failure would have been in the next clock tick
5502 after the received time, and all would be well.
5504 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5505 being recognized when the domain list was tested by the match_domain
5506 condition in an expansion string.
5508 16. Added the ${str2b64: operator.
5510 17. Exim was always calling setrlimit() to set a large limit for the number of
5511 processes, without checking whether the existing limit was already
5512 adequate. (It did check for the limit on file descriptors.) Furthermore,
5513 errors from getrlimit() and setrlimit() were being ignored. Now they are
5514 logged to the main and panic logs, but Exim does carry on, to try to do its
5515 job under whatever limits there are.
5517 18. Imported PCRE 5.0.
5519 19. Trivial typo in log message " temporarily refused connection" (the leading
5522 20. If the log selector return_path_on_delivery was set and an address was
5523 redirected to /dev/null, the delivery process crashed because it assumed
5524 that a return path would always be set for a "successful" delivery. In this
5525 case, the whole delivery is bypassed as an optimization, and therefore no
5528 21. Internal re-arrangement: the function for sending a challenge and reading
5529 a response while authentication was assuming a zero-terminated challenge
5530 string. It's now changed to take a pointer and a length, to allow for
5531 binary data in such strings.
5533 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5535 23. Exim was not respecting finduser_retries when seeking the login of the
5536 uid under which it was called; it was always trying 10 times. (The default
5537 setting of finduser_retries is zero.) Also, it was sleeping after the final
5538 failure, which is pointless.
5540 24. Implemented tls_on_connect_ports.
5542 25. Implemented acl_smtp_predata.
5544 26. If the domain in control=submission is set empty, Exim assumes that the
5545 authenticated id is a complete email address when it generates From: or
5546 Sender: header lines.
5548 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5549 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5550 chown and chgrp in /bin and hostname in /usr/bin.
5552 28. Exim was keeping the "process log" file open after each use, just as it
5553 does for the main log. This opens the possibility of it remaining open for
5554 long periods when the USR1 signal hits a daemon. Occasional processlog
5555 errors were reported, that could have been caused by this. Anyway, it seems
5556 much more sensible not to leave this file open at all, so that is what now
5559 29. The long-running daemon process does not normally write to the log once it
5560 has entered its main loop, and it closes the log before doing so. This is
5561 so that log files can straightforwardly be renamed and moved. However,
5562 there are a couple of unusual error situations where the daemon does write
5563 log entries, and I had neglected to close the log afterwards.
5565 30. The text of an SMTP error response that was received during a remote
5566 delivery was being truncated at 512 bytes. This is too short for some of
5567 the long messages that one sometimes sees. I've increased the limit to
5570 31. It is now possible to make retry rules that apply only when a message has a
5571 specific sender, in particular, an empty sender.
5573 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5574 it possible to be selective about when SMTP synchronization is enforced.
5576 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5578 32. Implemented hosts_connection_nolog.
5580 33. Added an ACL for QUIT.
5582 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5585 35. Added mailbox_size and mailbox_filecount to appendfile.
5587 36. Added control = no_multiline_responses to ACLs.
5589 37. There was a bug in the logic of the code that waits for the clock to tick
5590 in the case where the clock went backwards by a substantial amount such
5591 that the microsecond fraction of "now" was more than the microsecond
5592 fraction of "then" (but the whole seconds number was less).
5594 38. Added support for the libradius Radius client library this is found on
5595 FreeBSD (previously only the radiusclient library was supported).
5601 1. When certain lookups returned multiple values in the form name=value, the
5602 quoting of the values was not always being done properly. Specifically:
5603 (a) If the value started with a double quote, but contained no whitespace,
5605 (b) If the value contained whitespace other than a space character (i.e.
5606 tabs or newlines or carriage returns) it was not quoted.
5607 This fix has been applied to the mysql and pgsql lookups by writing a
5608 separate quoting function and calling it from the lookup code. The fix
5609 should probably also be applied to nisplus, ibase and oracle lookups, but
5610 since I cannot test any of those, I have not disturbed their existing code.
5612 2. A hit in the callout cache for a specific address caused a log line with no
5613 reason for rejecting RCPT. Now it says "Previous (cached) callout
5614 verification failure".
5616 3. There was an off-by-one bug in the queryprogram router. An over-long
5617 return line was truncated at 256 instead of 255 characters, thereby
5618 overflowing its buffer with the terminating zero. As well as fixing this, I
5619 have increased the buffer size to 1024 (and made a note to document this).
5621 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5622 when Exim is waiting for an SMTP response from a remote server, Exim
5623 restarts its select() call on the socket, thereby resetting its timeout.
5624 This is not a problem when such interrupts are rare. Somebody set up a cron
5625 job to run exiwhat every 2 minutes, which is less than the normal select()
5626 timeout (5 or 10 minutes). This meant that the select() timeout never
5627 kicked in because it was always reset. I have fixed this by comparing the
5628 time when an interrupt arrives with the time at the start of the first call
5629 to select(). If more time than the timeout has elapsed, the interrupt is
5630 treated as a timeout.
5632 5. Some internal re-factoring in preparation for the addition of Sieve
5633 extensions (by MH). In particular, the "personal" test is moved to a
5634 separate function, and given an option for scanning Cc: and Bcc: (which is
5635 not set for Exim filters).
5637 6. When Exim created an email address using the login of the caller as the
5638 local part (e.g. when creating a From: or Sender: header line), it was not
5639 quoting the local part when it contained special characters such as @.
5641 7. Installed new OpenBSD configuration files.
5643 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5644 try to make them clearer.
5646 9. Callout options, other than the timeout value, were being ignored when
5647 verifying sender addresses in header lines. For example, when using
5649 verify = header_sender/callout=no_cache
5651 the cache was (incorrectly) being used.
5653 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5654 only the Cygwin environment.
5656 11. When return_path_on_delivery was set as a log selector, if different remote
5657 addresses in the same message used different return paths and parallel
5658 remote delivery occurred, the wrong values would sometimes be logged.
5659 (Whenever a remote delivery process finished, the return path value from
5660 the most recently started remote delivery process was logged.)
5662 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5663 lines when AUTH and/or TLS are in use. This is the "received protocol"
5664 field. Exim used to use "asmtp" for authenticated SMTP, without any
5665 indication (in the protocol name) for TLS use. Now it follows the RFC and
5666 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5667 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5668 names appear in log lines as well as in Received: header lines.
5670 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5671 extensions, and comparison tests, and to fix some bugs.
5673 14. Changes to the "personal" filter test:
5675 (1) The test was buggy in that it was just doing the equivalent of
5676 "contains" tests on header lines. For example, if a user's address was
5677 anne@some.where, the "personal" test would incorrectly be true for
5679 To: susanne@some.where
5681 This test is now done by extracting each address from the header in turn,
5682 and checking the entire address. Other tests that are part of "personal"
5683 are now done using regular expressions (for example, to check local parts
5684 of addresses in From: header lines).
5686 (2) The list of non-personal local parts in From: addresses has been
5687 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5688 taken from the Sieve specification recommendations.
5690 (3) If the message contains any header line starting with "List-" it is
5691 treated as non-personal.
5693 (4) The test for "circular" in the Subject: header line has been removed
5694 because it now seems ill-conceived.
5696 15. Minor typos in src/EDITME comments corrected.
5698 16. Installed latest exipick from John Jetmore.
5700 17. If headers_add on a router specified a text string that was too long for
5701 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5702 of string_sprintf() is now avoided.
5704 18. $message_body_size was not set (it was always zero) when running the DATA
5705 ACL and the local_scan() function.
5707 19. For the "mail" command in an Exim filter, no default was being set for
5708 the once_repeat time, causing a random time value to be used if "once" was
5709 specified. (If the value happened to be <= 0, no repeat happened.) The
5710 default is now 0s, meaning "never repeat". The "vacation" command was OK
5711 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5712 (I found it when inspecting the code).
5714 20. There is now an overall timeout for performing a callout verification. It
5715 defaults to 4 times the callout timeout, which applies to individual SMTP
5716 commands during the callout. The overall timeout applies when there is more
5717 than one host that can be tried. The timeout is checked before trying the
5718 next host. This prevents very long delays if there are a large number of
5719 hosts and all are timing out (e.g. when the network connections are timing
5720 out). The value of the overall timeout can be changed by specifying an
5721 additional sub-option for "callout", called "maxwait". For example:
5723 verify = sender/callout=5s,maxwait=20s
5725 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5726 to the end before writing, but this should make it even safer).
5728 22. Exim was forgetting that it had advertised PIPELINING for the second and
5729 subsequent messages on an SMTP connection. It was also not resetting its
5730 memory on STARTTLS and an internal HELO.
5732 23. When Exim logs an SMTP synchronization error within a session, it now
5733 records whether PIPELINING has been advertised or not.
5735 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5736 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5737 rather than long int.
5739 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5741 26. Added the never_mail option to autoreply.
5747 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5748 crash if the getsockname() call failed; this can happen if a connection is
5749 closed very soon after it is established. The problem was simply in the
5750 order in which certain operations were done, causing Exim to try to write
5751 to the SMTP stream before it had set up the file descriptor. The bug has
5752 been fixed by making things happen in the correct order.
5758 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5759 before the connection was closed, thus losing the rejection response.
5761 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5762 some early Solaris releases, but causes trouble in current releases where
5763 socklen_t is defined.
5765 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5768 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5771 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5772 that releasing the top of it at the end releases what was used for sub-
5773 expansions (unless the block got too big). However, discard this block if
5774 the first thing is a variable or header, so that we can use its block when
5775 it is dynamic (useful for very large $message_headers, for example).
5777 6. Lookups now cache *every* query, not just the most recent. A new, separate
5778 store pool is used for this. It can be recovered when all lookup caches are
5779 flushed. Lookups now release memory at the end of their result strings.
5780 This has involved some general refactoring of the lookup sources.
5782 7. Some code has been added to the store_xxx() functions to reduce the amount
5783 of flapping under certain conditions.
5785 8. log_incoming_interface used to affect only the <= reception log lines. Now
5786 it causes the local interface and port to be added to several more SMTP log
5787 lines, for example "SMTP connection from", and rejection lines.
5789 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5791 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5793 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5794 could be overwritten at the end of the current message (or the start of a
5795 new message if it was set in a HELO ACL). The value is now preserved for
5796 the duration of the SMTP connection.
5798 12. If a transport had a headers_rewrite setting, and a matching header line
5799 contained an unqualified address, that address was qualified, even if it
5800 did not match any rewriting rules. The underlying bug was that the values
5801 of the flags that permit the existence of unqualified sender and recipient
5802 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5803 non-local messages, and by -bnq for local messages) were not being
5804 preserved with the message after it was received.
5806 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5807 "next input=" as part of the text comprising the host identity instead of
5808 the correct text. The code was using the same buffer for two different
5809 strings. However, depending on which order the printing function evaluated
5810 its arguments, the bug did not always show up. Under Linux, for example, my
5811 test suite worked just fine.
5813 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5814 doesn't work with some older versions of Perl. It has been changed to "my",
5815 which in any case is probably the better facility to use.
5817 15. A really picky compiler found some instances of statements for creating
5818 error messages that either had too many or two few arguments for the format
5821 16. The size of the buffer for calls to the DNS resolver has been increased
5822 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5823 for addresses that have a lot of PTR records. This alleviates a problem; it
5824 does not fully solve it.
5826 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5827 buffer now truncates the list and logs the incident, which is the same
5828 action as happens when Exim is looking up a host name and its aliases.
5829 Previously in this situation something unpredictable would happen;
5830 sometimes it was "internal error: store_reset failed".
5832 18. If a server dropped the connection unexpectedly when an Exim client was
5833 using GnuTLS and trying to read a response, the client delivery process
5834 crashed while trying to generate an error log message.
5836 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5837 string, for example:
5839 warn message = H1: something\nH2: something
5841 the text was added as a single header line from Exim's point of view
5842 though it ended up OK in the delivered message. However, searching for the
5843 second and subsequent header lines using $h_h2: did not work. This has been
5844 fixed. Similarly, if a system filter added multiple headers in this way,
5845 the routers could not see them.
5847 20. Expanded the error message when iplsearch is called with an invalid key to
5848 suggest using net-iplsearch in a host list.
5850 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5851 ACLs are no longer actually imposed (and a message to that effect is
5854 22. If a "gecos" field in a passwd entry contained escaped characters, in
5855 particular, if it contained a \" sequence, Exim got it wrong when building
5856 a From: or a Sender: header from that name. A second bug also caused
5857 incorrect handling when an unquoted " was present following a character
5858 that needed quoting.
5860 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5861 was not being matched caselessly.
5863 24. Arranged for all hyphens in the exim.8 source to be escaped with
5866 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5867 properly. Recipient callout cache records were still being keyed to include
5868 the sender, even when use_sender was set false. This led to far more
5869 callouts that were necessary. The sender is no longer included in the key
5870 when use_sender is false.
5872 26. Added "control = submission" modifier to ACLs.
5874 27. Added the ${base62d: operator to decode base 62 numbers.
5876 28. dnsdb lookups can now access SRV records.
5878 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5879 the configuration file.
5881 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5882 (-v) mode. This makes the output for a verbose queue run more intelligible.
5884 31. Added a use_postmaster feature to recipient callouts.
5886 32. Added the $body_zerocount variable, containing the number of binary zero
5887 bytes in the message body.
5889 33. The time of last modification of the "new" subdirectory is now used as the
5890 "mailbox time last read" when there is a quota error for a maildir
5893 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5895 35. Added +ignore_unknown as a special item in host lists.
5897 36. Code for decoding IPv6 addresses in host lists is now included, even if
5898 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5899 address was recognized as an IP address, but was then not correctly decoded
5900 into binary, causing unexpected and incorrect effects when compared with
5907 1. Very minor rewording of debugging text in manualroute to say "list of
5908 hosts" instead of "hostlist".
5910 2. If verify=header_syntax was set, and a header line with an unqualified
5911 address (no domain) and a large number of spaces between the end of the
5912 name and the colon was received, the reception process suffered a buffer
5913 overflow, and (when I tested it) crashed. This was caused by some obsolete
5914 code that should have been removed. The fix is to remove it!
5916 3. When running in the test harness, delay a bit after writing a bounce
5917 message to get a bit more predictability in the log output.
5919 4. Added a call to search_tidyup() just before forking a reception process. In
5920 theory, someone could use a lookup in the expansion of smtp_accept_max_
5921 per_host which, without the tidyup, could leave open a database connection.
5923 5. Added the variables $recipient_data and $sender_data which get set from a
5924 lookup success in an ACL "recipients" or "senders" condition, or a router
5925 "senders" option, similar to $domain_data and $local_part_data.
5927 6. Moved the writing of debug_print from before to after the "senders" test
5930 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5931 problems for message scanning, either using a data ACL, or using
5932 local_scan() because the Received: header was not generated till after they
5933 were called (in order to set the time as the time of reception completion).
5934 I have revised the way this works. The header is now generated after the
5935 body is received, but before the ACL or local_scan() are called. After they
5936 are run, the timestamp in the header is updated.
5942 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5943 before starting a queue runner without re-exec. This happened only when
5944 deliver_drop_privilege was set or when the Exim user was set to root. The
5945 effect of the bug was that timeouts during subsequent deliveries caused
5946 crashes instead of being properly handled. The handler is now left at its
5947 default (and expected) setting.
5949 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5950 message, again when deliver_drop_privilege is set or Exim is run as root.
5951 The bug described in (1) was not present in this case, but the tidying up
5952 of the other signals was missing. I have made the two cases consistent.
5954 3. The ignore_target_hosts setting on a manualroute router was being ignored
5955 for hosts that were looked up using the /MX notation.
5957 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5960 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5961 operated on the sender address. After changing the $sender_address to <>
5962 for the sender address verify, Exim was re-instated it as the original
5963 (before rewriting) address, but remembering that it had rewritten it, so it
5964 wasn't rewriting it again. This bug also had the effect of breaking the
5965 sender address verification caching when the sender address was rewritten.
5967 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5968 This has been changed so that if the ip literal address matches
5969 ignore_target_hosts, the router declines.
5971 7. Added expansion conditions match_domain, match_address, and match_local_
5972 part (NOT match_host).
5974 8. The placeholder for the Received: header didn't have a length field set.
5976 9. Added code to Exim itself and to exim_lock to test for a specific race
5977 condition that could lead to file corruption when using MBX delivery. The
5978 issue is with the lockfile that is created in /tmp. If this file is removed
5979 after a process has opened it but before that process has acquired a lock,
5980 there is the potential for a second process to recreate the file and also
5981 acquire a lock. This could lead to two Exim processes writing to the file
5982 at the same time. The added code performs the same test as UW imapd; it
5983 checks after acquiring the lock that its file descriptor still refers to
5984 the same named file.
5986 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5987 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5990 11. Added the smtp_active_hostname option. If used, this will typically be made
5991 to depend on the incoming interface address. Because $interface_address is
5992 not set up until the daemon has forked a reception process, error responses
5993 that can happen earlier (such as "too many connections") no longer contain
5996 12. If an expansion in a condition on a "warn" statement fails because a lookup
5997 defers, the "warn" statement is abandoned, and the next ACL statement is
5998 processed. Previously this caused the whole ACL to be aborted.
6000 13. Added the iplsearch lookup type.
6002 14. Added ident_timeout as a log selector.
6004 15. Added tls_certificate_verified as a log selector.
6006 16. Added a global option tls_require_ciphers (compare the smtp transport
6007 option of the same name). This controls incoming TLS connections.
6009 17. I finally figured out how to make tls_require_ciphers do a similar thing
6010 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6011 before starting the TLS session.
6013 18. Tabs are now shown as \t in -bP output.
6015 19. If the log selector return_path_on_delivery was set, Exim crashed when
6016 bouncing a message because it had too many Received: header lines.
6018 20. If two routers both had headers_remove settings, and the first one included
6019 a superfluous trailing colon, the final name in the first list and the
6020 first name in the second list were incorrectly joined into one item (with a
6021 colon in the middle).
6027 1. Added -C and -D options to the exinext utility, mainly to make it easier
6028 to include in the automated testing, but these could be helpful when
6029 multiple configurations are in use.
6031 2. The exinext utility was not formatting the output nicely when there was
6032 an alternate port involved in the retry record key, nor when there was a
6033 message id as well (for retries that were specific to a specific message
6034 and a specific host). It was also confused by IPv6 addresses, because of
6035 the additional colons they contain. I have fixed the IPv4 problem, and
6036 patched it up to do a reasonable job for IPv6.
6038 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6039 delivery, the log line now contains "pipelined" if PIPELINING was used.
6041 4. An SMTP transport process used to panic and die if the bind() call to set
6042 an explicit outgoing interface failed. This has been changed; it is now
6043 treated in the same way as a connect() failure.
6045 5. A reference to $sender_host_name in the part of a conditional expansion
6046 that was being skipped was still causing a DNS lookup. This no longer
6049 6. The def: expansion condition was not recognizing references to header lines
6050 that used bh_ and bheader_.
6052 7. Added the _cache feature to named lists.
6054 8. The code for checking quota_filecount in the appendfile transport was
6055 allowing one more file than it should have been.
6057 9. For compatibility with Sendmail, the command line option
6065 and sets the incoming protocol and host name (for trusted callers). The
6066 host name and its colon can be omitted when only the protocol is to be set.
6067 Note the Exim already has two private options, -pd and -ps, that refer to
6068 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6069 "s", but I don't think that's a major issue.
6071 10. A number of refactoring changes to the code, none of which should affect
6074 (a) The number of logging options was getting close to filling up the
6075 32-bit word that was used as a bit map. I have split them into two classes:
6076 those that are passed in the argument to log_write(), and those that are
6077 only ever tested independently outside of that function. These are now in
6078 separate 32-bit words, so there is plenty of room for expansion again.
6079 There is no change in the user interface or the logging behaviour.
6081 (b) When building, for example, log lines, the code previously used a
6082 macro that called string_cat() twice, in order to add two strings. This is
6083 not really sufficiently general. Furthermore, there was one instance where
6084 it was actually wrong because one of the argument was used twice, and in
6085 one call a function was used. (As it happened, calling the function twice
6086 did not affect the overall behaviour.) The macro has been replaced by a
6087 function that can join an arbitrary number of extra strings onto a growing
6090 (c) The code for expansion conditions now uses a table and a binary chop
6091 instead of a serial search (which was left over from when there were very
6092 few conditions). Also, it now recognizes conditions like "pam" even when
6093 the relevant support is not compiled in: a suitably worded error message is
6094 given if an attempt is made to use such a condition.
6096 11. Added ${time_interval:xxxxx}.
6098 12. A bug was causing one of the ddress fields not to be passed back correctly
6099 from remote delivery subprocesses. The field in question was not being
6100 subsequently used, so this caused to problems in practice.
6102 13. Added new log selectors queue_time and deliver_time.
6104 14. Might have fixed a bug in maildirsizefile handling that threw up
6105 "unexpected character" debug warnings, and recalculated the data
6106 unnecessarily. In any case, I expanded the warning message to give more
6109 15. Added the message "Restricted characters in address" to the statements in
6110 the default ACL that block characters like @ and % in local parts.
6112 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6113 Three changes have been made:
6115 (a) There was a serious bug; a negative response to MAIL caused the whole
6116 recipient domain to be cached as invalid, thereby blocking all messages
6117 to all local parts at the same domain, from all senders. This bug has
6118 been fixed. The domain is no longer cached after a negative response to
6119 MAIL if the sender used is not empty.
6121 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6124 (c) A new callout option, "use_sender" has been added for people who want
6125 the modified behaviour.
6131 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6134 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6135 indeed breaks things for older releases.
6137 3. Added additional logging to the case where there is a problem reading data
6138 from a filter that is running in a subprocess using a pipe, in order to
6139 try to track down a specific problem.
6141 4. Testing facility fudge: when running in the test harness and attempting
6142 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6143 getting "No route to host". Convert this to a timeout.
6145 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6148 6. Some OS don't have socklen_t but use size_t instead. This affects the
6149 fifth argument of getsockopt() amongst other things. This is now
6150 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6151 can be set for individual OS. I have set it for SunOS5, OSF1, and
6152 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6153 some earlier ones do not.
6155 7. Change 4.30/15 was not doing the test caselessly.
6157 8. The standard form for an IPv6 address literal was being rejected by address
6158 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6159 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6160 this, as well as the form without the "IPv6" on the front (but only when
6161 address literals are enabled, of course).
6163 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6165 10. Exim crashed if a message with an empty sender address specified by -f
6166 encountered a router with an errors_to setting. This could be provoked only
6167 by a command such as
6171 where an empty string was supplied; "<>" did not hit this bug.
6173 11. Installed PCRE release 4.5.
6175 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6176 remained set. It is now erased.
6178 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6179 times from message ids (which are base 36 rather than the normal 62).
6181 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6182 were being counted as actual protocol errors, and logged if the log
6183 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6184 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6185 and DATA following a set of rejected RCPTs do not count as protocol errors.
6186 In other words, Exim assumes they were pipelined, though this may not
6187 actually be the case. Of course, in all cases the client gets an
6188 appropriate error code.
6190 15. If a lookup fails in an ACL condition, a message about the failure may
6191 be available; it is used if testing the ACL cannot continue, because most
6192 such messages specify what the cause of the deferral is. However, some
6193 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6194 that caused an old message to be retained and used if a later statement
6195 caused a defer, replacing the real cause of the deferral.
6197 16. If an IP address had so many PTR records that the DNS lookup buffer
6198 was not large enough to hold them, Exim could crash while trying to process
6199 the truncated data. It now detects and logs this case.
6201 17. Further to 4.21/58, another change has been made: if (and only if) the
6202 first line of a message (the first header line) ends with CRLF, a bare LF
6203 in a subsequent header line has a space inserted after it, so as not to
6204 terminate the header.
6206 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6207 unnecessarily, used atoi() instead of strtol(), and didn't check the
6208 termination when getting file sizes from file names by regex.
6210 19. Completely re-implemented the support for maildirsize files, in the light
6211 of a number of problems with the previous contributed implementation
6212 (4.30/29). In particular:
6214 . If the quota is zero, the maildirsize file is maintained, but no quota is
6217 . If the maildir directory does not exist, it is created before any attempt
6218 to write a maildirsize file.
6220 . The quota value in the file is just a cache; if the quota is changed in
6221 the transport, the new value overrides.
6223 . A regular expression is available for excluding directories from the
6226 20. The autoreply transport checks the characters in options that define the
6227 message's headers; it allows continued headers, but it was checking with
6228 isspace() after an embedded newline instead of explicitly looking for a
6231 21. If all the "regular" hosts to which an address was routed had passed their
6232 expiry times, and had not reached their retry times, the address was
6233 bounced, even if fallback hosts were defined. Now Exim should go on to try
6236 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6237 equivalent code in the SMTP transport. Some hosts send humungous responses
6238 to HELO/EHLO, more than 1024 it seems.
6240 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6241 gives compiler warnings in some environments. I've now done it "properly",
6244 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6245 (because of problems with the built-in one) was declared to return uschar *
6246 instead of char *, causing compiler failure.
6248 25. Fixed a file descriptor leak when processing alias/forward files.
6250 26. Fixed a minor format string issue in dbfn.c.
6252 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6254 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6255 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6258 29. When a sender address is verified, it is cached, to save repeating the test
6259 when there is more than one recipient in a message. However, when the
6260 verification involves a callout, it is possible for different callout
6261 options to be set for different recipients. It is too complicated to keep
6262 track of this in the cache, so now Exim always runs a verification when a
6263 callout is required, relying on the callout cache for the optimization.
6264 The overhead is duplication of the address routing, but this should not be
6267 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6268 to be verified with callout=postmaster, and the main callout worked but the
6269 postmaster check failed, the verification correctly failed. However, if a
6270 subsequent RCPT command asked for sender verification *without* the
6271 postmaster check, incorrect caching caused this verification also to fail,
6274 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6275 it was not caching the DNS options (qualify_single, search_parents) that
6276 were used when the lookup failed. A subsequent lookup with different
6277 options therefore always gave the same answer, though there were cases
6278 where it should not have. (Example: a "domains = !$mx_any" option on a
6279 dnslookup router: the "domains" option is always processed without any
6280 widening, but the router might have qualify_single set.) Now Exim uses the
6281 cached value only when the same options are set.
6283 32. Added John Jetmore's "exipick" utility to the distribution.
6285 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6286 than a timeout (e.g. a certificate is required, and is not provided), an
6287 Exim server now closes the connection immediately. Previously it waited for
6288 the client to close - but if the client is SSL, it seems that they each
6289 wait for each other, leading to a delay before one of them times out.
6291 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6292 maintained 0.8.x compatibility because I don't think many are using it, and
6293 it is clearly obsolete.
6295 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6298 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6299 client certificate was expired. A simple patch fixes this, though I don't
6300 understand the full logic of why the verify callback is called multiple
6303 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6304 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6305 which causes problems with some clients (such as the Certicom SSL Plus
6306 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6307 disables the coutermeasure allowing Eudora to connect."
6309 38. Exim was not checking that a write() to a log file succeeded. This could
6310 lead to Bad Things if a log got too big, in particular if it hit a file
6311 size limit. Exim now panics and dies if it cannot write to a log file, just
6312 as it does if it cannot open a log file.
6314 39. Modified OS/Makefile-Linux so that it now contains
6316 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6318 The two -D definitions ensure that Exim is compiled with large file
6319 support, which makes it possible to handle log files that are bigger than
6322 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6323 instance) a domain was checked against a named list that involved a lookup,
6324 causing $domain_data to be set, then another domain was checked against the
6325 same list, then the first domain was re-checked, the value of $domain_data
6326 after the final check could be wrong. In particular, if the second check
6327 failed, it could be set empty. This bug probably also applied to
6330 41. The strip_trailing_dot option was not being applied to the address given
6331 with the -f command-line option.
6333 42. The code for reading a message's header from the spool was incrementing
6334 $received_count, but never initializing it. This meant that the value was
6335 incorrect (doubled) while delivering a message in the same process in which
6336 it was received. In the most common configuration of Exim, this never
6337 happens - a fresh exec is done - but it can happen when
6338 deliver_drop_privilege is set.
6340 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6341 it now includes up to 150 characters of the unexpected data in the log
6344 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6345 and building data strings. The size of both of these buffers was 10 000
6346 bytes - far larger than anybody would *ever* want, thought I. Needless to
6347 say, somebody hit the limit. I have increased the maximum line length to
6348 20 000 and the maximum data length of concatenated lines to 100 000. I have
6349 also fixed two bugs, because there was no checking on these buffers. Tsk,
6350 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6351 buffer is too small.
6353 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6354 lsearch lookups. Now it does.
6356 46. When parsing a route_list item in a manualroute router, a fixed-length
6357 buffer was used for the list of hosts. I made this 1024 bytes long,
6358 thinking that nobody would ever have a list of hosts that long. Wrong.
6359 Somebody had a whole pile of complicated expansion conditions, and the
6360 string was silently truncated, leading to an expansion error. It turns out
6361 that it is easier to change to an unlimited length (owing to other changes
6362 that have happened since this code was originally written) than to build
6363 structure for giving a limitation error. The length of the item that
6364 expands into the list of hosts is now unlimited.
6366 47. The lsearch lookup could not handle data where the length of text line was
6367 more than 4095 characters. Such lines were truncated, leading to shortened
6368 data being returned. It should now handle lines of any length.
6370 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6371 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6374 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6375 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6376 no longer happens in comment lines. A list of replacements is now placed
6377 at the head of all of the source files, except those whose only change is
6378 to replace PERL_COMMAND in the very first #! line.
6380 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6381 messages on the queue, with a bottom-up merge sort, using code contributed
6382 by Michael Haardt. This should make operations like -bp somewhat faster on
6383 large queues. It won't affect queue runners, except when queue_run_in_order
6386 51. Installed eximstats 1.31 in the distribution.
6388 52. Added support for SRV lookups to the dnslookup router.
6390 53. If an ACL referred to $message_body or $message_body_end, the value was not
6391 reset for any messages that followed in the same SMTP session.
6393 54. The store-handling optimization for building very long strings was not
6394 differentiating between the different store pools. I don't think this
6395 actually made any difference in practice, but I've tidied it.
6397 55. While running the routers to verify a sender address, $sender_address
6398 was still set to the sender address. This is wrong, because when routing to
6399 send a bounce to the sender, it would be empty. Therefore, I have changed
6400 it so that, while verifying a sender address, $sender_address is set to <>.
6401 (There is no change to what happens when verifying a recipient address.)
6403 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6404 target A or AAAA records (if not already returned) without resetting the
6405 qualify_single or search_parents options of the DNS resolver. These are
6406 inappropriate in this case because the targets of MX and SRV records must
6407 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6408 target that, when qualified, matched something in the local domain. These
6409 two options are now turned off when doing these lookups.
6411 57. It seems that at least some releases of Reiserfs (which does not have the
6412 concept of a fixed number of inodes) returns zero and not -1 for the
6413 number of available inodes. This interacted badly with check_spool_inodes,
6414 which assumed that -1 was the "no such thing" setting. What I have done is
6415 to check that the total number of inodes is greater than zero before doing
6416 the test of how many are available.
6418 58. When a "warn" ACL statement has a log_message modifier, the message is
6419 remembered, and not repeated. This is to avoid a lot of repetition when a
6420 message has many recipients that cause the same warning to be written.
6421 However, Exim was preserving the list of already written lines for an
6422 entire SMTP session, which doesn't seem right. The memory is now reset if a
6423 new message is started.
6425 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6426 debugging output unless log_rewrite was also set.
6428 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6429 of (int)(handle) when we know that handle contains (void *)(-1).
6431 61. The Exim daemon panic-logs an error return when it closes the incoming
6432 connection. However "connection reset by peer" seems to be common, and
6433 isn't really an error worthy of noting specially, so that particular error
6436 62. When Exim is trying to find all the local interfaces, it used to panic and
6437 die if the ioctl to get the interface flags failed. However, it seems that
6438 on at least one OS (Solaris 9) it is possible to have an interface that is
6439 included in the list of interfaces, but for which you get a failure error
6440 for this call. This happens when the interface is not "plumbed" into a
6441 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6442 failure of the "get flags" call assumes that the interface is down.
6444 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6445 makes life easier for people who are doing arithmetic on fields extracted
6446 from dates, where you often get leading zeros that should not be
6447 interpreted as octal.
6449 64. Added qualify_domain to the redirect router, to override the global
6452 65. If a pathologically long header line contained very many addresses (the
6453 report of this problem mentioned 10 000) and each of them was rewritten,
6454 Exim could use up a very large amount of memory. (It kept on making new
6455 copies of the header line as it rewrote, and never released the old ones.)
6456 At the expense of a bit more processing, the header rewriting function has
6457 been changed so that it no longer eats memory in this way.
6459 66. The generation of the Received: header has been moved from the time that a
6460 message starts to be received, to the time that it finishes. The timestamp
6461 in the Received: header should now be very close to that of the <= log
6462 line. There are two side-effects of this change:
6464 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6465 logged header lines no longer include the local Received: line, because
6466 it has not yet been created. The same applies to a copy of the message
6467 that is returned to a non-SMTP sender when a message is rejected.
6469 (b) When a filter file is tested using -bf, no additional Received: header
6470 is added to the test message. After some thought, I decided that this
6473 This change does not affect the value of $received_for. It is still set
6474 after address rewriting, but before local_scan() is called.
6476 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6478 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6479 gave an unhelpful panic error message, and a defer error. I have managed to
6480 change this behaviour so that it now rejects any supplied certificate,
6481 which seems right, as the list of acceptable certificates is empty.
6483 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6484 gave an unhelpful defer error. I have not managed to make this reject any
6485 supplied certificates, but the error message it gives is "no certificate
6486 supplied", which is not helpful.
6488 70. exigrep's output now also includes lines that are not associated with any
6489 message, but which match the given pattern. Implemented by a patch from
6490 Martin Sluka, which also tidied up the Perl a bit.
6492 71. Recipient callout verification, like sender verification, was using <> in
6493 the MAIL FROM command. This isn't really the right thing, since the actual
6494 sender may affect whether the remote host accepts the recipient or not. I
6495 have changed it to use the actual sender in the callout; this means that
6496 the cache record is now keyed on a recipient/sender pair, not just the
6497 recipient address. There doesn't seem to be a real danger of callout loops,
6498 since a callout by the remote host to check the sender would use <>.
6499 [SEE ABOVE: changed after hitting problems.]
6501 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6502 temporary errors. However, in the case of such a code being given after
6503 the end of a data transmission (i.e. after ".") Exim was failing to write
6504 a retry record for the message. (Yes, there was some broken host that was
6505 actually sending 8xx at this point.)
6507 73. An unknown lookup type in a host list could cause Exim to panic-die when
6508 the list was checked. (An example that provoked this was putting <; in the
6509 middle of a list instead of at the start.) If this happened during a DATA
6510 ACL check, a -D file could be left lying around. This kind of configuration
6511 error no longer causes Exim to die; instead it causes a defer error. The
6512 incident is still logged to the main and panic logs.
6514 74. Buglet left over from Exim 3 conversion. The message "too many messages
6515 in one connection" was written to the rejectlog but not the mainlog, except
6516 when address rewriting (yes!) was being logged.
6518 75. Added write_rejectlog option.
6520 76. When a system filter was run not as root (that is, when system_filter_user
6521 was set), the values of the $n variables were not being returned to the
6522 main process; thus, they were not subsequently available in the $sn
6525 77. Added +return_path_on_delivery log selector.
6527 78. A connection timeout was being treated differently from recipients deferred
6528 when testing hosts_max_try with a message that was older than the host's
6529 retry timeout. (The host should not be counted, thus allowing all hosts to
6530 be tried at least once before bouncing.) This may have been the cause of an
6531 occasionally reported bug whereby a message would remain on the queue
6532 longer than the retry timeout, but would be bounced if a delivery was
6533 forced. I say "may" because I never totally pinned down the problem;
6534 setting up timeout/retry tests is difficult. See also the next item.
6536 79. The ultimate address timeout was not being applied to errors that involved
6537 a combination of host plus message (for example, a timeout on a MAIL
6538 command). When an address resolved to a number of possible hosts, and they
6539 were not all tried for each delivery (e.g. because of hosts_max_try), a
6540 message could remain on the queue longer than the retry timeout.
6542 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6545 81. Fixed an obscure SMTP outgoing bug which required at least the following
6546 conditions: (a) there was another message waiting for the same server;
6547 (b) the server returned 5xx to all RCPT commands in the first message so
6548 that the message was not completed; (c) the server dropped the connection
6549 or gave a negative response to the RSET that Exim sends to abort the
6550 transaction. The observed case was a dropped connection after DATA that had
6551 been sent in pipelining mode. That is, the server had advertised PIPELINING
6552 but was not implementing it correctly. The effect of the bug was incorrect
6553 behaviour, such as trying another host, and this could lead to a crash.
6559 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6560 and daemon.c were passed as pointers to ints; they should have been
6561 pointers to socklen_t variables (which are typically unsigned ints).
6563 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6566 3. Fixed a really odd bug that affected only the testing scheme; patching a
6567 certain fixed string in the binary changed the value of another string that
6568 happened to be identical to the end of the original first string.
6570 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6571 name", it returns that address as the IP address. On some operating
6572 systems (e.g. Solaris), it also passes back the IP address string as the
6573 "host name". However, on others (e.g. Linux), it passes back an empty
6574 string. Exim wasn't checking for this, and was changing the host name to an
6575 empty string, assuming it had been canonicalized.
6577 5. Although rare, it is permitted to have more than one PTR record for a given
6578 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6579 all the names associated with an address, because they do in Solaris.
6580 However, it seems that they do not in Linux for data that comes from the
6581 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6582 I found this out when I moved to a new Linux workstation and tried to run
6583 the Exim test suite.
6585 To get round this problem I have changed the code so that it now does its
6586 own call to the DNS to look up PTR records when searching for a host name.
6587 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6588 addresses that are only in /etc/hosts are still found.
6590 This behaviour is, however, controlled by an option called host_lookup_
6591 order, which defaults to "bydns:byaddr". If people want to use the other
6592 order, or indeed, just use one or the other means of lookup, they can
6593 specify it in this variable.
6595 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6596 some operating systems, this comes back from gethostbyaddr() as an empty
6597 string, and this is what Exim used to test for. However, it seems that in
6598 other systems, "." is yielded. Exim now tests for this case too.
6600 7. The values of check_spool_space and check_log_space are now held internally
6601 as a number of kilobytes instead of an absolute number of bytes. If a
6602 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6603 kilobyte. This means that much larger values can be stored.
6605 8. Exim monitor: an attempt to get the action menu when not actually pointing
6606 at a message produces an empty menu entitled "No message selected". This
6607 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6608 no entries in it ("Shell widget menu has zero width and/or height"). So I
6609 have added a single, blank menu entry in this case.
6611 9. Added ${quote_local_part.
6613 10. MIME decoding is now applied to the contents of Subject: header lines when
6616 11. Now that a reference to $sender_host_address automatically causes a reverse
6617 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6618 host lookup before query-style lookups in lists that might use this
6619 variable. This has therefore been abolished, and the "net-" prefix is no
6620 longer necessary for query-style lookups.
6622 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6623 have been a typo for LFLAGS, so it has been changed.
6625 13. The install script calls Exim with "-C /dev/null" in order to find the
6626 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6627 to be output. However, since Exim outputs its version number before the
6628 error, it didn't break the script. It just looked ugly. I fixed this by
6629 always allowing "-C /dev/null" if the caller is root.
6631 14. Ignore overlarge ACL variable number when reading spool file - insurance
6632 against a later release with more variables having written the file.
6634 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6635 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6636 this, as well as the form without the "IPv6" on the front.
6638 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6639 OS/Makefile-Darwin file.
6641 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6642 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6644 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6645 variable. However, some OS use ints for inodes. I've added cast to long int
6646 to get rid of the compiler warning.
6648 19. I had forgotten to lock out "/../" in configuration file names when
6649 ALT_CONFIG_PREFIX was set.
6651 20. Routers used for verification do not need to specify transports. However,
6652 if such a router generated a host list, and callout was configured, Exim
6653 crashed, because it could not find a port number from the (non-existent)
6654 transport. It now assumes port 25 in this circumstance.
6656 21. Added the -t option to exigrep.
6658 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6659 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6660 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6661 them is set, LOOKUP_LSEARCH is forced.
6663 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6665 24. Added sender and host information to the "rejected by local_scan()" log
6666 line; previously there was no indication of these.
6668 25. Added .include_if_exists.
6670 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6671 Linux. It turns out that not all file systems support this. Apparently some
6672 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6673 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6674 not supported on the file descriptor, is now ignored when Exim is trying to
6675 sync a directory. This applies only to Linux.
6677 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6679 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6680 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6683 29. Added support for maildirsize files from supplied patch (modified a bit).
6685 30. The use of :fail: followed by an empty string could lead Exim to respond to
6686 sender verification failures with (e.g.):
6688 550 Verification failed for <xxx>
6689 550 Sender verify failed
6691 where the first response line was missing the '-' that indicates it is not
6692 the final line of the response.
6694 31. The loop for finding the name of the user that called Exim had a hardwired
6695 limit of 10; it now uses the value of finduser_retries, which is used for
6696 all other user lookups.
6698 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6701 33. Exim was neglecting to zero errno before one call of strtol() when
6702 expanding a string and expecting an integer value. On some systems this
6703 resulted in spurious "integer overflow" errors. Also, it was casting the
6704 result into an int without checking.
6706 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6707 did not work. The code looks as if it has *never* worked, though it appears
6708 to have been documented since at least release 1.62. I have made it work.
6710 35. The "timeout_DNS" error in retry rules, also documented since at least
6711 1.62, also never worked. As it isn't clear exactly what this means, and
6712 clearly it isn't a major issue, I have abolished the feature by treating it
6713 as "timeout", and writing a warning to the main and panic logs.
6715 36. The display of retry rules for -brt wasn't always showing the error code
6718 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6719 timeout_connect_A, timeout_connect_MX.
6721 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6722 to the empty sender.
6724 39. The daemon was not analysing the content of -oX till after it had closed
6725 stderr and disconnected from the controlling terminal. This meant that any
6726 syntax errors were only noted on the panic log, and the return code from
6727 the command was 0. By re-arranging the code a little, I've made the
6728 decoding happen first, so such errors now appear on stderr, and the return
6729 code is 1. However, the actual setting up of the sockets still happens in
6730 the disconnected process, so errors there are still only recorded on the
6733 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6734 connections (as happens on some IP stacks) was logged at start up time as
6735 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6736 it from "IPv6 and IPv4", which means that two separate sockets are being
6739 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6740 says whether AF_INET or AF_INET6 was passed as an argument.
6742 42. Exiwhat output was messed up when time zones were included in log
6745 43. Exiwhat now gives more information about the daemon's listening ports,
6746 and whether -tls-on-connect was used.
6748 44. The "port" option of the smtp transport is now expanded.
6750 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6751 silently ignored. Now an error message is written to the main and panic
6754 46. There's a new ACL modifier called "logwrite" which writes to a log file
6755 as soon as it is encountered.
6757 47. Added $local_user_uid and $local_user_gid at routing time.
6759 48. Exim crashed when trying to verify a sender address that was being
6762 49. Exim was recognizing only a space character after ".include". It now also
6763 recognizes a tab character.
6765 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6766 extracting the relevant information from the specification. The man page no
6767 longer contains scrambled data for the -d option, and I've added a section
6768 at the front about calling Exim under different names.
6770 51. Added "extra_headers" argument to the "mail" command in filter files.
6772 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6775 53. Installed eximstats 1.29.
6777 54. Added transport_filter_timeout as a generic transport option.
6779 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6780 Cc: header lines. This was required by RFC 822, but it not required by RFC
6783 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6784 incoming messages that did not have them. Now it does so only if the
6785 message originates locally, that is, if there is no associated remote host
6786 address. When Resent- header lines are present, this applies to the Resent-
6787 lines rather than the non-Resent- lines.
6789 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6790 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6792 58. Messages for configuration errors now include the name of the main
6793 configuration file - useful now that there may be more than one file in a
6794 list (.included file names were always shown).
6796 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6797 for those rare installations that do not start the daemon as root or run it
6798 setuid root. I've cut out the call to initgroups() if the daemon is not
6801 60. The Exim user and group can now be bound into the binary as text strings
6802 that are looked up at the start of Exim's processing.
6804 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6806 62. Added $mailstore_basename variable.
6808 63. Installed patch to sieve.c from Michael Haardt.
6810 64. When Exim failed to open the panic log after failing to open the main log,
6811 the original message it was trying to log was written to stderr and debug
6812 output, but if they were not available (the usual case in production), it
6813 was lost. Now it is written to syslog before the two lines that record the
6814 failures to open the logs.
6816 65. Users' Exim filters run in subprocesses under the user's uid. It is
6817 possible for a "deliver" command or an alias in a "personal" command to
6818 provoke an address rewrite. If logging of address rewriting is configured,
6819 this fails because the process is not running as root or exim. There may be
6820 a better way of dealing with this, but for the moment (because 4.30 needs
6821 to be released), I have disabled address rewrite logging when running a
6822 filter in a non-root, non-exim process.
6828 1. The buildconfig auxiliary program wasn't quoting the value set for
6829 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6830 not defined. This bug was masked in 4.22 by the effect that was fixed in
6833 2. Some messages that were rejected after a message id was allocated were
6834 shown as "incomplete" by exigrep. It no longer does this for messages that
6835 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6837 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6838 have allow_domain_literals set, the ID did not get logged in the <= line.
6839 Domain literals are now always recognized in Message-ID: header lines.
6841 4. The first argument for a ${extract expansion item is the key name or field
6842 number. Leading and trailing spaces in this item were not being ignored,
6843 causing some misleading effects.
6845 5. When deliver_drop_privilege was set, single queue runner processes started
6846 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6847 same command in the process it spins off) were not dropping privilege.
6849 6. When the daemon running as "exim" started a queue runner, it always
6850 re-executed Exim in the spun-off process. This is a waste of effort when
6851 deliver_drop_privilege is set. The new process now just calls the
6852 queue-runner function directly.
6858 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6861 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6862 ignored. Though the use of -oP was forcing the writing of a pid file, it
6863 was always written to the default place.
6865 3. If the message "no IP address found for host xxxx" is generated during
6866 incoming verification, it is now followed by identification of the incoming
6867 connection (so you can more easily find what provoked it).
6869 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6871 5. Added some features to "harden" Exim a bit more against certain attacks:
6873 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6874 be put in Local/Makefile. This is like the never_users runtime option,
6875 but it cannot be overridden. The default setting is "root".
6877 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6878 prefix string with which any file named in a -C command line option
6881 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6882 is retained for -C and -D only if the caller of Exim is root. Without
6883 it, the exim user may also use -C and -D and retain privilege.
6885 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6886 command line option is disabled.
6888 6. Macro names set by the -D option must start with an upper case letter, just
6889 like macro names defined in the configuration file.
6891 7. Added "dereference=" facility to LDAP.
6893 8. Two instances of the typo "uknown" in the source files are fixed.
6895 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6896 the Configure-Makefile script screwed up while processing it.
6898 10. Incorporated PCRE 4.4.
6900 11. The SMTP synchronization check was not operating right at the start of an
6901 SMTP session. For example, it could not catch a HELO sent before the client
6902 waited for the greeting. There is now a check for outstanding input at the
6903 point when the greeting is written. Because of the duplex, asynchronous
6904 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6905 way, but not yet received, when the check is performed.
6907 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6908 on TCP/IP sockets, because this apparently causes some broken clients to
6911 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6912 unchanged) from the Cygwin maintainer.
6914 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6915 was supported instead of testing for mbx. Effectively a typo.
6917 15. The spa authenticator server code was not checking that the input it
6918 received was valid base64.
6920 16. The debug output line for the "set" modifier in ACLs was not showing the
6921 name of the variable that was being set.
6923 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6925 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6926 lookup on its own. Something else was needed to trigger the lookup. For
6927 example, a match in host_lookup or the need for a host name in a host list.
6928 Now, if $sender_host_name is referenced and the host name has not yet been
6929 looked up, a lookup is performed. If the lookup fails, the variable remains
6930 empty, and $host_lookup_failed is set to "1".
6932 19. Added "eqi" as a case-independent comparison operator.
6934 20. The saslauthd authentication condition could segfault if neither service
6935 nor realm was specified.
6937 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6938 error message that was logged was misleading, and incoming SMTP
6939 connections were dropped. The message is now more accurate, and temporary
6940 errors are given to SMTP connections.
6942 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6943 (except RSET) with a 421 error, until QUIT is received. However, it was
6944 failing to send a response to QUIT.
6946 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6947 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6948 ACL is now run after the helo_try_verify_hosts code.
6950 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6951 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6952 case-independent, so other case variants are also recognized). Apparently
6953 some systems use these upper case variants.
6955 25. If more than two messages were waiting for the same host, and a transport
6956 filter was specified for the transport, Exim sent two messages over the
6957 same TCP/IP connection, and then failed with "socket operation on non-
6958 socket" when it tried to send the third.
6960 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6962 27. The extern definition of crypt16() in expand.c was not being excluded when
6963 the OS had its own crypt16() function.
6965 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6966 as a preferred synonym for return_size_limit, both as an option and as an
6969 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6971 30. Changed the default configuration ACL to relax the local part checking rule
6972 for addresses that are not in any local domains. For these addresses,
6973 slashes and pipe symbols are allowed within local parts, but the sequence
6974 /../ is explicitly forbidden.
6976 31. SPA server authentication was not clearing the challenge buffer before
6979 32. log_message in a "warn" ACL statement was writing to the reject log as
6980 well as to the main log, which contradicts the documentation and doesn't
6981 seem right (because no rejection is happening). So I have stopped it.
6983 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6984 However, I am unable to do any testing of this.
6986 34. Fixed an infelicity in the appendfile transport. When checking directories
6987 for a mailbox, to see if any needed to be created, it was accidentally
6988 using path names with one or more superfluous leading slashes; tracing
6989 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6991 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6992 contents are added to the log line that is written for every discarded
6993 recipient. (Previously a log_message setting was ignored.)
6995 36. The ${quote: operator now quotes the string if it is empty.
6997 37. The install script runs exim in order to find its version number. If for
6998 some reason other than non-existence or emptiness, which it checks, it
6999 could not run './exim', it was installing it with an empty version number,
7000 i.e. as "exim-". This error state is now caught, and the installation is
7003 38. An argument was missing from the function that creates an error message
7004 when Exim fails to connect to the socket for saslauthd authentication.
7005 This could cause Exim to crash, or give a corrupted message.
7007 39. Added isip, isip4, and isip6 to ${if conditions.
7009 40. The ACL variables $acl_xx are now saved with the message, and can be
7010 accessed later in routers, transports, and filters.
7012 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7013 strings in the file are not string-expanded.
7015 42. If a MAIL command specified a SIZE value that was too large to fit into an
7016 int variable, the check against message_size_limit failed. Such values are
7017 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7018 day this will have to be increased, but I don't think I want to be around
7019 when emails are that large.
7026 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7027 iconv() is not standard in FreeBSD.
7029 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7030 IPv6 enabled. The observed symptom was a segmentation fault on return from
7031 the function os_common_find_running_interfaces() in src/os.c.
7033 3. In the check_special_case() function in daemon.c I had used "errno" as an
7034 argument name, which causes warnings on some systems. This was basically a
7035 typo, since it was named "eno" in the comments!
7037 4. The code that waits for the clock to tick (at a resolution of some fraction
7038 of a second) so as to ensure message-id uniqueness was always waiting for
7039 at least one whole tick, when it could have waited for less. [This is
7040 almost certainly not relevant at current processor speeds, where it is
7041 unlikely to ever wait at all. But we try to future-proof.]
7043 5. The function that sleeps for a time interval that includes fractions of a
7044 second contained a race. It did not block SIGALRM between setting the
7045 timer, and suspending (a couple of lines later). If the interval was short
7046 and the sigsuspend() was delayed until after it had expired, the suspension
7047 never ended. On busy systems this could lead to processes getting stuck for
7050 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7051 process, before it forks any delivery processes. The open lookup caching
7052 mechanism meant that the open file or database connection was passed into
7053 the delivery process. The problem was that delivery processes always tidy
7054 up cached lookup data. This could cause a problem for the next delivery
7055 process started by the queue runner, because the external queue runner
7056 process does not know about the closure. So the next delivery process
7057 still has data in the lookup cache. In the case of a file lookup, there was
7058 no problem because closing a file descriptor in a subprocess doesn't affect
7059 the parent. However, if the lookup was caching a connection to a database,
7060 the connection was closed, and the second delivery process was likely to
7061 see errors such as "PGSQL: query failed: server closed the connection
7062 unexpectedly". The problem has been fixed by closing all cached lookups
7063 in a queue runner before running a delivery process.
7065 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7066 seem to have the "const" qualifier which it has on other OS. I've
7069 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7070 *of the same type* (client or server) with the same public name that an
7071 error should be diagnosed.
7073 9. When Exim looked up a host name for an IP address, but failed to find the
7074 original IP address when looking up the host name (a safety check), it
7075 output the message "<ip address> does not match any IP for NULL", which was
7076 confusing, to say the least. The bug was that the host name should have
7077 appeared instead of "NULL".
7079 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7080 user that is built into the binary, and the -C or -D options is used, root
7081 privilege is dropped before the configuration file is read. In addition,
7082 logging is switched to stderr instead of the normal log files. If the
7083 configuration then re-defines the Exim user, the unprivileged environment
7084 is probably not what is expected, so Exim logs a panic warning message (but
7087 However, if deliver_drop_privilege is set, the unprivileged state may well
7088 be exactly what is intended, so the warning has been cut out in that case,
7089 and Exim is allowed to try to write to its normal log files.
7095 1. smtp_return_error_details was not giving details for temporary sender
7096 or receiver verification errors.
7098 2. Diagnose a configuration error if two authenticators have the same public
7101 3. Exim used not to create the message log file for a message until the first
7102 delivery attempt. This could be confusing when incoming messages were held
7103 for policy or load reasons. The message log file is now created at the time
7104 the message is received, and an initial "Received" line is written to it.
7106 4. The automatically generated man page for command line options had a minor
7107 bug that caused no ill effects; however, a more serious problem was that
7108 the procedure for building the man page automatically didn't always
7109 operate. Consequently, release 4.20 contains an out-of-date version. This
7110 shouldn't happen again.
7112 5. When building Exim with embedded Perl support, the script that builds the
7113 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7114 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7116 6. The freeze_tell option was not being used for messages that were frozen on
7117 arrival, either by an ACL or by local_scan().
7119 7. Added the smtp_incomplete_transaction log selector.
7121 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7122 was accepting AUTH without a new EHLO.
7124 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7125 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7128 10. Logging of TCP/IP connections (when configured) now happens in the main
7129 daemon process instead of the child process, so that the TCP/IP connection
7130 count is more accurate (but it can never be perfect).
7132 11. The use of "drop" in a nested ACL was not being handled correctly in the
7133 outer ACL. Now, if condition failure induced by the nested "drop" causes
7134 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7135 or "require"), the connection is dropped.
7137 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7138 that yield "discard" can now be used with an "accept" or a "discard" verb,
7139 but an error is generated for any others (because I can't see a useful way
7140 to define what should happen).
7142 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7143 are now processed in the same way as lines in the Exim configuration file.
7144 In particular, continuation lines are supported.
7146 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7148 15. Added -ti meaning -t -i.
7150 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7151 domains, and warn by logging if others are found.
7153 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7154 returned from GIFCONF when Exim is trying to find the list of interfaces on
7155 a host. The code in os.c has been modified to copy each ifreq to an aligned
7156 structure in all cases.
7158 Also, in some cases, the returned ifreq's were being copied to a 'struct
7159 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7160 means the last couple of bytes of an IPv6 address could be chopped if the
7161 ifreq contained only a normal sockaddr (14 bytes storage).
7163 18. Named domain lists were not supported in the hosts_treat_as_local option.
7164 An entry such as +xxxx was not recognized, and was treated as a literal
7167 19. Ensure that header lines added by a DATA ACL are included in the reject log
7168 if the ACL subsequently rejects the message.
7170 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7171 MD5 (which is deprecated).
7173 21. When testing a filter file using -bf, Exim was writing a message when it
7174 took the sender from a "From " line in the message, but it was not doing so
7175 when it took $return_path from a Return-Path: header line. It now does.
7177 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7178 with a valid header line field name (a series of printing characters
7179 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7181 23. Changed "disc" in the source to "disk" to conform to the documentation and
7182 the book and for uniformity.
7184 24. Ignore Sendmail's -Ooption=value command line item.
7186 25. When execve() failed while trying to run a command in a pipe transport,
7187 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7188 could be confused with a return value of 69 from the command itself. This
7189 has been changed to 127, the value the shell returns if it is asked to run
7190 a non-existent command. The wording for the related log line suggests a
7191 non-existent command as the problem.
7193 26. If received_header_text expands to an empty string, do not add a Received:
7194 header line to the message. (Well, it adds a token one on the spool, but
7195 marks it "old" so that it doesn't get used or transmitted.)
7197 27. Installed eximstats 1.28 (addition of -nt option).
7199 28. There was no check for failure on the call to getsockname() in the daemon
7200 code. This can fail if there is a shortage of resources on the system, with
7201 ENOMEM, for example. A temporary error is now given on failure.
7203 29. Contrary to the C standard, it seems that in some environments, the
7204 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7205 program. Exim now does this explicitly; it affects the formatting of
7206 timestamps using strftime().
7208 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7209 complaints. I've now initialized all the variables, to avoid this.
7211 32. Header lines added by a system filter were not being "seen" during
7212 transport-time rewrites.
7214 33. The info_callback() function passed to OpenSSL is set up with type void
7215 (*)(SSL *, int, int), as described somewhere. However, when calling the
7216 function (actually a macro) that sets it up, the type void(*)() is
7217 expected. I've put in a cast to prevent warnings from picky compilers.
7219 34. If a DNS black list lookup found a CNAME record, but there were no A
7220 records associated with the domain it pointed at, Exim crashed.
7222 35. If a DNS black list lookup returned more than one A record, Exim ignored
7223 all but the first. It now scans all returned addresses if a particular IP
7224 value is being sought. In this situation, the contents of the
7225 $dnslist_value variable are a list of all the addresses, separated by a
7228 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7229 to accept a host name and all its aliases if the forward lookup for any of
7230 them yielded the IP address of the incoming connection. Now it accepts only
7231 those names whose forward lookup yields the correct IP address. Any other
7232 names are discarded. This closes a loophole whereby a rogue DNS
7233 administrator could create reverse DNS records to break through a
7234 wildcarded host restriction in an ACL.
7236 37. If a user filter or a system filter that ran in a subprocess used any of
7237 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7238 the wrong values were passed to the pipe command ($thisaddress had the
7239 value of $0, $0 had the value of $1, etc). This bug was introduced by
7240 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7242 38. Improved the line breaking for long SMTP error messages from ACLs.
7243 Previously, if there was no break point between 40 and 75 characters, Exim
7244 left the rest of the message alone. Two changes have been made: (a) I've
7245 reduced the minimum length to 35 characters; (b) if it can't find a break
7246 point between 35 and 75 characters, it looks ahead and uses the first one
7247 that it finds. This may give the occasional overlong line, but at least the
7248 remaining text gets split now.
7250 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7251 file descriptors might be low, and that setting 1000 would always raise it.
7252 It turns out that in some environments, the limit is already over 1000 and
7253 that lowering it causes trouble. So now Exim takes care not to decrease it.
7255 40. When delivering a message, the value of $return_path is set to $sender_
7256 address at the start of routing (routers may change the value). By an
7257 oversight, this default was not being set up when an address was tested by
7258 -bt or -bv, which affected the outcome if any router or filter referred to
7261 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7262 log only when "message" or "log_message" are set. However, if one of the
7263 conditions was an address verification, or a call to a nested ACL, the
7264 messages generated by the underlying test were being passed through. This
7265 no longer happens. The underlying message is available in $acl_verify_
7266 message for both "message" and "log_message" expansions, so it can be
7267 passed through if needed.
7269 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7270 new expansion $bh_ to give the encoded byte string without charset
7271 translation. Translation happens only if iconv() is available; HAVE_ICONV
7272 indicates this at build time. HEADERS_CHARSET gives the charset to
7273 translate to; headers_charset can change it in the configuration, and
7274 "headers charset" can change it in an individual filter file.
7276 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7277 that creates RFC 2047 encoded "words" labels them as that charset instead
7278 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7279 expansion operator; (ii) when Exim creates a From: line for a local
7280 message; (iii) when a header line is rewritten to include a "phrase" part.
7282 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7283 buggy, causing it to skip the first lines of messages whose message ID
7284 ended in 'D'. This would not have bitten before Exim release 4.14, because
7285 message IDs were unlikely to end in 'D' before then. The effect was to have
7286 incorrect size information for certain domains.
7288 45. #include "config.h" was missing at the start of the crypt16.c module. This
7289 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7292 46. If there was a timeout during a "random" callout check, Exim treated it as
7293 a failure of the random address, and carried on sending RSET and the real
7294 address. If the delay was just some slowness somewhere, the response to the
7295 original RCPT would be taken as a response to RSET and so on, causing
7296 mayhem of various kinds.
7298 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7299 when I implemented it. It didn't allow for the fact that some option values
7300 may legitimately be negative (e.g. size_addition), and it didn't even do
7301 the right test for positive values.
7303 48. Domain names in DNS records are case-independent. Exim always looks them up
7304 in lower case. Some resolvers return domain names in exactly the case they
7305 appear in the zone file, that is, they may contain uppercase letters. Not
7306 all resolvers do this - some return always lower case. Exim was treating a
7307 change of case by a resolver as a change of domain, similar to a widening
7308 of a domain abbreviation. This triggered its re-routing code and so it was
7309 trying to route what was effectively the same domain again. This normally
7310 caused routing to fail (because the router wouldn't handle the domain
7311 twice). Now Exim checks for this case specially, and just changes the
7312 casing of the domain that it ultimately uses when it transmits the message
7315 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7318 50. If a filter generated a file delivery with a non-absolute name (possible if
7319 no home directory exists for the router), the forbid_file option was not
7322 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7323 the existing equality matching.
7325 52. Exim was using ints instead of ino_t variables in some places where it was
7326 dealing with inode numbers.
7328 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7329 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7330 called TMPDIR, and if it finds it is different, it changes its value.
7332 54. The smtp_printf() function is now made available to local_scan() so
7333 additional output lines can be written before returning. There is also an
7334 smtp_fflush() function to enable the detection of a dropped connection.
7335 The variables smtp_input and smtp_batched_input are exported to
7338 55. Changed the default runtime configuration: the message "Unknown user"
7339 has been removed from the ACL, and instead placed on the localuser router,
7340 using the cannot_route_message feature. This means that any verification
7341 failures that generate their own messages won't get overridden. Similarly,
7342 the "Unrouteable address" message that was in the ACL for unverifiable
7343 relay addresses has also been removed.
7345 56. Added hosts_avoid_esmtp to the smtp transport.
7347 57. The exicyclog script was not checking for the esoteric option
7348 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7349 will work only if exicyclog is run under the appropriate euid.
7351 58. Following a discussion on the list, the rules by which Exim recognises line
7352 endings on incoming messages have been changed. The -dropcr and drop_cr
7353 options are now no-ops, retained only for backwards compatibility. The
7354 following line terminators are recognized: LF CRLF CR. However, special
7355 processing applies to CR:
7357 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7358 nor a local message in the state where . is a terminator.
7360 (ii) If a bare CR is encountered in a header line, an extra space is added
7361 after the line terminator so as not to end the header. The reasoning
7362 behind this is that bare CRs in header lines are most likely either
7363 to be mistakes, or people trying to play silly games.
7365 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7366 was being incorrectly given as 18 bytes larger than it should have been.
7367 This is a VOB (very old bug).
7369 60. This may never have affected anything current, but just in case it has:
7370 When the local host is found other than at the start of a list of hosts,
7371 the local host, those with the same MX, and any that follow, are discarded.
7372 When the list in question was part of a longer list of hosts, the following
7373 hosts (not currently being processed) were also being discarded. This no
7374 longer happens. I'm not sure if this situation could ever has previously
7377 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7380 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7381 header. This is something that is recommended in a new Internet Draft, and
7382 is something that is documented as being done by Sendmail. There are two
7383 possible values. For messages generated by the autoreply transport, Exim
7386 Auto-Submitted: auto-replied
7388 whereas for all other generated messages (e.g. bounces) it adds
7390 Auto-Submitted: auto-generated
7392 63. The "personal" condition in filters now includes a test for the
7393 Auto-Submitted: header. If it contains the string "auto-" the message it
7394 not considered personal.
7396 64. Added rcpt_include_affixes as a generic transport option.
7398 65. Added queue_only_override (default true).
7400 66. Added the syslog_duplication option.
7402 67. If what should have been the first header line of a message consisted of
7403 a space followed by a colon, Exim was mis-interpreting it as a header line.
7404 It isn't of course - it is syntactically invalid and should therefore be
7405 treated as the start of the message body. The misbehaviour could have
7406 caused a number of strange effects, including loss of data in subsequent
7407 header lines, and spool format errors.
7409 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7410 client host had authenticated. This control can now be exercised by an ACL
7411 for more flexibility.
7413 69. By default, callouts do not happen when testing with -bh. There is now a
7414 variant, -bhc, which does actually run the callout code, including
7415 consulting and updating the callout cache.
7417 70. Added support for saslauthd authentication, courtesy of Alexander
7420 71. If statvfs() failed on the spool or log directories while checking their
7421 size for availability, Exim confusingly gave the error "space shortage".
7422 Furthermore, in debugging mode it crashed with a floating point exception.
7423 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7424 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7425 serious problem, Exim now writes to the main and panic logs when this
7426 happens, with details of the failure. It then refuses to accept the
7427 incoming message, giving the message "spool directory problem" or "log
7428 directory problem" with a 421 code for SMTP messages.
7430 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7431 0, 1, and 2 exist, because some OS complain for execs without them (see
7432 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7433 descriptors. However, the code omitted to check that the open succeeded,
7434 causing mysterious errors if for some reason the permissions on /dev/null
7435 got screwed. Now Exim writes a message to the main and panic logs, and
7436 bombs out if it can't open /dev/null.
7438 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7439 interact so that it is all more flexible. It is supposed to remain
7440 backwards compatible. Also added extra_local_interfaces.
7442 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7443 to bomb out with an assertion failure - to the client this appears as a
7444 connection drop. This problem occurs in the part of the code that was taken
7445 from the Samba project. Fortunately, the assertion is in a very simple
7446 function, so I have fixed this by reproducing the function inline in the
7447 one place where it is called, and arranging for authentication to fail
7448 instead of killing the process with assert().
7450 75. The SPA client code was not working when the server requested OEM rather
7451 than Unicode encoding.
7453 76. Added code to make require_files with a specific uid setting more usable in
7454 the case where statting the file as root fails - usually a non-root-mounted
7455 NFS file system. When this happens and the failure is EACCES, Exim now
7456 forks a subprocess and does the per-uid checking as the relevant uid.
7458 77. Added process_log_path.
7460 78. If log_file_path was not explicitly set, a setting of check_log_space or
7461 check_log_inodes was ignored.
7463 79. If a space check for the spool or log partitions fails, the incident is now
7464 logged. Of course, in the latter case the data may get lost...
7466 80. Added the %p formatting code to string_format() so that it can be used to
7467 print addresses in debug_print(). Adjusted all the address printing in the
7468 debugging in store.c to use %p rather than %d.
7470 81. There was a concern that a line of code in smtp_in.c could overflow a
7471 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7472 initially expressed, the concern was not well-founded, because trailing
7473 spaces are removed early. However, if the trailing spaces were followed by
7474 a NULL, they did not get removed, so the overflow was possible. Two fixes
7477 (a) I re-wrote the offending code in a cleaner fashion.
7478 (b) If an incoming SMTP command contains a NULL character, it is rejected
7481 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7482 runs initgroups(), so that if the Exim user is in any additional groups,
7483 they will be used during message reception.
7489 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob