1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Avoid costly startup code when not strictly needed. This reduces time
10 for some exim process initialisations. It does mean that the logging
11 of TLS configuration problems is only done for the daemon startup.
13 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
15 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
16 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
19 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
20 DKIM signing is being done. Previously a transport_filter would always
21 disable CHUNKING, falling back to traditional DATA.
23 JH/05 Regard command-line receipients as tainted.
25 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
27 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
28 PAM library frees one of the arguments given to it, despite the
29 documentation. Therefore a plain malloc must be used.
31 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
32 on-stack buffers were used, resulting in a taint trap when DSN information
33 copied from a received message was written into the buffer.
35 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
36 the ordering of its ARC headers. This caused a crash.
38 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
39 a new record was being constructed with information from the peer, a trap
42 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
43 installation would get error messages from DMARC verify, when it hit the
44 nonexistent file indicated by the default. Distros wanting DMARC enabled
45 should both provide the file and set the option.
46 Also enforce no DMARC verification for command-line sourced messages.
48 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
49 could, depending on the platform, hang at the STARTTLS response.
51 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
52 message on a connection. Previously if one message had ARC headers and
53 the following one did not, a crash could result when adding an
54 Authentication-Results: header.
56 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
57 Exim main code and Exim-related utities. The introduction of taint
58 tracking also did many adjustments to string handling. Since then, eximon
59 frequently terminated with an assert failure.
65 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
66 side, to decode a TLS 1.3 packet capture.
68 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
69 Previously the default library behaviour applied, sending two, each in
72 JH/03 Debug output for ACL now gives the config file name and line number for
75 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
77 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
79 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
80 buffer overrun for (non-chunking) other transports.
82 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
83 TLS1.3, means that a server rejecting a client certificate is not visible
84 to the client until the first read of encrypted data (typically the
85 response to EHLO). Add detection for that case and treat it as a failed
86 TLS connection attempt, so that the normal retry-in-clear can work (if
89 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
90 and/or domain. Found and fixed by Jason Betts.
92 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
93 configuration). If a CNAME target was not a wellformed name pattern, a
96 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
97 the OS reports them interleaved with other addresses.
99 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
100 used both for input and for a verify callout, both encrypted, SMTP
101 responses being sent by the server could be lost. This resulted in
102 dropped connections and sometimes bounces generated by a peer sending
105 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
106 string. Previously it was possible to cause undefined behaviour in a
107 library routine (usually a crash). Found by "zerons".
109 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
112 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
113 API was removed, so update to use the newer ones.
115 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
116 any timeout set, is taking a long time. Previously we would hang on to a
117 rotated logfile "forever" if the input was arriving with long gaps
118 (a previous attempt to fix addressed lack, for a long time, of initial
121 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
122 shared (NFS) environment. The length of the tempfile name is now
123 4 + 16 ("hdr.$message_exim_id") which might break on file
124 systems which restrict the file name length to lower values.
127 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
128 shared (NFS) environment.
130 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
131 did for all versions <4.90). Notably -M, -m, --invert, -I may be
134 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
135 on some platforms for bit 31.
137 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
138 to changes apparently associated with TLS1.3 handling some of the APIs
139 previously used were either nonfunctional or inappropriate. Strings
140 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
141 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
142 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
143 This affects log line X= elements, the $tls_{in,out}_cipher variables,
144 and the use of specific cipher names in the encrypted= ACL condition.
146 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
148 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
149 verification result was not updated unless hosts_require_ocsp applied.
151 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
152 queue_list_requires_admin set to false, non-admin users were denied the
155 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
156 directory-of-certs mode. Previously they were advertised despite the
159 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
160 A single TCP connection by a client will now hold a TLS connection open
161 for multiple message deliveries, by default. Previoud the default was to
164 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
165 default. If built with the facility, DANE will be used. The facility
166 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
168 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
169 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
170 must be defined and you must still, unless you define DISABLE_TLS, manage
171 the the include-dir and library-file requirements that go with that
172 choice. Non-TLS builds are still supported.
174 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
177 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
178 default. If the platform supports and has the facility enabled, it will
179 be requested on all coneections.
181 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
182 controlled by the build-time option SUPPORT_PIPE_CONNECT.
184 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
186 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
187 success-DSN messages. Previously the From: header was always the default
188 one for these; the option was ignored.
190 JH/28 Fix the timeout on smtp response to apply to the whole response.
191 Previously it was reset for every read, so a teergrubing peer sending
192 single bytes within the time limit could extend the connection for a
193 long time. Credit to Qualsys Security Advisory Team for the discovery.
195 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
196 delivery address, which leaked information of the results of local
197 forwarding. Change to the original envelope recipient address, per
200 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
201 requested. Previously not bounce was generated and a log entry of
202 error ignored was made.
204 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
206 JH/32 Introduce a general tainting mechanism for values read from the input
207 channel, and values derived from them. Refuse to expand any tainted
208 values, to catch one form of exploit.
210 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
211 was unused and the unexpanded text used for the test. Found and
212 fixed by Ruben Jenster.
214 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
215 an attempt to use a TLS library read routine dereffed a nul pointer,
218 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
219 them in our smtp response.
221 JH/36 Have the general router option retry_use_local_part default to true when
222 any of the restrictive preconditions are set (to anything). Previously it
223 was only for check_local user. The change removes one item of manual
224 configuration which is required for proper retries when a remote router
225 handles a subset of addresses for a domain.
227 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
228 link count into consideration.
230 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
231 caused the extension of big_buffer, the following lines were ignored.
233 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
234 accordance with RFC 2308. Previously there was no expiry, so a longlived
235 receive process (eg. due to ACL delays) versus a short SOA value could
238 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
240 JH/39 Promote DMARC support to mainline.
242 JH/40 Bug 2452: Add a References: header to DSNs.
244 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
245 parameters. The relevant library call is documented as "Deprecated: This
246 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
247 3.6.0, DH parameters are negotiated following RFC7919."
249 HS/06 Change the default of dnssec_request_domains to "*"
251 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
252 carried on and emitted a BDAT command, even when PIPELINING was not
255 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
256 buffer was used for the filename, resulting in a trap when tainted
257 arguments (eg. $domain) were used.
259 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
260 recommended to avoid a possible server-load attack. The feature can be
261 re-enabled via the openssl_options main cofiguration option.
263 JH/45 local_scan API: documented the current smtp_printf() call. This changed
264 for version 4.90 - adding a "more data" boolean to the arguments.
265 Bumped the ABI version number also, this having been missed previously;
266 release versions 4.90 to 4.92.3 inclusive were effectively broken in
267 respect of usage of smtp_printf() by either local_scan code or libraries
268 accessed via the ${dlfunc } expansion item. Both will need coding
269 adjustment for any calls to smtp_printf() to match the new function
270 signature; a FALSE value for the new argument is always safe.
272 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
273 the file-offset (which the Linux syscall does, and exim expects); this
274 resulted in an indefinite loop.
276 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
277 to do ARC verification. The Authentication-Results: header line added
278 by the configuration then had no ARC item.
284 JH/01 Remove code calling the customisable local_scan function, unless a new
285 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
287 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
288 non-signal-safe functions being used.
290 JH/03 Bug 2269: When presented with a received message having a stupidly large
291 number of DKIM-Signature headers, disable DKIM verification to avoid
292 a resource-consumption attack. The limit is set at twenty.
294 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
295 report of oldest_pass in ${authres } in consequence, and separate out
296 some descriptions of reasons for verification fail.
298 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
299 files in the spool were present and unlocked. A queue-runner could spot
300 them, resulting in a duplicate delivery. Fix that by doing the unlock
301 after the unlink. Investigation by Tim Stewart. Take the opportunity to
302 add more error-checking on spoolfile handling while that code is being
305 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
306 No known attacks, no CVE, this is defensive hardening.
308 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
309 a queue-runner could start a delivery while other operations were ongoing.
310 Cutthrough delivery was a common victim, resulting in duplicate delivery.
311 Found and investigated by Tim Stewart. Fix by using the open message data
312 file handle rather than opening another, and not locally closing it (which
313 releases a lock) for that case, while creating the temporary .eml format
314 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
316 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
317 $sender_verify_failure/$recipient_verify_failure to "random".
319 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
322 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
323 Previously this would segfault.
325 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
328 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
329 like zero, since the resolver should be doing this for us, But we need one
330 as a CNAME but no MX presence gets the CNAME returned; we need to check
331 that doesn't point to an MX to declare it "no MX returned" rather than
332 "error, loop". A new main option is added so the older capability of
333 following some limited number of chain links is maintained.
335 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
337 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
338 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
339 now-deprecated earlier definitions used only specified the range up to TLS
340 1.2 (in the older-version library docs).
342 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
344 JH/15 Rework TLS client-side context management. Stop using a global, and
345 explicitly pass a context around. This enables future use of TLS for
346 connections to service-daemons (eg. malware scanning) while a client smtp
347 connection is using TLS; with cutthrough connections this is quite likely.
349 JH/16 Fix ARC verification to do AS checks in reverse order.
351 JH/17 Support a "tls" option on the ${readsocket } expansion item.
353 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
354 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
355 Previously the "utf8" would be re-prepended for every additional message.
357 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
358 Previously thery were accepted, resulting in issues when attempting to
359 forward messages to a non-supporting MTA.
361 PP/02 Let -n work with printing macros too, not just options.
363 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
364 one parent address was copied, and bogus data was used at delivery-logging
365 time. Either a crash (after delivery) or bogus log data could result.
366 Discovery and analysis by Tim Stewart.
368 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
369 Previously if the string ended mid-character, we did not insert the
370 promised '?' replacement.
372 PP/04 Documentation: current string operators work on bytes, not codepoints.
374 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
375 should pack well giving a smaller memory footprint so better caching and
376 therefore performance. Group the declarations where this can't be done so
377 that the byte-sized flag variables are not interspersed among pointer
378 variables, giving a better chance of good packing by the compiler.
380 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
381 non-null, to avoid issues with sites running BATV. Previously reports were
382 sent with an empty envelope sender so looked like bounces.
384 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
385 The ignore_error flag wasn't being returned from the filter subprocess so
386 was not set for later routers. Investigation and fix by Matthias Kurz.
388 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
389 and a msg:complete for the whole, when a message is manually removed using
390 -Mrm. Developement by Matthias Kurz, hacked on by JH.
392 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
393 a "Gnu special" function, asprintf() in the DB utility binary builds; I
394 hope that is portable enough.
396 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
397 requiring a known-CA anchor certificate; make it now rely entirely on the
398 TLSA as an anchor. Checking the name on the leaf cert against the name
399 on the A-record for the host is still done for TA (but not for EE mode).
401 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
402 included in delivery lines for non-proxied connections, when compiled with
403 SUPPORT_SOCKS and running with proxy logging enabled.
405 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
406 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
407 move the existing event to fire before the normal logging of message
408 failure so that custom logging is bracketed by normal logging.
410 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
411 msg:fail:internal event. Developement by Matthias Kurz.
413 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
414 far too small for todays use of crypto signatures stored there. Go all
415 the way to the max DNS message size of 64kB, even though this might be
416 overmuch for IOT constrained device use.
418 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
419 copy a string over itself. The library routine is documented as not
420 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
422 JH/32 For main options check_spool_space and check_inode_space, where the
423 platform supports 64b integers, support more than the previous 2^31 kB
424 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
425 the previous G, M, k.
427 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
428 $authenticated_fail_id variable on authentication failure. Previously
431 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
432 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
433 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
434 GNUTLS_SEC_PARAM_MEDIUM.
436 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
437 side. Previously we would continue as if no SNI had been received.
439 JH/36 Harden the handling of string-lists. When a list consisted of a sole
440 "<" character, which should be a list-separator specification, we walked
441 off past the nul-terimation.
443 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
444 causes) even when the retry time is not yet met. Previously they were
445 not, meaning that when (say) an account was over-quota and temp-rejecting,
446 and multiple senders' messages were queued, only one sender would get
447 notified on each configured delay_warning cycle.
449 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
451 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
452 cipher-suites, an error can be left on the stack even for a succeeding
453 accept; this results in impossible error messages when a later operation
456 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
457 return error codes indicating retry. Under TLS1.3 this becomes required.
459 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
460 it only wrote the new authenticators, resulting in a lack of tracking of
461 peer changes of ESMTP extensions until the next cache flush.
463 JH/41 Fix the loop reading a message header line to check for integer overflow,
464 and more-often against header_maxsize. Previously a crafted message could
465 induce a crash of the recive process; now the message is cleanly rejected.
467 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
468 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
474 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
475 When redis_servers is set to a list of > 1 element, and the Redis servers
476 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
477 case of MOVED into a DEFER case instead, thus moving the query onto the
478 next server in the list. For a cluster of N elements, all N servers must
479 be defined in redis_servers.
481 GF/02 Catch and remove uninitialized value warning in exiqsumm
482 Check for existence of @ARGV before looking at $ARGV[0]
484 JH/01 Replace the store_release() internal interface with store_newblock(),
485 which internalises the check required to safely use the old one, plus
486 the allocate and data copy operations duplicated in both (!) of the
487 extant use locations.
489 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
490 modifier. This matches the restriction on the commandline.
492 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
493 Previously only the last row was returned.
495 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
496 we assumed that tags in the header were well-formed, and parsed the
497 element content after inspecting only the first char of the tag.
498 Assumptions at that stage could crash the receive process on malformed
501 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
502 While running the DKIM ACL we operate on the Permanent memory pool so that
503 variables created with "set" persist to the DATA ACL. Also (at any time)
504 DNS lookups that fail create cache records using the Permanent pool. But
505 expansions release any allocations made on the current pool - so a dnsdb
506 lookup expansion done in the DKIM ACL releases the memory used for the
507 DNS negative-cache, and bad things result. Solution is to switch to the
508 Main pool for expansions.
509 While we're in that code, add checks on the DNS cache during store_reset,
510 active in the testsuite.
511 Problem spotted, and debugging aided, by Wolfgang Breyha.
513 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
514 When none of the hosts presented to a transport match an already-open
515 connection, close it and proceed with the list. Previously we would
516 queue the message. Spotted by Lena with Yahoo, probably involving
519 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
520 Previously a spurious "250 OK id=" response was appended to the proper
523 JH/08 The "support for" informational output now, which built with Content
524 Scanning support, has a line for the malware scanner interfaces compiled
525 in. Interface can be individually included or not at build time.
527 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
528 by the template makefile "src/EDITME". The "STREAM" support for an older
529 ClamAV interface method is removed.
531 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
532 rows affected is given instead).
534 JH/11 The runtime Berkeley DB library version is now additionally output by
535 "exim -d -bV". Previously only the compile-time version was shown.
537 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
538 SMTP connection. Previously, when one had more receipients than the
539 first, an abortive onward connection was made. Move to full support for
540 multiple onward connections in sequence, handling cutthrough connection
541 for all multi-message initiating connections.
543 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
544 routers. Previously, a multi-recipient message would fail to match the
545 onward-connection opened for the first recipient, and cause its closure.
547 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
548 a timeout on read on a GnuTLS initiating connection, resulting in the
549 initiating connection being dropped. This mattered most when the callout
550 was marked defer_ok. Fix to keep the two timeout-detection methods
553 JH/15 Relax results from ACL control request to enable cutthrough, in
554 unsupported situations, from error to silently (except under debug)
555 ignoring. This covers use with PRDR, frozen messages, queue-only and
558 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
560 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
561 metadata, resulting in a crash in free().
563 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
564 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
565 Broken also in d185889f4, with init system revamp.
567 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
568 Previously we abruptly closed the connection after reading a malware-
569 found indication; now we go on to read the "scan ok" response line,
572 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
573 ACL. Previously, a crash would result.
575 JH/19 Speed up macro lookups during configuration file read, by skipping non-
576 macro text after a replacement (previously it was only once per line) and
577 by skipping builtin macros when searching for an uppercase lead character.
579 JH/20 DANE support moved from Experimental to mainline. The Makefile control
580 for the build is renamed.
582 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
583 was allocated for every new TLS startup, meaning one per message. Fix
584 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
586 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
587 reported the original. Fix to report (as far as possible) the ACL
588 result replacing the original.
590 JH/23 Fix memory leak during multi-message connections using STARTTLS under
591 OpenSSL. Certificate information is loaded for every new TLS startup,
592 and the resources needed to be freed.
594 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
596 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
599 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
600 DATA response info to the (existing) per-recipient response info for
601 the "C=" log element. It can have useful tracking info from the
602 destination system. Patch from Simon Arlott.
604 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
605 length value. Previously this would segfault.
607 HS/02 Support Avast multiline protoocol, this allows passing flags to
608 newer versions of the scanner.
610 JH/28 Ensure that variables possibly set during message acceptance are marked
611 dead before release of memory in the daemon loop. This stops complaints
612 about them when the debug_store option is enabled. Discovered specifically
613 for sender_rate_period, but applies to a whole set of variables.
614 Do the same for the queue-runner and queue-list loops, for variables set
615 from spool message files. Do the same for the SMTP per-message loop, for
616 certain variables indirectly set in ACL operations.
618 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
619 as a multi-recipient message from a mailinglist manager). The coding had
620 an arbitrary cutoff number of characters while checking for more input;
621 enforced by writing a NUL into the buffer. This corrupted long / fast
622 input. The problem was exposed more widely when more pipelineing of SMTP
623 responses was introduced, and one Exim system was feeding another.
624 The symptom is log complaints of SMTP syntax error (NUL chars) on the
625 receiving system, and refused recipients seen by the sending system
626 (propating to people being dropped from mailing lists).
627 Discovered and pinpointed by David Carter.
629 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
630 replaced by the ${authresults } expansion.
632 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
634 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
635 allows proper process termination in container environments.
637 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
638 Previously the "final dot" had a newline after it; ensure it is CR,LF.
640 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
641 and "err_perm", deprecated since 4.83 when the RFC-defined words
642 "temperror" and "permerror" were introduced.
644 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
645 transport-filters or DKIM-signing. The restriction was lost in the
646 consolidation of verify-callout and delivery SMTP handling.
647 Extend the restriction to also cover ARC-signing.
649 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
650 in defer=pass mode supply a 450 to the initiator. Previously the message
653 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
654 tls_require_ciphers is used as before.
656 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
657 "pass_unscanned". Only tmpfails from the scanner are written to
658 the paniclog, as they may require admin intervention (permission
659 denied, license issues). Other scanner errors (like decompression
660 bombs) do not cause a paniclog entry.
662 JH/36 Fix reinitialisation of DKIM logging variable between messages.
663 Previously it was possible to log spurious information in receive log
666 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
667 triggered odd behaviour from Outlook Express clients.
669 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
672 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
673 since the IETF WG has not yet settled on that versus the original
674 "bare" representation.
676 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
677 Previously the millisecond value corrupted the output.
678 Fix also for syslog_pid=no and log_selector +pid, for which the pid
679 corrupted the output.
685 JH/01 Rework error string handling in TLS interface so that the caller in
686 more cases is responsible for logging. This permits library-sourced
687 string to be attached to addresses during delivery, and collapses
688 pairs of long lines into single ones.
690 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
691 during configuration. Wildcards are allowed and expanded.
693 JH/02 Rework error string handling in DKIM to pass more info back to callers.
694 This permits better logging.
696 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
697 do not close it down and have the child transport start it up again on
698 the passed-on TCP connection. Instead, proxy the child (and any
699 subsequent ones) for TLS via a unix-domain socket channel. Logging is
700 affected: the continued delivery log lines do not have any DNSSEC, TLS
701 Certificate or OCSP information. TLS cipher information is still logged.
703 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
704 identical IP addresses on different listening ports. Will also affect
707 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
708 add noisy ifdef guards to special-case this sillyness.
709 Patch from Bernd Kuhls.
711 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
712 than 255 are no longer allowed.
714 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
715 Disable the session-cache too, which might reduce our load. Since we
716 currrectly use a new context for every connection, both as server and
717 client, there is no benefit for these.
718 GnuTLS appears to not support tickets server-side by default (we don't
719 call gnutls_session_ticket_enable_server()) but client side is enabled
720 by default on recent versions (3.1.3 +) unless the PFS priority string
723 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
724 <https://reproducible-builds.org/specs/source-date-epoch/>.
726 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
727 the check for any unsuccessful recipients did not notice the limit, and
728 erroneously found still-pending ones.
730 JH/08 Pipeline CHUNKING command and data together, on kernels that support
731 MSG_MORE. Only in-clear (not on TLS connections).
733 JH/09 Avoid using a temporary file during transport using dkim. Unless a
734 transport-filter is involved we can buffer the headers in memory for
735 creating the signature, and read the spool data file once for the
736 signature and again for transmission.
738 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
739 4.77 as the kernel support then wasn't solid, having issues in 64bit
740 mode. Now, it's been long enough. Add support for FreeBSD also.
742 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
743 case where the routing stage had gathered several addresses to send to
744 a host before calling the transport for the first, we previously failed
745 to close down TLS in the old transport process before passing the TCP
746 connection to the new process. The new one sent a STARTTLS command
747 which naturally failed, giving a failed delivery and bloating the retry
748 database. Investigation and fix prototype from Wolfgang Breyha.
750 JH/12 Fix check on SMTP command input synchronisation. Previously there were
751 false-negatives in the check that the sender had not preempted a response
752 or prompt from Exim (running as a server), due to that code's lack of
753 awareness of the SMTP input buffering.
755 PP/04 Add commandline_checks_require_admin option.
756 Exim drops privileges sanely, various checks such as -be aren't a
757 security problem, as long as you trust local users with access to their
758 own account. When invoked by services which pass untrusted data to
759 Exim, this might be an issue. Set this option in main configuration
760 AND make fixes to the calling application, such as using `--` to stop
763 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
764 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
765 responses to those, into a single TLS record each way (this usually means
766 a single packet). As a side issue, smtp_enforce_sync now works on TLS
769 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
770 affects you only if you're dancing at the edge of the param size limits.
771 If you are, and this message makes sense to you, then: raise the
772 configured limit or use OpenSSL 1.1. Nothing we can do for older
775 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
776 cmdline element to get the documented default one. Previously it was
779 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
780 are unsuable for non-message-specific reasons - eg. connection timeout,
781 banner-time rejection.
783 JH/16 Fix logging of delivery remote port, when specified by router, under
786 PP/06 Repair manualroute's ability to take options in any order, even if one
787 is the name of a transport.
790 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
792 JH/17 Change the list-building routines interface to use the expanding-string
793 triplet model, for better allocation and copying behaviour.
795 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
796 Previously it was constructed the first time a possibly-matching string
797 was met in the configuration file input during startup; now it is done
800 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
801 compatible one, to avoid the (poorly documented) possibility of a config
802 file in the working directory redirecting the DB files, possibly correpting
803 some existing file. CVE-2017-10140 assigned for BDB.
805 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
806 cache-hot. Previously, although the result was properly cached, the
807 initial verify call returned a defer.
809 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
810 the main verify for receipient in uncached-mode.
812 JH/22 Retire historical build files to an "unsupported" subdir. These are
813 defined as "ones for which we have no current evidence of testing".
815 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
816 if present. Previously it was ignored.
818 JH/24 Start using specified-initialisers in C structure init coding. This is
819 a C99 feature (it's 2017, so now considered safe).
821 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
822 if was a fixed-sized field and bitmask ops via macros; it is now more
825 PP/07 GitHub PR 56: Apply MariaDB build fix.
826 Patch provided by Jaroslav Škarvada.
828 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
829 during Coverity cleanups [4.87 JH/47]
830 Diagnosis and fix provided by Michael Fischer v. Mollard.
832 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
833 the right size to place the terminating semicolon on its own folded
834 line, the header hash was calculated to an incorrect value thanks to
835 the (relaxed) space the fold became.
837 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
838 and confused the parent.
840 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
841 which could crash as a result. This could lead to undeliverable messages.
843 JH/28 Logging: "next input sent too soon" now shows where input was truncated
846 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
847 matters on fast-turnover and PID-randomising systems, which were getting
848 out-of-order delivery.
850 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
851 a possibly-overlapping copy. The symptom was that "Remote host closed
852 connection in response to HELO" was logged instead of the actual 4xx
855 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
856 Previously only that bufferd was discarded, resulting in SYMTP command
859 JH/32 DKIM: when a message has multiple signatures matching an identity given
860 in dkim_verify_signers, run the dkim acl once for each. Previously only
861 one run was done. Bug 2189.
863 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
864 "panic the current process" to "deliberately defer". The panic log is
865 still written with the problem list name; the mail and reject logs now
866 get a temp-reject line for the message that was being handled, saying
867 something like "domains check lookup or other defer". The SMTP 451
868 message is still "Temporary local problem".
870 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
871 A crafted sequence of BDAT commands could result in in-use memory beeing
872 freed. CVE-2017-16943.
874 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
875 from SMTP input. Previously it was always done; now only done for DATA
876 and not BDAT commands. CVE-2017-16944.
878 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
879 to the message (such as an overlong header line). Previously this was
880 not done and we did not exit BDAT mode. Followon from the previous item
881 though a different problem.
887 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
888 than -2003 did; needs libidn2 in addition to libidn.
890 JH/02 The path option on a pipe transport is now expanded before use.
892 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
893 Patch provided by "Björn", documentation fix added too.
895 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
896 missing a wire-to-host endian conversion.
898 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
899 close after a BDAT command line could be taken as a following command,
900 giving a synch failure. Fix by only checking for synch immediately
901 before acknowledging the chunk.
903 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
904 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
905 macro. Patches provided by Josh Soref.
907 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
908 Previously we did not; the RFC seems ambiguous and VRFY is not listed
909 by IANA as a service extension. However, John Klensin suggests that we
912 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
913 the dkim code may be unix-mode line endings rather than smtp wire-format
914 CRLF, so prepend a CR to any bare LF.
916 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
917 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
919 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
920 the first were themselves being wrongly included in the feed into dkim
921 processing; with most chunk sizes in use this resulted in an incorrect
922 body hash calculated value.
924 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
925 DKIM signature block, for verification. Although advised against by
926 standards it is specifically not ruled illegal.
928 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
930 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
931 missing a body hash (the bh= tag).
933 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
934 It seems that HAProxy sends the Proxy Protocol information in clear and
935 only then does a TLS startup, so do the same.
937 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
938 TCP connections (such as for Spamd) unless the daemon successfully set
939 Fast Open mode on its listening sockets. This fixes breakage seen on
940 too-old kernels or those not configured for Fast Open, at the cost of
941 requiring both directions being enabled for TFO, and TFO never being used
942 by non-daemon-related Exim processes.
944 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
945 endings, at least on the first header line. Try to canonify any that get
946 past that check, despite the cost.
948 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
949 now limited to an arbitrary five deep, while parsing addresses with the
950 strip_excess_angle_brackets option enabled.
952 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
953 instead leave the unprompted TLS handshake in socket buffer for the
954 TLS library to consume.
956 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
958 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
960 JH/16 Drop variables when they go out of scope. Memory management drops a whole
961 region in one operation, for speed, and this leaves assigned pointers
962 dangling. Add checks run only under the testsuite which checks all
963 variables at a store-reset and panics on a dangling pointer; add code
964 explicitly nulling out all the variables discovered. Fixes one known
965 bug: a transport crash, where a dangling pointer for $sending_ip_address
966 originally assigned in a verify callout, is re-used.
968 PP/06 Drop '.' from @INC in various Perl scripts.
970 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
972 PP/08 Reduce a number of compilation warnings under clang; building with
973 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
974 should be warning-free.
976 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
978 HS/01 Fix portability problems introduced by PP/08 for platforms where
979 realloc(NULL) is not equivalent to malloc() [SunOS et al].
981 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
982 chunk. This allows us to accept broken chunked messages. We need a more
983 general solution here.
985 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
986 already-broken messages in the queue.
988 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
990 JH/19 Fix reference counting bug in routing-generated-address tracking.
996 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
997 supports it and a size is available (ie. the sending peer gave us one).
999 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1000 years of being deprecated). The replacements are the ACLs
1001 acl_smtp_mime and acl_not_smtp_mime.
1003 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1004 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1005 or even an in-clear connection were permitted. Now, if the host lookup
1006 was dnssec and dane was requested then the host is only used if the
1007 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1008 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1009 if one fails this test.
1010 This means that a poorly-configured remote DNS will make it incommunicado;
1011 but it protects against a DNS-interception attack on it.
1013 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1014 non-noisy when a race steals the message being considered.
1016 JH/05 If main configuration option tls_certificate is unset, generate a
1017 self-signed certificate for inbound TLS connections.
1019 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1020 in rewrites and routers.
1022 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1023 and logged a warning sing 4.83; now they are a configuration file error.
1025 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1026 (lacking @domain). Apply the same qualification processing as RCPT.
1028 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1030 JH/10 Support ${sha256:} applied to a string (as well as the previous
1033 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1034 a cutthrough deliver is pending, as we always want to make a connection.
1035 This also avoids re-routing the message when later placing the cutthrough
1036 connection after a verify cache hit.
1037 Do not update it with the verify result either.
1039 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1040 when routing results in more than one destination address.
1042 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1043 signing (which inhibits the cutthrough capability). Previously only
1044 the presence of an option was tested; now an expansion evaluating as
1045 empty is permissible (obviously it should depend only on data available
1046 when the cutthrough connection is made).
1048 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1049 the relevant preceding SMTP command did not note the pipelining mode.
1051 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1052 Previously they were not counted.
1054 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1055 as one having no matching records. Previously we deferred the message
1056 that needed the lookup.
1058 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1059 distinguished as "(=".
1061 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1062 for missing MX records. Previously it only worked for missing A records.
1064 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1066 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1067 after the data-go-ahead and data-ack. Patch from Jason Betts.
1069 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1070 even for a "none" policy. Patch from Tony Meyer.
1072 JH/22 Fix continued use of a connection for further deliveries. If a port was
1073 specified by a router, it must also match for the delivery to be
1076 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1077 When one of the recipients of a message was unsuitable for the connection
1078 (has no matching addresses), we lost track of needing to mark it
1079 deferred. As a result mail would be lost.
1081 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1083 JH/25 Decoding ACL controls is now done using a binary search; the source code
1084 takes up less space and should be simpler to maintain. Merge the ACL
1085 condition decode tables also, with similar effect.
1087 JH/26 Fix problem with one_time used on a redirect router which returned the
1088 parent address unchanged. A retry would see the parent address marked as
1089 delivered, so not attempt the (identical) child. As a result mail would
1092 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1093 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1094 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1097 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1098 Main config options check_{log,spool}_{inodes,space} are now
1099 100 inodes, 10MB unless set otherwise in the configuration.
1101 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1102 Previously it only applied to the main-section connection policy
1105 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1107 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1108 by me. Added RFC7919 DH primes as an alternative.
1110 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1111 are not in the system include path.
1113 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1114 GnuTLS, when a session startup failed (eg because the client disconnected)
1115 Exim did stdio operations after fclose. This was exposed by a recent
1116 change which nulled out the file handle after the fclose.
1118 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1119 signed directly by the cert-signing cert, rather than an intermediate
1120 OCSP-signing cert. This is the model used by LetsEncrypt.
1122 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1124 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1125 an incoming connection.
1127 HS/02 Bug 1802: Do not half-close the connection after sending a request
1130 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1131 fallback to "prime256v1".
1133 JH/34 SECURITY: Use proper copy of DATA command in error message.
1134 Could leak key material. Remotely exploitable. CVE-2016-9963.
1140 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1141 and 3.4.4 - once the server is enabled to respond to an OCSP request
1142 it does even when not requested, resulting in a stapling non-aware
1143 client dropping the TLS connection.
1145 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1146 support variable-length bit vectors. No functional change.
1148 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1149 The I= interface field on outgoing lines is now after the H= remote
1150 host field, same as incoming lines. There is a separate
1151 outgoing_interface log selector which allows you to disable the
1154 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1155 If not running log_selector +smtp_connection the mainlog would be held
1156 open indefinitely after a "too many connections" event, including to a
1157 deleted file after a log rotate. Leave the per net connection logging
1158 leaving it open for efficiency as that will be quickly detected by the
1159 check on the next write.
1161 HS/01 Bug 1671: Fix post transport crash.
1162 Processing the wait-<transport> messages could crash the delivery
1163 process if the message IDs didn't exist for some reason. When
1164 using 'split_spool_directory=yes' the construction of the spool
1165 file name failed already, exposing the same netto behaviour.
1167 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1168 mime_regex ACL conditions.
1170 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1171 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1172 to HELO, local diagnostic string.
1174 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1175 log line to debug. Even when configured with a tls authenticator many
1176 client connections are expected to not authenticate in this way, so
1177 an authenticate fail is not an error.
1179 HS/02 Add the Exim version string to the process info. This way exiwhat
1180 gives some more detail about the running daemon.
1182 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1183 matter for fast-change records such as DNSBLs.
1185 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1186 retry record, even if constant. There may be multiple transports with
1187 different interface settings and the retry behaviour needs to be kept
1190 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1192 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1194 JH/10 Bug 840: fix log_defer_output option of pipe transport
1196 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1197 hosts_randomize. This matters a lot when combined with mua_wrapper.
1199 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1200 ${quote_pgsql:<string>} operator.
1202 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1204 JH/14 Change header file name for freeradius-client. Relevant if compiling
1205 with Radius support; from the Gentoo tree and checked under Fedora.
1207 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1209 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1210 option transport_filter, meaning no filtering.
1212 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1214 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1215 defaults to "*" (all hosts). The variable is now available when not built
1216 with TLS, default unset, mainly to enable keeping the testsuite sane.
1217 If a server certificate is not supplied (via tls_certificate) an error is
1218 logged, and clients will find TLS connections fail on startup. Presumably
1219 they will retry in-clear.
1220 Packagers of Exim are strongly encouraged to create a server certificate
1221 at installation time.
1223 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1224 with the $config_file variable.
1226 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1227 in transport context, after the attempt, and per-recipient. The latter type
1228 is per host attempted. The event data is the error message, and the errno
1229 information encodes the lookup type (A vs. MX) used for the (first) host,
1230 and the trailing two digits of the smtp 4xx response.
1232 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1233 to write to mainlog (or rejectlog, paniclog) in the window between file
1234 creation and permissions/ownership being changed. Particularly affects
1235 installations where exicyclog is run as root, rather than exim user;
1236 result is that the running daemon panics and dies.
1238 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1240 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1241 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1242 "pri" and "weight". Note that the previous implicit priority given by the
1243 list order is no longer honoured.
1245 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1246 for DKIM processing.
1248 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1249 by defining SUPPORT_SOCKS.
1251 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1252 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1253 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1254 variables are renamed to proxy_{local,external}_{address,port}.
1256 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1257 for a build by defining SUPPORT_I18N
1259 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1260 of the query string, and make ${quote_redis:} do that quoting.
1262 JH/29 Move Events support from Experimental to mainline, enabled by default
1263 and removable for a build by defining DISABLE_EVENT.
1265 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1267 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1268 cached by the daemon.
1270 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1271 by defining LOOKUP_REDIS. The libhiredis library is required.
1273 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1274 keys are given for lookup.
1276 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1277 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1278 only supported when built with TLS support. The PolarSSL SHA routines
1279 are still used when the TLS library is too old for convenient support.
1281 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1282 openssl_options), for security. OpenSSL forces this from version 1.1.0
1283 server-side so match that on older versions.
1285 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1286 allocation for $value could be released as the expansion processing
1287 concluded, but leaving the global pointer active for it.
1289 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1290 and to use the domains and local_parts ACL conditions.
1292 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1293 incorrectly not doubled on cutthrough transmission, hence seen as a
1294 body-termination at the receiving system - resulting in truncated mails.
1295 Commonly the sender saw a TCP-level error, and retransmitted the message
1296 via the normal store-and-forward channel. This could result in duplicates
1297 received - but deduplicating mailstores were liable to retain only the
1298 initial truncated version.
1300 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1302 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1304 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1305 we're in there, support oversigning also; bug 1309.
1307 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1309 HS/04 Add support for keep_environment and add_environment options.
1311 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1312 either intentional arithmetic overflow during PRNG, or testing config-
1315 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1316 delivery resulted in actual delivery. Cancel cutthrough before DATA
1319 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1320 rejects a recipient: pass the reject to the originator.
1322 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1323 Many were false-positives and ignorable, but it's worth fixing the
1326 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1327 for the new environment-manipulation done at startup. Move the routines
1328 from being local to tls.c to being global via the os.c file.
1330 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1331 an extract embedded as result-arg for a map, the first arg for extract
1332 is unavailable so we cannot tell if this is a numbered or keyed
1333 extraction. Accept either.
1339 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1342 JH/02 The smtp transport option "multi_domain" is now expanded.
1344 JH/03 The smtp transport now requests PRDR by default, if the server offers
1347 JH/04 Certificate name checking on server certificates, when exim is a client,
1348 is now done by default. The transport option tls_verify_cert_hostnames
1349 can be used to disable this per-host. The build option
1350 EXPERIMENTAL_CERTNAMES is withdrawn.
1352 JH/05 The value of the tls_verify_certificates smtp transport and main options
1353 default to the word "system" to access the system default CA bundle.
1354 For GnuTLS, only version 3.0.20 or later.
1356 JH/06 Verification of the server certificate for a TLS connection is now tried
1357 (but not required) by default. The verification status is now logged by
1358 default, for both outbound TLS and client-certificate supplying inbound
1361 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1364 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1365 Status Notification (bounce) messages are now MIME format per RFC 3464.
1366 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1367 under the control of the dsn_advertise_hosts option, and routers may
1368 have a dsn_lasthop option.
1370 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1371 default, modifiable by a malware= option. The list separator for
1372 the options can now be changed in the usual way. Bug 68.
1374 JH/10 The smtp_receive_timeout main option is now expanded before use.
1376 JH/11 The incoming_interface log option now also enables logging of the
1377 local interface on delivery outgoing connections.
1379 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1380 if the interface and destination host and port all match.
1382 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1385 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1386 Patch from Andrew Lewis.
1388 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1389 now supports optional time-restrictions, weighting, and priority
1390 modifiers per server. Patch originally by <rommer@active.by>.
1392 JH/16 The spamd_address main option now supports a mixed list of local
1393 and remote servers. Remote servers can be IPv6 addresses, and
1394 specify a port-range.
1396 JH/17 Bug 68: The spamd_address main option now supports an optional
1397 timeout value per server.
1399 JH/18 Bug 1581: Router and transport options headers_add/remove can
1400 now have the list separator specified.
1402 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1405 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1408 JH/21 Support for the A6 type of dns record is withdrawn.
1410 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1411 rather than the verbs used.
1413 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1414 from 255 to 1024 chars.
1416 JH/24 Verification callouts now attempt to use TLS by default.
1418 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1419 are generic router options now. The defaults didn't change.
1421 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1422 Original patch from Alexander Shikoff, worked over by JH.
1424 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1425 files if ZCAT_COMMAND is not executable.
1427 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1429 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1431 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1432 Normally benign, it bites when the pair was led to by a CNAME;
1433 modern usage is to not canonicalize the domain to a CNAME target
1434 (and we were inconsistent anyway for A-only vs AAAA+A).
1436 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1438 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1439 when evaluating $sender_host_dnssec.
1441 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1442 $sender_helo_dnssec variable.
1444 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1446 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1448 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1450 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1451 documented as working, but never had. Support all but $spam_report.
1453 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1454 added for tls authenticator.
1456 HS/03 Add perl_taintmode main config option
1462 TL/01 When running the test suite, the README says that variables such as
1463 no_msglog_check are global and can be placed anywhere in a specific
1464 test's script, however it was observed that placement needed to be near
1465 the beginning for it to behave that way. Changed the runtest perl
1466 script to read through the entire script once to detect and set these
1467 variables, reset to the beginning of the script, and then run through
1468 the script parsing/test process like normal.
1470 TL/02 The BSD's have an arc4random API. One of the functions to induce
1471 adding randomness was arc4random_stir(), but it has been removed in
1472 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1473 function when detected.
1475 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1476 cause callback expansion.
1478 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1479 syntax errors in an expansion can be treated as a string instead of
1480 logging or causing an error, due to the internal use of bool_lax
1481 instead of bool when processing it.
1483 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1484 server certificates when making smtp deliveries.
1486 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1488 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1490 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1492 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1493 Merged patch from Sebastian Wiedenroth.
1495 JH/05 Fix results-pipe from transport process. Several recipients, combined
1496 with certificate use, exposed issues where response data items split
1497 over buffer boundaries were not parsed properly. This eventually
1498 resulted in duplicates being sent. This issue only became common enough
1499 to notice due to the introduction of connection certificate information,
1500 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1502 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1503 size buffer was used, resulting in syntax errors when an expansion
1506 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1507 version 3.3.6 or later.
1509 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1510 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1511 both become "event_action", the variables become $event_name, $event_data
1512 and $event_defer_errno. There is a new variable $verify_mode, usable in
1513 routers, transports and related events. The tls:cert event is now also
1514 raised for inbound connections, if the main configuration event_action
1517 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1518 early OCSP support, but no stapling (appears to be less than 1.0.0).
1520 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1521 server certificate names available under the smtp transport option
1522 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1525 JH/10 Time-related extraction expansions from certificates now use the main
1526 option "timezone" setting for output formatting, and are consistent
1527 between OpenSSL and GnuTLS compilations. Bug 1541.
1529 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1530 encoded parameter in the incoming message. Bug 1558.
1532 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1533 include certificate info, eximon was claiming there were spoolfile
1536 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1538 JH/14 Log delivery-related information more consistently, using the sequence
1539 "H=<name> [<ip>]" wherever possible.
1541 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1542 are problematic for Debian distribution, omit them from the release
1545 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1547 JH/16 Fix string representation of time values on 64bit time_t architectures.
1550 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1551 output list separator was used.
1556 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1557 checkers that were complaining about end of non-void function with no
1560 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1561 This was a regression introduced in 4.83 by another bugfix.
1563 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1565 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1566 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1572 TF/01 Correctly close the server side of TLS when forking for delivery.
1574 When a message was received over SMTP with TLS, Exim failed to clear up
1575 the incoming connection properly after forking off the child process to
1576 deliver the message. In some situations the subsequent outgoing
1577 delivery connection happened to have the same fd number as the incoming
1578 connection previously had. Exim would try to use TLS and fail, logging
1579 a "Bad file descriptor" error.
1581 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1582 utilities have not been installed.
1584 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1585 temporary space as the ACL may create new global variables.
1587 TL/01 LDAP support uses per connection or global context settings, depending
1588 upon the detected version of the libraries at build time.
1590 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1591 to extract and use the src ip:port in logging and expansions as if it
1592 were a direct connection from the outside internet. PPv2 support was
1593 updated based on HAProxy spec change in May 2014.
1595 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1597 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1598 Properly escape header and check for NULL return.
1600 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1603 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1605 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1606 characters in header names, implemented as a verify condition.
1607 Contributed by Michael Fischer v. Mollard.
1609 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1610 results permerror and temperror. Previous values are deprecated but
1611 still accepted. In a future release, err_perm and err_temp will be
1612 completely removed, which will be a backward incompatibility if the
1613 ACL tests for either of these two old results. Patch contributed by
1614 user bes-internal on the mailing list.
1616 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1618 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1619 selectors, in both main and reject logs.
1621 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1624 JH/07 Add malware type "sock" for talking to simple daemon.
1626 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1628 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1629 routers/transports under cutthrough routing.
1631 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1632 numbers. Touch up "bool" conditional to keep the same definition.
1634 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1636 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1638 JH/12 Expand items in router/transport headers_add or headers_remove lists
1639 individually rather than the list as a whole. Bug 1452.
1641 Required for reasonable handling of multiple headers_ options when
1642 they may be empty; requires that headers_remove items with embedded
1643 colons must have them doubled (or the list-separator changed).
1645 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1646 view the policy declared in the DMARC record. Currently, $dmarc_status
1647 is a combined value of both the record presence and the result of the
1650 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1652 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1653 dnslookup router and the smtp transport (applying to the forward
1656 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1657 of ldap servers used for a specific lookup. Patch provided by Heiko
1660 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1661 New variable $lookup_dnssec_authenticated for observability.
1663 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1664 Patch submitted by Lars Timman.
1666 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1668 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1669 Requires trusted mode and valid format message id, aborts otherwise.
1670 Patch contributed by Heiko Schlichting.
1672 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1673 certextract with support for various fields. Bug 1358.
1675 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1676 is requested by default, modifiable by smtp transport option
1679 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1680 operate on certificate variables to give certificate fingerprints
1681 Also new ${sha256:cert_variable}.
1683 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1685 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1686 Christian Aistleitner.
1688 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1690 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1691 file. Patch from Wolfgang Breyha.
1693 JH/25 Expand the coverage of the delivery $host and $host_address to
1694 client authenticators run in verify callout. Bug 1476.
1696 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1697 align with daemon_smtp_ports. Bug 72.
1699 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1700 support and error reporting did not work properly.
1702 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1703 and is readable. Patch from Andrew Colin Kissa.
1705 TL/14 Enhance documentation of ${run expansion and how it parses the
1706 commandline after expansion, particularly in the case when an
1707 unquoted variable expansion results in an empty value.
1709 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1711 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1712 by renaming away from it. Addresses GH issue 15, reported by
1715 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1717 TL/15 SECURITY: prevent double expansion in math comparison functions
1718 (can expand unsanitized data). Not remotely exploitable.
1725 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1727 PP/02 Make -n do something, by making it not do something.
1728 When combined with -bP, the name of an option is not output.
1730 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1733 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1734 $sender_host_name and config options to manage this, and basic check
1737 PP/05 DSCP support for outbound connections and control modifier for inbound.
1739 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1740 (Only plugin which currently uses this is kerberos4, which nobody should
1741 be using, but we should make it available and other future plugins might
1742 conceivably use it, even though it would break NAT; stuff *should* be
1743 using channel bindings instead).
1745 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1746 name; added for Sendmail compatibility; requires admin caller.
1747 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1748 just ignore it); requires trusted caller.
1749 Also parse but ignore: -Ac -Am -X<logfile>
1752 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1754 TL/02 Add +smtp_confirmation as a default logging option.
1756 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1757 Patch by Magnus Holmgren from 2007-02-20.
1759 TL/04 Bugzilla 1281 - Spec typo.
1760 Bugzilla 1283 - Spec typo.
1761 Bugzilla 1290 - Spec grammar fixes.
1763 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1765 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1767 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1768 mailing list by by Dmitry Isaikin.
1770 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1772 JH/02 Support "G" suffix to numbers in ${if comparisons.
1774 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1776 NM/01 Bugzilla 1197 - Spec typo
1777 Bugzilla 1196 - Spec examples corrections
1779 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1781 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1782 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1785 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1786 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1788 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1789 "acl {{name}{arg}...}", and optional args on acl condition
1792 JH/05 Permit multiple router/transport headers_add/remove lines.
1794 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1796 JH/07 Avoid using a waiting database for a single-message-only transport.
1797 Performance patch from Paul Fisher. Bugzilla 1262.
1799 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1802 JH/09 Add $headers_added variable, with content from use of ACL modifier
1803 add_header (but not yet added to the message). Bugzilla 199.
1805 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1806 Pulled from Bugzilla 817 by Wolfgang Breyha.
1808 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1810 (nb: this is the same fix as in Exim 4.80.1)
1812 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1815 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1818 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1820 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1821 advertise SMTP AUTH mechanism to us, instead of a generic
1822 protocol violation error. Also, make Exim more robust to bad
1823 data from the Dovecot auth socket.
1825 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1827 When a queue runner is handling a message, Exim first routes the
1828 recipient addresses, during which it prunes them based on the retry
1829 hints database. After that it attempts to deliver the message to
1830 any remaining recipients. It then updates the hints database using
1833 So if a recipient address works intermittently, it can get repeatedly
1834 deferred at routing time. The retry hints record remains fresh so the
1835 address never reaches the final cutoff time.
1837 This is a fairly common occurrence when a user is bumping up against
1838 their storage quota. Exim had some logic in its local delivery code
1839 to deal with this. However it did not apply to per-recipient defers
1840 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1842 This change adds a proper retry rule check during routing so that the
1843 final cutoff time is checked against the message's age. We only do
1844 this check if there is an address retry record and there is not a
1845 domain retry record; this implies that previous attempts to handle
1846 the address had the retry_use_local_parts option turned on. We use
1847 this as an approximation for the destination being like a local
1848 delivery, as in LMTP.
1850 I suspect this new check makes the old local delivery cutoff check
1851 redundant, but I have not verified this so I left the code in place.
1853 TF/02 Correct gecos expansion when From: is a prefix of the username.
1855 Test 0254 submits a message to Exim with the header
1859 When I ran the test suite under the user fanf2, Exim expanded
1860 the header to contain my full name, whereas it should have added
1861 a Resent-Sender: header. It erroneously treats any prefix of the
1862 username as equal to the username.
1864 This change corrects that bug.
1866 GF/01 DCC debug and logging tidyup
1867 Error conditions log to paniclog rather than rejectlog.
1868 Debug lines prefixed by "DCC: " to remove any ambiguity.
1870 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1872 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1873 Bug spotted by Jeremy Harris; was flawed since initial commit.
1874 Would have resulted in OCSP responses post-SNI triggering an Exim
1875 NULL dereference and crash.
1877 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1879 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1880 Bug detection, analysis and fix by Samuel Thibault.
1881 Bugzilla 1331, Debian bug #698092.
1883 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1885 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1886 Server implementation by Todd Lyons, client by JH.
1887 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1888 config variable "prdr_enable" controls whether the server
1889 advertises the facility. If the client requests PRDR a new
1890 acl_data_smtp_prdr ACL is called once for each recipient, after
1891 the body content is received and before the acl_smtp_data ACL.
1892 The client is controlled by both of: a hosts_try_prdr option
1893 on the smtp transport, and the server advertisement.
1894 Default client logging of deliveries and rejections involving
1895 PRDR are flagged with the string "PRDR".
1897 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1898 fclose(). Diagnosis by Todd Lyons.
1900 PP/17 Update configure.default to handle IPv6 localhost better.
1901 Patch by Alain Williams (plus minor tweaks).
1904 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1905 This is now consistent with GnuTLS, and is now documented: the
1906 previous undocumented portable approach to treating the option as
1907 unset was to force an expansion failure. That still works, and
1908 an empty string is now equivalent.
1910 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1911 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1912 not performing validation itself.
1914 PP/20 Added force_command boolean option to pipe transport.
1915 Patch from Nick Koston, of cPanel Inc.
1917 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1920 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1922 PP/21 Fix eximon continuous updating with timestamped log-files.
1923 Broken in a format-string cleanup in 4.80, missed when I repaired the
1924 other false fix of the same issue.
1925 Report and fix from Heiko Schlichting.
1928 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1929 Report from Prashanth Katuri.
1931 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1932 It's SecureTransport, so affects any MacOS clients which use the
1933 system-integrated TLS libraries, including email clients.
1935 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1936 using a MIME ACL for non-SMTP local injection.
1937 Report and assistance in diagnosis by Warren Baker.
1939 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1941 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1943 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1944 last id that failed so it may be referenced in subsequent ACL's.
1946 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1949 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1950 ldap library initialization, allowing self-signed CA's to be
1951 used. Also properly sets require_cert option later in code by
1952 using NULL (global ldap config) instead of ldap handle (per
1953 session). Bug diagnosis and testing by alxgomz.
1955 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1956 the src/util/ subdirectory.
1958 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1959 renamed to Transport Post Delivery Action by Jeremy Harris, as
1962 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1963 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1964 redis_servers = needs to be configured which will be used by the redis
1965 lookup. Patch from Warren Baker, of The Packet Hub.
1967 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1969 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1970 hostname or reverse DNS when processing a host list. Used suggestions
1971 from multiple comments on this bug.
1973 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1975 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1976 Missed a few lines, added it to make the runtest require no keyboard
1979 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1980 contains upper case chars. Make router use caseful_local_part.
1982 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1983 support when GnuTLS has been built with p11-kit.
1989 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1991 This, or similar/improved, will also be change PP/11 of 4.82.
1997 PP/01 Handle short writes when writing local log-files.
1998 In practice, only affects FreeBSD (8 onwards).
1999 Bugzilla 1053, with thanks to Dmitry Isaikin.
2001 NM/01 Bugzilla 949 - Documentation tweak
2003 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2006 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2008 PP/02 Implemented gsasl authenticator.
2010 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2012 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2013 `pkg-config foo` for cflags/libs.
2015 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2016 with rest of GSASL and with heimdal_gssapi.
2018 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2019 `pkg-config foo` for cflags/libs for the TLS implementation.
2021 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2022 properties get this fed in as external SSF. A number of robustness
2023 and debugging improvements to the cyrus_sasl authenticator.
2025 PP/08 cyrus_sasl server now expands the server_realm option.
2027 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2028 Patch by Jeremy Harris.
2030 PP/10 Added dbmjz lookup type.
2032 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2034 PP/12 MAIL args handles TAB as well as SP, for better interop with
2035 non-compliant senders.
2036 Analysis and variant patch by Todd Lyons.
2038 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2039 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2040 Patch from Dirk Mueller <dmueller@suse.com>
2042 PP/13 tls_peerdn now print-escaped for spool files.
2043 Observed some $tls_peerdn in wild which contained \n, which resulted
2044 in spool file corruption.
2046 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2047 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2048 or write after TLS renegotiation, which otherwise led to messages
2051 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2052 as a tracking header (ie: a signed header comes before the signature).
2053 Patch from Wolfgang Breyha.
2055 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2056 comma-sep list; embedded commas doubled.
2058 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2060 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2062 Report and patch from Dmitry Banschikov.
2064 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2065 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2066 failures. We appear to now support TLS1.1+ with Exim.
2068 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2069 lets Exim select keys and certificates based upon TLS SNI from client.
2070 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2071 before an outbound SMTP session. New log_selector, +tls_sni.
2073 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2074 NULL dereference. Report and patch from Alun Jones.
2076 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2077 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2078 Not seeing resolver debug output on NetBSD, but suspect this is a
2079 resolver implementation change.
2081 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2082 Left warnings. Added "eximon gdb" invocation mode.
2084 PP/21 Defaulting "accept_8bitmime" to true, not false.
2086 PP/22 Added -bw for inetd wait mode support.
2088 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2089 locate the relevant includes and libraries. Made this the default.
2091 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2092 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2094 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2095 This may cause build issues on older platforms.
2097 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2098 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2099 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2100 Added SNI support via GnuTLS too.
2101 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2103 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2105 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2106 Applied second patch from Janne, implementing suggestion to default
2107 multiple-strings-in-record handling to match SPF spec.
2109 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2111 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2112 read-only, out of scope).
2113 Patch from Wolfgang Breyha, report from Stuart Northfield.
2115 PP/29 Fix three issues highlighted by clang analyser static analysis.
2116 Only crash-plausible issue would require the Cambridge-specific
2117 iplookup router and a misconfiguration.
2118 Report from Marcin Mirosław.
2120 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2122 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2123 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2124 As part of this, removing so much warning spew let me fix some minor
2125 real issues in debug logging.
2127 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2128 assignment on my part. Fixed.
2130 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2131 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2132 Janne Snabb (who went above and beyond: thank you).
2134 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2135 string otherwise requires a connection and a bunch more work and it's
2136 relatively easy to get wrong. Should also expose TLS library linkage
2139 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2140 64-bit ${eval} (JH/03).
2142 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2143 GNU libc to support some of the 64-bit stuff, should not lead to
2144 conflicts. Defined before os.h is pulled in, so if a given platform
2145 needs to override this, it can.
2147 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2148 protection layer was required, which is not implemented.
2149 Bugzilla 1254, patch from Wolfgang Breyha.
2151 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2152 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2153 tls_dhparam take prime identifiers. Also unbreak combination of
2154 OpenSSL+DH_params+TLSSNI.
2156 PP/39 Disable SSLv2 by default in OpenSSL support.
2162 PP/01 Solaris build fix for Oracle's LDAP libraries.
2163 Bugzilla 1109, patch from Stephen Usher.
2165 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2167 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2170 TF/02 Fix a couple more cases where we did not log the error message
2171 when unlink() failed. See also change 4.74-TF/03.
2173 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2174 lock up or crash if it happened to be inside a call to libc when it
2175 got a SIGUSR1 from exiwhat.
2177 The SIGUSR1 handler appends the current process status to the process
2178 log which is later printed by exiwhat. It used to use the general
2179 purpose logging code to do this, but several functions it calls are
2180 not safe for signals.
2182 The new output code in the SIGUSR1 handler is specific to the process
2183 log, and simple enough that it's easy to inspect for signal safety.
2184 Removing some special cases also simplifies the general logging code.
2185 Removing the spurious timestamps from the process log simplifies
2188 TF/04 Improved ratelimit ACL condition.
2190 The /noupdate option has been deprecated in favour of /readonly which
2191 has clearer semantics. The /leaky, /strict, and /readonly update modes
2192 are mutually exclusive. The update mode is no longer included in the
2193 database key; it just determines when the database is updated. (This
2194 means that when you upgrade Exim will forget old rate measurements.)
2196 Exim now checks that the per_* options are used with an update mode that
2197 makes sense for the current ACL. For example, when Exim is processing a
2198 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2199 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2200 must specify per_mail/readonly. If you omit the update mode it defaults to
2201 /leaky where that makes sense (as before) or /readonly where required.
2203 The /noupdate option is now undocumented but still supported for
2204 backwards compatibility. It is equivalent to /readonly except that in
2205 ACLs where /readonly is required you may specify /leaky/noupdate or
2206 /strict/noupdate which are treated the same as /readonly.
2208 A useful new feature is the /count= option. This is a generalization
2209 of the per_byte option, so that you can measure the throughput of other
2210 aggregate values. For example, the per_byte option is now equivalent
2211 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2213 The per_rcpt option has been generalized using the /count= mechanism
2214 (though it's more complicated than the per_byte equivalence). When it is
2215 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2216 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2217 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2218 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2219 Note that using per_rcpt with a non-readonly update mode in more than
2220 one ACL will cause the recipients to be double-counted. (The per_mail
2221 and per_byte options don't have this problem.)
2223 The handling of very low rates has changed slightly. If the computed rate
2224 is less than the event's count (usually one) then this event is the first
2225 after a long gap. In this case the rate is set to the same as this event's
2226 count, so that the first message of a spam run is counted properly.
2228 The major new feature is a mechanism for counting the rate of unique
2229 events. The new per_addr option counts the number of different
2230 recipients that someone has sent messages to in the last time period. It
2231 behaves like per_rcpt if all the recipient addresses are different, but
2232 duplicate recipient addresses do not increase the measured rate. Like
2233 the /count= option this is a general mechanism, so the per_addr option
2234 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2235 example, measure the rate that a client uses different sender addresses
2236 with the options per_mail/unique=$sender_address. There are further
2237 details in the main documentation.
2239 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2241 TF/06 Removed a few PCRE remnants.
2243 TF/07 Automatically extract Exim's version number from tags in the git
2244 repository when doing development or release builds.
2246 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2247 Bugzilla 879. Patch from Paul Fisher.
2249 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2250 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2253 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2255 PP/05 Variable $av_failed, true if the AV scanner deferred.
2256 Bugzilla 1078. Patch from John Horne.
2258 PP/06 Stop make process more reliably on build failure.
2259 Bugzilla 1087. Patch from Heiko Schlittermann.
2261 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2262 Bugzilla 1089. Patch from Heiko Schlittermann.
2264 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2265 Bugzilla 1131. Patch from Holger Weiß.
2267 PP/09 Handle IPv6 addresses with SPF.
2268 Bugzilla 860. Patch from Wolfgang Breyha.
2270 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2272 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2275 PP/11 match_* no longer expand right-hand-side by default.
2276 New compile-time build option, EXPAND_LISTMATCH_RHS.
2277 New expansion conditions, "inlist", "inlisti".
2279 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2281 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2283 PP/14 fix log_write() format string regression from TF/03.
2284 Bugzilla 1152. Patch from Dmitry Isaikin.
2290 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2292 PP/02 Harmonised TLS library version reporting; only show if debugging.
2293 Layout now matches that introduced for other libraries in 4.74 PP/03.
2295 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2297 PP/04 New "dns_use_edns0" global option.
2299 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2302 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2303 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2305 TK/01 Updated PolarSSL code to 0.14.2.
2306 Bugzilla 1097. Patch from Andreas Metzler.
2308 PP/07 Catch divide-by-zero in ${eval:...}.
2309 Fixes bugzilla 1102.
2311 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2314 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2315 format-string attack -- SECURITY: remote arbitrary code execution.
2317 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2318 time unintentionally subject to list matching rules, letting the header
2319 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2320 arbitrary string expansion). This allowed for information disclosure.
2322 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2323 INT_MIN/-1 -- value coerced to INT_MAX.
2329 NM/01 Workaround for PCRE version dependency in version reporting
2332 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2333 This fixes portability to compilers other than gcc, notably
2334 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2336 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2337 makefiles for portability to HP-UX and POSIX correctness.
2339 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2340 Also via indented variable definition in the Makefile.
2341 (Debugging by Oliver Heesakkers).
2343 PP/02 Restore caching of spamd results with expanded spamd_address.
2344 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2346 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2347 Improves build reliability. Fix from: Frank Elsner
2349 NM/02 Fix wide character breakage in the rfc2047 coding
2350 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2352 NM/03 Allow underscore in dnslist lookups
2353 Fixes bug 1026. Patch from Graeme Fowler
2355 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2356 Code patches from Adam Ciarcinski of NetBSD.
2358 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2361 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2362 is logged, to avoid truncation. Patch from John Horne.
2364 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2365 Patch from Jakob Hirsch.
2367 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2368 SQL string expansion failure details.
2369 Patch from Andrey Oktyabrski.
2371 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2372 Patch from Simon Arlott.
2374 PP/09 New lookups functionality failed to compile on old gcc which rejects
2375 extern declarations in function scope.
2376 Patch from Oliver Fleischmann
2378 PP/10 Use sig_atomic_t for flags set from signal handlers.
2379 Check getgroups() return and improve debugging.
2380 Fixed developed for diagnosis in bug 927 (which turned out to be
2383 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2384 Patch from Mark Zealey.
2386 PP/12 Bugzilla 1056: Improved spamd server selection.
2387 Patch from Mark Zealey.
2389 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2390 Based on patch from Heiko Schlittermann.
2392 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2393 Patch from Uwe Doering, sign-off by Michael Haardt.
2395 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2396 variable declaration deep within a block. Bug and patch from
2399 PP/15 lookups-Makefile IRIX compatibility coercion.
2401 PP/16 Make DISABLE_DKIM build knob functional.
2403 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2404 Patch by Simon Arlott
2406 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2407 variable argument macros. Our copy now differs from upstream.
2413 TF/01 Failure to get a lock on a hints database can have serious
2414 consequences so log it to the panic log.
2416 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2417 controlled using the smtp_confirmation log selector.
2419 TF/03 Include the error message when we fail to unlink a spool file.
2421 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2422 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2423 for maintaining out-of-tree patches for some time.
2425 PP/01 Bugzilla 139: Documentation and portability issues.
2426 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2427 Handle per-OS dynamic-module compilation flags.
2429 PP/02 Let /dev/null have normal permissions.
2430 The 4.73 fixes were a little too stringent and complained about the
2431 permissions on /dev/null. Exempt it from some checks.
2432 Reported by Andreas M. Kirchwitz.
2434 PP/03 Report version information for many libraries, including
2435 Exim version information for dynamically loaded libraries. Created
2436 version.h, now support a version extension string for distributors
2437 who patch heavily. Dynamic module ABI change.
2439 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2440 privilege escalation vulnerability whereby the Exim run-time user
2441 can cause root to append content of the attacker's choosing to
2444 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2447 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2448 If dropping privileges for untrusted macros, we disabled normal logging
2449 on the basis that it would fail; for the Exim run-time user, this is not
2450 the case, and it resulted in successful deliveries going unlogged.
2451 Fixed. Reported by Andreas Metzler.
2457 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2458 only prepend for the Resent-* case. Fixes regression introduced in
2459 Exim 4.70 by NM/22 for Bugzilla 607.
2461 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2462 increasing numbers of administrators be bitten by this.
2464 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2466 PP/03 Bugzilla 994: added openssl_options main configuration option.
2468 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2470 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2472 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2474 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2475 without a peer certificate, leading to a segfault because of an
2476 assumption that peers always have certificates. Be a little more
2477 paranoid. Problem reported by Martin Tscholak.
2479 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2480 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2481 NB: ClamAV planning to remove STREAM in "middle of 2010".
2482 CL also introduces -bmalware, various -d+acl logging additions and
2483 more caution in buffer sizes.
2485 PP/09 Implemented reverse_ip expansion operator.
2487 PP/10 Bugzilla 937: provide a "debug" ACL control.
2489 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2491 PP/12 Bugzilla 973: Implement --version.
2493 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2495 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2497 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2499 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2500 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2501 ignore trailing whitespace.
2503 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2505 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2508 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2509 Notification from Dr Andrew Aitchison.
2511 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2512 ExtendedDetectionInfo response format.
2513 Notification from John Horne.
2515 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2518 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2519 XSL and documented dependency on system catalogs, with examples of how
2522 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2525 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2526 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2527 configuration file which is writeable by the Exim user or group.
2529 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2530 of configuration files to cover files specified with the -C option if
2531 they are going to be used with root privileges, not just the default
2534 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2535 option (effectively making it always true).
2537 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2538 files to be used while preserving root privileges.
2540 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2541 that rogue child processes cannot use them.
2543 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2544 run-time user, instead of root.
2546 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2547 Exim run-time user without dropping privileges.
2549 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2550 result string, instead of calling string_vformat() twice with the same
2553 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2554 for other users. Others should always drop root privileges if they use
2555 -C on the command line, even for a whitelisted configure file.
2557 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2559 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2565 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2566 $data_path, and $header_path variables; fixed documentation bugs and
2569 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2570 exipick to access non-standard spools, including the "frozen" queue
2573 NM/01 Bugzilla 965: Support mysql stored procedures.
2574 Patch from Alain Williams
2576 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2578 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2579 Patch from Andreas Metzler
2581 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2582 Patch from Kirill Miazine
2584 NM/05 Bugzilla 671: Added umask to procmail example.
2586 JJ/03 installed exipick 20100323.0, fixing doc bug
2588 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2589 directory. Notification and patch from Dan Rosenberg.
2591 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2593 TK/02 Improve log output when DKIM signing operation fails.
2595 MH/01 Treat the transport option dkim_domain as a colon separated
2596 list, not as a single string, and sign the message with each element,
2597 omitting multiple occurences of the same signer.
2599 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2600 Bugzilla 985, 986. Patch by Simon Arlott
2602 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2603 Patch by Simon Arlott
2605 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2606 MBX locking. Notification from Dan Rosenberg.
2612 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2614 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2616 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2618 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2620 NM/04 Bugzilla 915: Flag broken perl installation during build.
2626 TK/01 Added patch by Johannes Berg that expands the main option
2627 "spamd_address" if it starts with a dollar sign.
2629 TK/02 Write list of recipients to X-Envelope-Sender header when building
2630 the mbox-format spool file for content scanning (suggested by Jakob
2633 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2634 (http://www.dcc-servers.net/) support via dccifd. Activated by
2635 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2637 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2638 by Mark Daniel Reidel <mr@df.eu>.
2640 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2641 When building exim an external PCRE library is now needed -
2642 PCRE is a system library on the majority of modern systems.
2643 See entry on PCRE_LIBS in EDITME file.
2645 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2646 conversation. Added nologin parameter to request.
2647 Patch contributed by Kirill Miazine.
2649 TF/01 Do not log submission mode rewrites if they do not change the address.
2651 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2653 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2654 log files in place. Contributed by Roberto Lima.
2656 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2658 TF/03 Bugzilla 615: When checking the local_parts router precondition
2659 after a local_part_suffix or local_part_prefix option, Exim now
2660 does not use the address's named list lookup cache, since this
2661 contains cached lookups for the whole local part.
2663 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2664 Robert Millan. Documentation is in experimental-spec.txt.
2666 TF/04 Bugzilla 668: Fix parallel build (make -j).
2668 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2670 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2671 Patch provided by Jan Srzednicki.
2673 TF/05 Leading white space used to be stripped from $spam_report which
2674 wrecked the formatting. Now it is preserved.
2676 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2677 that they are available at delivery time.
2679 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2681 TF/08 TLS error reporting now respects the incoming_interface and
2682 incoming_port log selectors.
2684 TF/09 Produce a more useful error message if an SMTP transport's hosts
2685 setting expands to an empty string.
2687 NM/06 Bugzilla 744: EXPN did not work under TLS.
2688 Patch provided by Phil Pennock.
2690 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2691 Patch provided by Richard Godbee.
2693 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2694 acl_smtp_notquit, added index entry.
2696 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2697 Patch provided by Eugene Bujak.
2699 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2700 accept(). Patch provided by Maxim Dounin.
2702 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2703 Patch provided by Phil Pennock.
2705 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2707 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2708 Patch provided by Brad "anomie" Jorsch.
2710 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2711 Patch provided by Dean Brooks.
2713 TK/05 Add native DKIM support (does not depend on external libraries).
2715 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2716 Patch provided by Graeme Fowler.
2718 NM/16 Bugzilla 851: Documentation example syntax fix.
2720 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2722 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2725 NM/19 Bugzilla 745: TLS version reporting.
2726 Patch provided by Phil Pennock.
2728 NM/20 Bugzilla 167: bool: condition support.
2729 Patch provided by Phil Pennock.
2731 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2732 clients. Patch provided by Phil Pennock.
2734 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2735 Patch provided by Brad "anomie" Jorsch.
2737 NM/23 Bugzilla 687: Fix misparses in eximstats.
2738 Patch provided by Heiko Schlittermann.
2740 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2741 Patch provided by Heiko Schlittermann.
2743 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2744 plus update to original patch.
2746 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2748 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2749 Patch provided by David Brownlee.
2751 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2753 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2755 NM/30 Bugzilla 888: TLS documentation bugfixes.
2757 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2759 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2760 Unlike the original bugzilla I have changed all shell scripts in src tree.
2762 NM/33 Bugzilla 898: Transport filter timeout fix.
2763 Patch by Todd Rinaldo.
2765 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2766 Patch by Serge Demonchaux.
2768 NM/35 Bugzilla 39: Base64 decode bug fixes.
2769 Patch by Jakob Hirsch.
2771 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2773 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2775 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2777 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2783 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2784 ALT-N's libdkim that I have put here:
2785 http://duncanthrax.net/exim-experimental/
2787 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2788 were called 'true' and it seems that C99 defines that as a reserved
2789 keyword to be used with 'bool' variable types. That means you could
2790 not include C99-style headers which use bools without triggering
2791 build errors in sieve.c.
2793 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2794 as mailq or other aliases. Changed the --help handling significantly
2795 to do whats expected. exim_usage() emits usage/help information.
2797 SC/01 Added the -bylocaldomain option to eximstats.
2799 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2801 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2803 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2809 PH/01 Another patch from the Sieve maintainer.
2811 PH/02 When an IPv6 address is converted to a string for single-key lookup
2812 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2813 dots are used instead of colons so that keys in lsearch files need not
2814 contain colons. This was done some time before quoting was made available
2815 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2816 (notated using the quote facility) so as to distinguish them from IPv4
2817 keys. This meant that lookups for IP addresses in host lists did not work
2818 for iplsearch lookups.
2820 This has been fixed by arranging for IPv6 addresses to be expressed with
2821 colons if the lookup type is iplsearch. This is not incompatible, because
2822 previously such lookups could never work.
2824 The situation is now rather anomalous, since one *can* have colons in
2825 ordinary lsearch keys. However, making the change in all cases is
2826 incompatible and would probably break a number of configurations.
2828 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2831 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2832 conversion specification without a maximum field width, thereby enabling
2833 a rogue spamd server to cause a buffer overflow. While nobody in their
2834 right mind would setup Exim to query an untrusted spamd server, an
2835 attacker that gains access to a server running spamd could potentially
2836 exploit this vulnerability to run arbitrary code as the Exim user.
2838 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2839 $primary_hostname instead of what libspf2 thinks the hosts name is.
2841 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2842 a directory entry by the name of the lookup key. Previously, if a
2843 symlink pointed to a non-existing file or a file in a directory that
2844 Exim lacked permissions to read, a lookup for a key matching that
2845 symlink would fail. Now it is enough that a matching directory entry
2846 exists, symlink or not. (Bugzilla 503.)
2848 PH/03 The body_linecount and body_zerocount variables are now exported in the
2851 PH/04 Added the $dnslist_matched variable.
2853 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2854 This means they are set thereafter only if the connection becomes
2857 PH/06 Added the client_condition to authenticators so that some can be skipped
2858 by clients under certain conditions.
2860 PH/07 The error message for a badly-placed control=no_multiline_responses left
2861 "_responses" off the end of the name.
2863 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2865 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2866 (without spaces) instead of just copying the configuration text.
2868 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2870 PH/11 Added $max_received_linelength.
2872 PH/12 Added +ignore_defer and +include_defer to host lists.
2874 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2875 way in which PCRE > 7.0 is built.
2877 PH/14 Implemented queue_only_load_latch.
2879 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2880 MAIL command. The effect was to mangle the value on 64-bit systems.
2882 PH/16 Another patch from the Sieve maintainer.
2884 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2886 PH/18 If a system quota error occurred while trying to create the file for
2887 a maildir delivery, the message "Mailbox is full" was not appended to the
2888 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2889 only to a quota excession during the actual writing of the file.
2891 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2892 characters?) which causes problems in log lines. The DN values are now
2893 passed through string_printing() before being added to log lines.
2895 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2896 and InterBase are left for another time.)
2898 PH/21 Added message_body_newlines option.
2900 PH/22 Guard against possible overflow in moan_check_errorcopy().
2902 PH/23 POSIX allows open() to be a macro; guard against that.
2904 PH/24 If the recipient of an error message contained an @ in the local part
2905 (suitably quoted, of course), incorrect values were put in $domain and
2906 $local_part during the evaluation of errors_copy.
2912 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2913 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2916 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2917 issue a MAIL command.
2919 PH/02 In an ACL statement such as
2921 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2923 if a client was not listed at all, or was listed with a value other than
2924 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2925 the condition was not true (as it should be), so access was not denied.
2926 The bug was that the ! inversion was incorrectly passed on to the second
2927 item. This has been fixed.
2929 PH/03 Added additional dnslists conditions == and =& which are different from
2930 = and & when the dns lookup returns more than one IP address.
2932 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2933 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2935 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2936 FSYNC, which compiles an option called disable_fsync that allows for
2937 bypassing fsync(). The documentation is heavily laced with warnings.
2939 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2941 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2942 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2943 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2944 including adding "make clean"; (3) Added -fPIC when compiling the test
2945 dynamically loaded module, to get rid of a warning.
2947 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2948 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2949 = 0s. The bug is otherwise harmless.
2951 PH/07 There was a bug in the dovecot authenticator such that the value of
2952 $auth1 could be overwritten, and so not correctly preserved, after a
2953 successful authentication. This usually meant that the value preserved by
2954 the server_setid option was incorrect.
2956 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2958 PH/09 Installed PCRE release 7.0.
2960 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2961 run for batched SMTP input. It is now run at the start of every message
2962 in the batch. While fixing this I discovered that the process information
2963 (output by running exiwhat) was not always getting set for -bs and -bS
2964 input. This is fixed, and it now also says "batched" for BSMTP.
2966 PH/11 Added control=no_pipelining.
2968 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2969 patch, slightly modified), and move the expansion of helo_data till after
2970 the connection is made in the smtp transport (so it can use these
2973 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2975 PH/14 Added log_selector = +pid.
2977 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2979 PH/16 Add ${if forany and ${if forall.
2981 PH/17 Added dsn_from option to vary the From: line in DSNs.
2983 PH/18 Flush SMTP output before performing a callout, unless control =
2984 no_callout_flush is set.
2986 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2987 was true (the default) a successful delivery failed to delete the retry
2988 item, thus causing premature timeout of the address. The bug is now
2991 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2993 PH/21 Long custom messages for fakedefer and fakereject are now split up
2994 into multiline responses in the same way that messages for "deny" and
2995 other ACL rejections are.
2997 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2998 with slight modification.
3000 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3001 draft, changing the syntax and factoring some duplicate code".
3003 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3004 for deliveries of the second and subsequent messages over the same SMTP
3007 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3008 ${reduce, with only minor "tidies".
3010 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3012 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3013 expansion side effects.
3015 PH/27 When a message times out after an over-quota error from an Exim-imposed
3016 quota, the bounce message says "mailbox is full". This message was not
3017 being given when it was a system quota that was exceeded. It now should
3020 MH/03 Made $recipients available in local_scan(). local_scan() already has
3021 better access to the recipient list through recipients_list[], but
3022 $recipients can be useful in postmaster-provided expansion strings.
3024 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3025 in the case of a MAIL command with additional options following the
3026 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3027 were accidentally chopped off.
3029 PH/29 SMTP synchronization checks are implemented when a command is read -
3030 there is a check that no more input is waiting when there shouldn't be
3031 any. However, for some commands, a delay in an ACL can mean that it is
3032 some time before the response is written. In this time, more input might
3033 arrive, invalidly. So now there are extra checks after an ACL has run for
3034 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3035 pipelining has not been advertised.
3037 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3039 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3040 returned in the SMTP response when smtp_return_error_details was false.
3041 This has been fixed.
3043 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3044 instead of the C I/O that was originally supplied, because problems were
3045 reported on Solaris.
3047 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3048 Exim which did not show up earlier: it was assuming that a call to
3049 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3050 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3051 was a macro that became an assignment, so it seemed to work. This has
3052 changed to a proper function call with a void return, hence the compile
3053 error. Exim's code has been fixed.
3055 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3058 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3060 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3062 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3063 because it turns out to be too big or there is a timeout) there is no
3064 "Completed" line in the log. When some messages of this type were
3065 selected by exigrep, they were listed as "not completed". Others were
3066 picked up by some special patterns. I have improved the selection
3067 criteria to be more general.
3069 PH/38 The host_find_failed option in the manualroute router can now be set
3070 to "ignore", to completely ignore a host whose IP address cannot be
3071 found. If all hosts are ignored, the behaviour is controlled by the new
3072 host_all_ignored option.
3074 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3075 homing or because of multiple MX records with /mx) generated more than
3076 one IP address, and the following item turned out to be the local host,
3077 all the secondary addresses of the first item were incorrectly removed
3078 from the list, along with the local host and any following hosts (which
3079 is what is supposed to happen).
3081 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3082 whoever called Exim into the -H file. In the case of the daemon it was
3083 behaving confusingly. When first started, it used values for whoever
3084 started the daemon, but after a SIGHUP it used the Exim user (because it
3085 calls itself on a restart). I have changed the code so that it now always
3088 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3089 message are rejected with the same error (e.g. no authentication or bad
3090 sender address), and a DATA command is nevertheless sent (as can happen
3091 with PIPELINING or a stupid MUA), the error message that was given to the
3092 RCPT commands is included in the rejection of the DATA command. This is
3093 intended to be helpful for MUAs that show only the final error to their
3096 PH/42 Another patch from the Sieve maintainer.
3098 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3099 Eximstats - Fixed some broken HTML links and added missing column headers
3101 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3102 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3104 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3106 PH/43 Yet another patch from the Sieve maintainer.
3108 PH/44 I found a way to check for a TCP/IP connection going away before sending
3109 the response to the final '.' that terminates a message, but only in the
3110 case where the client has not sent further data following the '.'
3111 (unfortunately, this is allowed). However, in many cases there won't be
3112 any further data because there won't be any more messages to send. A call
3113 to select() can be used: if it shows that the input is "ready", there is
3114 either input waiting, or the socket has been closed. An attempt to read
3115 the next input character can distinguish the two cases. Previously, Exim
3116 would have sent an OK response which the client would never have see.
3117 This could lead to message repetition. This fix should cure that, at
3118 least in a lot of common cases.
3120 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3121 advertised in response to EHLO.
3127 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3128 fixed by 4.65/MH/01 (is this a record?) are fixed:
3130 (i) An empty string was always treated as zero by the numeric comparison
3131 operators. This behaviour has been restored.
3133 (ii) It is documented that the numeric comparison operators always treat
3134 their arguments as decimal numbers. This was broken in that numbers
3135 starting with 0 were being interpreted as octal.
3137 While fixing these problems I realized that there was another issue that
3138 hadn't been noticed. Values of message_size_limit (both the global option
3139 and the transport option) were treated as octal if they started with 0.
3140 The documentation was vague. These values are now always treated as
3141 decimal, and I will make that clear in the documentation.
3147 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3148 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3151 MH/01 Don't check that the operands of numeric comparison operators are
3152 integers when their expansion is in "skipping" mode (fixes bug
3153 introduced by 4.64-PH/07).
3155 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3156 child addresses, Exim now panics and dies. Previously, because the count
3157 is held in a short int, deliveries were likely to be lost. As such a
3158 large number of recipients for a single message is ridiculous
3159 (performance will be very, very poor), I have chosen to impose a limit
3160 rather than extend the field.
3166 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3167 leftover -K file (the existence of which was triggered by #402).
3168 While we were at it, introduced process PID as part of the -K
3169 filename. This should rule out race conditions when creating
3172 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3173 processing considerably. Previous code took too long for large mails,
3174 triggering a timeout which in turn triggers #401.
3176 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3177 in the DK code in transports.c. sendfile() is not really portable,
3178 hence the _LINUX specificness.
3180 TF/01 In the add_headers option to the mail command in an Exim filter,
3181 there was a bug that Exim would claim a syntax error in any
3182 header after the first one which had an odd number of characters
3185 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3186 callout verification, Exim cached a "reject" for the entire domain. This
3187 is correct for most verifications, but it is not correct for a recipient
3188 verification with use_sender or use_postmaster set, because in that case
3189 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3190 case of MAIL FROM:<> rejection from other early rejections (e.g.
3191 rejection of HELO). When verifying a recipient using a non-null MAIL
3192 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3193 Whatever the result of the callout, the value of the domain cache is
3194 left unchanged (for any other kind of callout, getting as far as trying
3195 RCPT means that the domain itself is ok).
3197 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3200 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3201 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3202 session, the daemon ignores ECONNECT errors and logs others; it now
3203 ignores EPIPE as well.
3205 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3206 (quoted-printable decoding).
3208 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3209 later the small subsequent patch to fix an introduced bug.
3211 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3213 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3215 PH/08 An error is now given if message_size_limit is specified negative.
3217 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3218 to be given (somewhat) arbitrary names.
3220 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3223 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3224 miscellaneous code fixes
3226 PH/10 Added the log_reject_target ACL modifier to specify where to log
3229 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3230 hostname. This is wrong, because it relates to the incoming message (and
3231 probably the interface on which it is arriving) and not to the outgoing
3232 callout (which could be using a different interface). This has been
3233 changed to use the value of the helo_data option from the smtp transport
3234 instead - this is what is used when a message is actually being sent. If
3235 there is no remote transport (possible with a router that sets up host
3236 addresses), $smtp_active_hostname is used.
3238 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3239 tweaks were necessary in order to get it to work (see also 21 below):
3240 (a) The code assumed that strncpy() returns a negative number on buffer
3241 overflow, which isn't the case. Replaced with Exim's string_format()
3243 (b) There were several signed/unsigned issues. I just did the minimum
3244 hacking in of casts. There is scope for a larger refactoring.
3245 (c) The code used strcasecmp() which is not a standard C function.
3246 Replaced with Exim's strcmpic() function.
3247 (d) The code set only $1; it now sets $auth1 as well.
3248 (e) A simple test gave the error "authentication client didn't specify
3249 service in request". It would seem that Dovecot has changed its
3250 interface. Fortunately there's a specification; I followed it and
3251 changed what the client sends and it appears to be working now.
3253 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3256 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3257 address A is aliased to B and C, where B exists and C does not. Without
3258 -v the output is "A verified" because verification stops after a
3259 successful redirection if more than one address is generated. However,
3260 with -v the child addresses are also verified. Exim was outputting "A
3261 failed to verify" and then showing the successful verification for C,
3262 with its parentage. It now outputs "B failed to verify", showing B's
3263 parentage before showing the successful verification of C.
3265 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3266 look up a TXT record in a specific list after matching in a combined
3269 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3270 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3271 they consult the DNS. I had assumed they would set it the way they
3272 wanted; and indeed my experiments on Linux seem to show that in some
3273 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3274 To be on the safe side, however, I have now made the interface to
3275 host_find_byname() similar to host_find_bydns(), with an argument
3276 containing the DNS resolver options. The host_find_byname() function now
3277 sets these options at its start, just as host_find_bydns() does. The smtp
3278 transport options dns_qualify_single and dns_search_parents are passed to
3279 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3280 of host_find_byname() use the default settings of RES_DEFNAMES
3281 (qualify_single) but not RES_DNSRCH (search_parents).
3283 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3284 spool_read_header() do less string testing, by means of a preliminary
3285 switch on the second character of optional "-foo" lines. (This is
3286 overdue, caused by the large number of possibilities that now exist.
3287 Originally there were few.) While I was there, I also converted the
3288 str(n)cmp tests so they don't re-test the leading "-" and the first
3289 character, in the hope this might squeeze out yet more improvement.
3291 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3292 flag allowing group syntax was set by the header_syntax check but not
3293 turned off, possible causing trouble later; (2) The flag was not being
3294 set at all for the header_verify test, causing "group"-style headers to
3295 be rejected. I have now set it in this case, and also caused header_
3296 verify to ignore an empty address taken from a group. While doing this, I
3297 came across some other cases where the code for allowing group syntax
3298 while scanning a header line wasn't quite right (mostly, not resetting
3299 the flag correctly in the right place). These bugs could have caused
3300 trouble for malformed header lines. I hope it is now all correct.
3302 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3303 with the "reply" argument non-NULL. The code, however (which originally
3304 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3305 but it didn't always do it. This confused somebody who was copying the
3306 code for some other use. I have removed all the tests.
3308 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3309 feature that was used to support insecure browsers during the U.S. crypto
3310 embargo. It requires special client support, and Exim is probably the
3311 only MTA that supported it -- and would never use it because real RSA is
3312 always available. This code has been removed, because it had the bad
3313 effect of slowing Exim down by computing (never used) parameters for the
3314 RSA_EXPORT functionality.
3316 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3317 authenticator to fail if there's a tab character in the incoming data
3318 (there should never be unless someone is messing about, as it's supposed
3319 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3320 if the connection is using TLS or if the remote IP is the same as the
3321 local IP, and the "valid-client-cert option" if a client certificate has
3324 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3325 authenticators. This can be used for authorization after authentication
3326 succeeds. (In the case of plaintext, it servers for both authentication
3329 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3330 if any retry times were supplied.
3332 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3333 connection, where there is no client IP address to check. In this
3334 situation, the verify now always succeeds.
3336 PH/25 Applied John Jetmore's -Mset patch.
3338 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3340 PH/27 In a string expansion for a processed (not raw) header when multiple
3341 headers of the same name were present, leading whitespace was being
3342 removed from all of them, but trailing whitespace was being removed only
3343 from the last one. Now trailing whitespace is removed from each header
3344 before concatenation. Completely empty headers in a concatenation (as
3345 before) are ignored.
3347 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3348 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3350 PH/29 [Removed. This was a change that I later backed out, and forgot to
3351 correct the ChangeLog entry (that I had efficiently created) before
3352 committing the later change.]
3354 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3355 address errors (4xx response to RCPT) over the same connection as other
3356 messages routed to the same hosts. Such deliveries are always "forced",
3357 so retry times are not inspected. This resulted in far too many retries
3358 for the affected addresses. The effect occurred only when there were more
3359 hosts than the hosts_max_try setting in the smtp transport when it had
3360 the 4xx errors. Those hosts that it had tried were not added to the list
3361 of hosts for which the message was waiting, so if all were tried, there
3362 was no problem. Two fixes have been applied:
3364 (i) If there are any address or message errors in an SMTP delivery, none
3365 of the hosts (tried or untried) are now added to the list of hosts
3366 for which the message is waiting, so the message should not be a
3367 candidate for sending over the same connection that was used for a
3368 successful delivery of some other message. This seems entirely
3369 reasonable: after all the message is NOT "waiting for some host".
3370 This is so "obvious" that I'm not sure why it wasn't done
3371 previously. Hope I haven't missed anything, but it can't do any
3372 harm, as the worst effect is to miss an optimization.
3374 (ii) If, despite (i), such a delivery is accidentally attempted, the
3375 routing retry time is respected, so at least it doesn't keep
3376 hammering the server.
3378 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3379 in ${readsocket because some servers need this prod.
3381 PH/32 Added some extra debug output when updating a wait-xxx database.
3383 PH/33 The hint "could be header name not terminated by colon", which has been
3384 given for certain expansion errors for a long time, was not being given
3385 for the ${if def:h_colon_omitted{... case.
3387 PH/34 The spec says: "With one important exception, whenever a domain list is
3388 being scanned, $domain contains the subject domain." There was at least
3389 one case where this was not true.
3391 PH/35 The error "getsockname() failed: connection reset by peer" was being
3392 written to the panic log as well as the main log, but it isn't really
3393 panic-worthy as it just means the connection died rather early on. I have
3394 removed the panic log writing for the ECONNRESET error when getsockname()
3397 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3398 runs only) independently of the message's sender address. This meant
3399 that, if the 4xx error was in fact related to the sender, a different
3400 message to the same recipient with a different sender could confuse
3401 things. In particular, this can happen when sending to a greylisting
3402 server, but other circumstances could also provoke similar problems.
3403 I have changed the default so that the retry time for these errors is now
3404 based a combination of the sender and recipient addresses. This change
3405 can be overridden by setting address_retry_include_sender=false in the
3408 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3409 remote server are returned as part of bounce messages. This was not
3410 happening for LMTP over a pipe (the lmtp transport), but now it is the
3411 same for both kinds of LMTP.
3413 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3414 in header lines that were in fact CNAMEs. This is no longer the case.
3416 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3417 and queue runs started by the daemon processed all messages. This has
3418 been fixed so that -R and -S can now usefully be given with -q<time>.
3420 PH/40 Import PCRE release 6.7 (fixes some bugs).
3422 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3424 PH/42 Give an error if -q is specified more than once.
3426 PH/43 Renamed the variables $interface_address and $interface_port as
3427 $received_ip_address and $received_port, to make it clear that these
3428 values apply to message reception, and not to the outgoing interface when
3429 a message is delivered. (The old names remain recognized, of course.)
3431 PH/44 There was no timeout on the connect() call when using a Unix domain
3432 socket in the ${readsocket expansion. There now is.
3434 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3435 be meaningful with "accept".
3437 SC/01 Eximstats V1.43
3438 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3440 SC/02 Eximstats V1.44
3441 Use a glob alias rather than an array ref in the generated
3442 parser. This improves both readability and performance.
3444 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3445 Collect SpamAssassin and rejection statistics.
3446 Don't display local sender or destination tables unless
3447 there is data to show.
3448 Added average volumes into the top table text output.
3450 SC/04 Eximstats V1.46
3451 Collect data on the number of addresses (recipients)
3452 as well as the number of messages.
3454 SC/05 Eximstats V1.47
3455 Added 'Message too big' to the list of mail rejection
3456 reasons (thanks to Marco Gaiarin).
3458 SC/06 Eximstats V1.48
3459 Mainlog lines which have GMT offsets and are too short to
3460 have a flag are now skipped.
3462 SC/07 Eximstats V1.49 (Alain Williams)
3463 Added the -emptyok flag.
3465 SC/08 Eximstats V1.50
3466 Fixes for obtaining the IP address from reject messages.
3468 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3469 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3470 whitespace changes from 4.64-PH/27
3472 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3475 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3476 are found, allow negative numbers in numeric criteria)
3478 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3480 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3481 to match changes made in 4.64-PH/43
3483 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3485 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3486 "domain" is actually an IP address, and force a failure. This locks out
3487 those revolvers/nameservers that support "A-for-A" lookups, in
3488 contravention of the specifications.
3490 PH/48 When a host name was looked up from an IP address, and the subsequent
3491 forward lookup of the name timed out, the host name was left in
3492 $sender_host_name, contrary to the specification.
3494 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3495 restricted to single-key lookups, Exim was not diagnosing an error if
3496 * or *@ was used with a query-style lookup.
3498 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3500 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3501 long ago, but noone interested enough thought of it. Let's just say that
3502 the "1.1" means that there are some new functions that weren't there at
3503 some point in the past.
3505 PH/51 Error processing for expansion failure of helo_data from an smtp
3506 transport during callout processing was broken.
3508 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3509 tested/used via the -bh/-bhc/-bs options.
3511 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3512 bug, fixed in subsequent PCRE releases).
3514 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3515 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3517 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3523 SC/01 Use a glob alias rather than an array ref in eximstats generated
3524 parser. This improves both readability and performance.
3526 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3527 Don't display local sender or destination tables in eximstats unless
3528 there is data to show.
3529 Added average volumes into the eximstats top table text output.
3531 SC/03 Collect data on the number of addresses (recipients) as well
3532 as the number of messages in eximstats.
3534 TF/01 Correct an error in the documentation for the redirect router. Exim
3535 does (usually) call initgroups() when daemonizing.
3537 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3538 with consistent privilege compared to when running as a daemon.
3540 TF/03 Note in the spec that $authenticated_id is not set for local
3541 submissions from trusted users.
3543 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3544 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3546 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3547 by adding some example configuration directives to the default
3548 configuration file. A little bit of work is required to uncomment the
3549 directives and define how usernames and passwords are checked, but
3550 there is now a framework to start from.
3552 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3553 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3554 without this. I don't know how relevant this is to other LDAP libraries.
3556 PH/02 Add the verb name to the "unknown ACL verb" error.
3558 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3560 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3562 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3563 directory not expanded when it should be if an expanded home directory
3564 was set for the address (which is overridden by the transport).
3566 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3569 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3570 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3571 because it is too late at that time, and has no effect.
3573 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3574 security issue with \' (bugzilla #107). I could not use the
3575 PQescapeStringConn() function, because it needs a PGconn value as one of
3578 PH/08 When testing addresses using -bt, indicate those final addresses that
3579 are duplicates that would not cause an additional delivery. At least one
3580 person was confused, thinking that -bt output corresponded to deliveries.
3581 (Suppressing duplicates isn't a good idea as you lose the information
3582 about possibly different redirections that led to the duplicates.)
3584 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3585 systems where poll() doesn't work, in particular OS X.
3587 PH/10 Added more information to debugging output for retry time not reached.
3589 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3590 operations in malware.c.
3592 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3595 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3596 syslog_duplication set false, log lines that would normally be written
3597 both the the main log and to the reject log were not written to syslog at
3600 PH/14 In the default configuration, change the use of "message" in ACL warn
3601 statements to "add_header".
3603 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3604 not followed by a command (e.g. "seen endif").
3606 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3607 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3610 PH/17 Added extra conditions to the default value of delay_warning_condition
3614 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3615 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3616 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3619 The Auto-Submitted: and various List- headers are standardised, whereas I
3620 don't think Precedence: ever was.
3622 PH/18 Refactored debugging code in route_finduser() to show more information,
3623 in particular, the error code if getpwnam() issues one.
3625 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3626 This is apparently needed in addition to the PH/07 change above to avoid
3627 any possible encoding problems.
3629 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3630 but not after initializing Perl.
3632 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3633 output them only if debugging. By default they are written stderr,
3634 apparently, which is not desirable.
3636 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3639 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3642 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3644 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3645 authenticated or an ident call has been made. Suppress the default
3646 values for $authenticated_id and $authenticated_sender (but permit -oMai
3647 and -oMas) when testing with -bh.
3649 PH/24 Re-jigged the order of the tests in the default configuration so that the
3650 tests for valid domains and recipients precede the DNS black list and CSA
3651 tests, on the grounds that those ones are more expensive.
3653 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3654 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3655 This bug exists in every version of Exim that I still have, right back to
3658 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3659 However, an attempt to turn on case-sensitivity in a regex key by
3660 including (?-i) didn't work because the subject string was already
3661 lowercased, and the effects were non-intuitive. It turns out that a
3662 one-line patch can be used to allow (?-i) to work as expected.
3668 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3669 other effects) broke the use of negated acl sub-conditions.
3671 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3674 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3675 "Deny" causes Exim to reject the incoming connection with a 554 error.
3676 Unfortunately, if there is a major crisis, such as a disk failure,
3677 tcp-wrappers gives "deny", whereas what one would like would be some
3678 kind of temporary error. A kludge has been added to help with this.
3679 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3680 554 error is used if errno is still zero or contains ENOENT (which occurs
3681 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3684 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3686 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3687 errors. Otherwise a message that provokes a temporary error (when other
3688 messages do not) can cause a whole host to time out.
3690 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3691 addresses were routed directly to files or pipes from a redirect router.
3692 File deliveries just didn't batch; pipe deliveries might have suffered
3695 PH/06 A failure to get a lock for a hints database would erroneously always say
3696 "Failed to get write lock", even when it was really a read lock.
3698 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3699 of 0600. This has been changed to use the value of the lockfile_mode
3700 option (which defaults to 0600).
3702 PH/08 Applied small patch from the Sieve maintainer.
3704 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3705 folder from quota calculations, a direct delivery into this folder messed
3706 up the contents of the maildirsize file. This was because the regex was
3707 used only to exclude .Trash (or whatever) when the size of the mailbox
3708 was calculated. There was no check that a delivery was happening into an
3709 excluded directory. This bug has been fixed by ignoring all quota
3710 processing for deliveries into excluded directories.
3712 PH/10 Added the maildirfolder_create_regex option to appendfile.
3718 PH/01 The code for finding all the local interface addresses on a FreeBSD
3719 system running IPv6 was broken. This may well have applied to all BSD
3720 systems, as well as to others that have similar system calls. The broken
3721 code found IPv4 interfaces correctly, but gave incorrect values for the
3722 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3723 that it would not match correctly against @[] and not recognize the IPv6
3726 PH/02 The ipliteral router was not recognizing addresses of the form user@
3727 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3729 PH/03 Added disable_ipv6.
3731 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3732 decoded form, because it is most often used to construct To: headers
3733 lines in autoreplies, and the decoded form may well be syntactically
3734 invalid. However, $reply_address has leading white space removed, and all
3735 newlines turned into spaces so that the autoreply transport does not
3738 PH/05 If group was specified without a user on a router, and no group or user
3739 was specified on a transport, the group from the router was ignored.
3741 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3742 for visible compile-time settings that can be used to change these
3743 numbers, for those that want even more. Backwards compatibility with old
3744 spool files has been maintained. However, going back to a previous Exim
3745 release will lost any variables that are in spool files.
3747 PH/07 Two small changes when running in the test harness: increase delay when
3748 passing a TCP/IP connection to a new process, in case the original
3749 process has to generate a bounce, and remove special handling of
3750 127.0.0.2 (sic), which is no longer necessary.
3752 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3753 be the same on different OS.
3755 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3758 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3759 whether --show-vars was specified or not
3761 JJ/02 exipick: Added support for new ACL variable spool format introduced
3764 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3765 syntactically invalid From: or Reply-to: line, and a filter used this to
3766 generate an autoreply, and therefore failed to obtain an address for the
3767 autoreply, Exim could try to deliver to a non-existent relative file
3768 name, causing unrelated and misleading errors. What now happens is that
3769 it logs this as a hard delivery error, but does not attempt to create a
3772 PH/11 The exinext utility has a -C option for testing purposes, but although
3773 the given file was scanned by exinext itself; it wasn't being passed on
3774 when Exim was called.
3776 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3777 an end-of-file indication when reading a command response.
3779 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3780 compiled. In many other places in Exim, IPv6 addresses are always
3781 recognized, so I have changed this. It also means that IPv4 domain
3782 literals of the form [IPV4:n.n.n.n] are now always recognized.
3784 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3785 used if the router is not running as root, for example, when verifying at
3786 ACL time, or when using -bh. The debugging output from this situation was
3787 non-existent - all you got was a failure to exec. I have made two
3790 (a) Failures to set uid/gid, the current directory, or a process leader
3791 in a subprocess such as that created by queryprogram now generate
3792 suitable debugging output when -d is set.
3794 (b) The queryprogram router detects when it is not running as root,
3795 outputs suitable debugging information if -d is set, and then runs
3796 the subprocess without attempting to change uid/gid.
3798 PH/15 Minor change to Makefile for building test_host (undocumented testing
3801 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3802 additional section of a DNS packet that returns MX or SRV records.
3803 Instead, it always explicitly searches for A/AAAA records. This avoids
3804 major problems that occur when a DNS server includes only records of one
3805 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3806 fixed another bug: if SRV records were looked up and the corresponding
3807 address records were *not* found in the additional section, the port
3808 values from the SRV records were lost.
3810 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3811 using the correct key (the original address) when searching the retry
3812 rules in order to find which one to use for generating the retry hint.
3814 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3815 adding the default one. Similarly, if it contains a Reply-To: header, the
3816 errors_reply_to option, if set, is not used.
3818 PH/19 When calculating a retry time, Exim used to measure the "time since
3819 failure" by looking at the "first failed" field in the retry record. Now
3820 it does not use this if it is later than than the arrival time of the
3821 message. Instead it uses the arrival time. This makes for better
3822 behaviour in cases where some deliveries succeed, thus re-setting the
3823 "first failed" field. An example is a quota failure for a huge message
3824 when small messages continue to be delivered. Without this change, the
3825 "time since failure" will always be short, possible causing more frequent
3826 delivery attempts for the huge message than are intended.
3827 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3829 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3830 $1, $2, $3) because the numerical variables can be reset during some
3831 expansion items (e.g. "match"), thereby losing the authentication data.
3833 PH/21 Make -bV show the size of off_t variables so that the test suite can
3834 decide whether to run tests for quotas > 2G.
3836 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3837 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3838 filecount value is greater than 2G or if a quota value is greater than 2G
3839 on a system where the size of off_t is not greater than 4, a panic error
3842 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3843 never match. The debug and -bh output now contains an explicit error
3844 message indicating a malformed IPv4 address or mask.
3846 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3847 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3848 PH/23 above applies.
3850 PH/25 Do not write to syslog when running in the test harness. The only
3851 occasion when this arises is a failure to open the main or panic logs
3852 (for which there is an explicit test).
3854 PH/26 Added the /no_tell option to "control=freeze".
3856 PH/27 If a host name lookup failed very early in a connection, for example, if
3857 the IP address matched host_lookup and the reverse lookup yielded a name
3858 that did not have a forward lookup, an error message of the form "no IP
3859 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3860 could be logged. Now it outputs the IP address instead of "NULL".
3862 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3863 allows the sender and the authenticated sender to be set when
3864 submitting a message from within Exim. Since child_open_exim() is
3865 documented for local_scan(), the new function should be too.
3867 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3868 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3869 results in an empty string is now treated as unset.
3871 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3873 PH/31 Added sender_verify_fail logging option.
3875 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3876 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3877 tidied the source and removed it altogether.
3879 PH/33 When a queue run was abandoned because the load average was too high, a
3880 log line was always written; now it is written only if the queue_run log
3881 selector is set. In addition, the log line for abandonment now contains
3882 information about the queue run such as the pid. This is always present
3883 in "start" and "stop" lines but was omitted from the "abandon" line.
3885 PH/34 Omit spaces between a header name and the colon in the error message that
3886 is given when verify = headers_syntax fails (if there are lots of them,
3887 the message gets confusing).
3889 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3890 names, as there are now some PTR records that contain slashes. This check
3891 is only to protect against broken name servers that fall over on strange
3892 characters, so the fact that it applies to all lookups doesn't matter.
3894 PH/36 Now that the new test suite is complete, we can remove some of the
3895 special code in Exim that was needed for the old test suite. For example,
3896 sorting DNS records because real resolvers return them in an arbitrary
3897 order. The new test suite's fake resolver always returns records in the
3900 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3901 bounces) except when queue_only is set, to avoid logging races between
3902 the different processes.
3904 PH/38 Panic-die if .include specifies a non-absolute path.
3906 PH/39 A tweak to the "H" retry rule from its user.
3908 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3909 a label. They prevented compilation on older perls.
3911 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3912 a warning to be raised on newish perls.
3914 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3915 on queue. Changes to match documented behaviour of showing count of
3916 messages matching specified criteria.
3918 PH/40 Changed the default ident timeout from 30s to 5s.
3920 PH/41 Added support for the use of login_cap features, on those BSD systems
3921 that have them, for controlling the resources used by pipe deliveries.
3923 PH/42 The content-scanning code uses fopen() to create files in which to put
3924 message data. Previously it was not paying any attention to the mode of
3925 the files. Exim runs with umask(0) because the rest of the code creates
3926 files with open(), and sets the required mode explicitly. Thus, these
3927 files were ending up world-writeable. This was not a big issue, because,
3928 being within the spool directory, they were not world-accessible. I have
3929 created a function called modefopen, which takes an additional mode
3930 argument. It sets umask(777), creates the file, chmods it to the required
3931 mode, then resets the umask. All the relevant calls to fopen() in the
3932 content scanning code have been changed to use this function.
3934 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3935 to 24 hours. This avoids potential overflow problems when processing G
3936 and H retry rules. I suspect nobody ever tinkers with this value.
3938 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3940 PH/45 When the plaintext authenticator is running as a client, the server's
3941 challenges are checked to ensure they are valid base64 strings. By
3942 default, the authentication attempt is cancelled if an invalid string is
3943 received. Setting client_ignore_invalid_base64 true ignores these errors.
3944 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3945 they are received. Thus, the responses can be made to depend on the
3946 challenges. If an invalid string is ignored, an empty string is placed in
3949 PH/46 Messages that are created by the autoreply transport now contains a
3950 References: header, in accordance with RFCs 2822 and 3834.
3952 PH/47 Added authenticated_sender_force to the smtp transport.
3954 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3956 PH/49 Installed latest patch from the Sieve maintainer.
3958 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3959 was also set, the appendfile transport was unnecessarily scanning a
3960 directory of message files (e.g. for maildir delivery) to find the count
3961 of files (along with the size), even though it did not need this
3962 information. It now does the scan only if it needs to find either the
3963 size of the count of files.
3965 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3967 PH/52 Two bugs concerned with error handling when the smtp transport is
3970 (i) Exim was not creating retry information for temporary errors given
3971 for individual recipients after the DATA command when the smtp transport
3972 was used in LMTP mode. This meant that they could be retried too
3973 frequently, and not timed out correctly.
3975 (ii) Exim was setting the flag that allows error details to be returned
3976 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3977 recipients that were returned after the DATA command.
3979 PH/53 This is related to PH/52, but is more general: for any failing address,
3980 when detailed error information was permitted to be returned to the
3981 sender, but the error was temporary, then after the final timeout, only
3982 "retry timeout exceeded" was returned. Now it returns the full error as
3983 well as "retry timeout exceeded".
3985 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3986 do this, and (what is worse) MTAs that accept it.
3988 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3989 will now be deprecated.
3991 PH/56 New os.c-cygwin from the Cygwin maintainer.
3993 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3994 formats (previously only available in exim formats via -bpr, -bpru,
3995 and -bpra. Now also available in native and exiqgrep formats)
3997 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3998 with very large, slow to parse queues
4000 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4002 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4004 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4005 responses to authentication challenges, though it was showing the
4006 challenges; (ii) I've removed the CR characters from the debug output for
4009 PH/58 Allow for the insertion of a newline as well as a space when a string
4010 is turned into more than one encoded-word during RFC 2047 encoding. The
4011 Sieve code now uses this.
4013 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4014 data_4xx, lost_connection, tls_required.
4016 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4017 message was being sent as an SMTP response.
4019 PH/61 Add -l and -k options to exicyclog.
4021 PH/62 When verifying, if an address was redirected to one new address, so that
4022 verification continued, and the new address failed or deferred after
4023 having set something in $address_data, the value of $address_data was not
4024 passed back to the ACL. This was different to the case when no
4025 redirection occurred. The value is now passed back in both cases.
4027 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4028 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4029 use login_cap.h, so on its own it isn't the distinguishing feature. The
4030 new name refers directly to the setclassresources() function.
4032 PH/65 Added configuration files for NetBSD3.
4034 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4036 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4037 is preferred over IPv4.
4039 PH/68 The bounce_return_message and bounce_return_body options were not being
4040 honoured for bounces generated during the reception of non-SMTP messages.
4041 In particular, this applied to messages rejected by the ACL. This bug has
4042 been fixed. However, if bounce_return_message is true and bounce_return_
4043 body is false, the headers that are returned for a non-SMTP message
4044 include only those that have been read before the error was detected.
4045 (In the case of an ACL rejection, they have all been read.)
4047 PH/69 The HTML version of the specification is now built in a directory called
4048 spec_html instead of spec.html, because the latter looks like a path with
4049 a MIME-type, and this confuses some software.
4051 PH/70 Catch two compiler warnings in sieve.c.
4053 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4054 function verify_get_ident() calls ip_connect() to connect a socket, but
4055 if the "connect()" function timed out, ip_connect() used to close the
4056 socket. However, verify_get_ident() also closes the socket later, and in
4057 between Exim writes to the log, which may get opened at this point. When
4058 the socket was closed in ip_connect(), the log could get the same file
4059 descriptor number as the socket. This naturally causes chaos. The fix is
4060 not to close the socket in ip_connect(); the socket should be closed by
4061 the function that creates it. There was only one place in the code where
4062 this was missing, in the iplookup router, which I don't think anybody now
4063 uses, but I've fixed it anyway.
4065 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4066 well as to direct DNS lookups. Otherwise the handling of names in host
4067 lists is inconsistent and therefore confusing.
4073 PH/01 Two changes to the default runtime configuration:
4075 (1) Move the checks for relay_from_hosts and authenticated clients from
4076 after to before the (commented out) DNS black list checks.
4078 (2) Add control=submission to the relay_from_hosts and authenticated
4079 clients checks, on the grounds that messages accepted by these
4080 statements are most likely to be submissions.
4082 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4084 (1) Generate an error if the third argument for the ${prvs expansion is
4087 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4090 (3) Reset the variables that are obtained from the first argument of
4091 ${prvscheck and used in the second argument before leaving the code,
4092 because their memory is reclaimed, so using them afterwards may do
4095 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4096 one (it's much easier than Tom thought :-).
4098 (5) Because of (4), we can now allow for the use of $prvscheck_result
4099 inside the third argument.
4101 PH/03 For some reason, the default setting of PATH when running a command from
4102 a pipe transport was just "/usr/bin". I have changed it to
4105 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4106 anything to be listed in the output from -bV.
4108 PH/05 When a filter generated an autoreply, the entire To: header line was
4109 quoted in the delivery log line, like this:
4111 => >A.N.Other <ano@some.domain> <original@ddress> ...
4113 This has been changed so that it extracts the operative address. There
4114 may be more than one such address. If so, they are comma-separated, like
4117 => >ano@some.domain,ona@other.domain <original@ddress> ...
4119 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4120 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4121 not being looked up in the rDNS to get a host name, Exim was showing the
4122 IP address twice in Received: lines, even though the IP addresses were
4123 identical. For example:
4125 Received: from [1.2.3.4] (helo=[1.2.3.4])
4127 However, if the real host name was known, it was omitting the HELO data
4128 if it matched the actual IP address. This has been tidied up so that it
4129 doesn't show the same IP address twice.
4131 PH/07 When both +timestamp and +memory debugging was on, the value given by
4132 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4133 called by the debug printing, thereby overwriting the timestamp buffer.
4134 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4136 PH/08 When the original message was included in an autoreply transport, it
4137 always said "this is a copy of the message, including all the headers",
4138 even if body_only or headers_only was set. It now gives an appropriate
4141 PH/09 Applied a patch from the Sieve maintainer which:
4143 o fixes some comments
4144 o adds the (disabled) notify extension core
4145 o adds some debug output for the result of if/elsif tests
4146 o points to the current vacation draft in the documentation
4147 and documents the missing references header update
4151 o fixes a bug in processing the envelope test (when testing
4152 multiple envelope elements, the last element determined the
4155 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4156 Electronic Mail") by including:
4158 Auto-submitted: auto-generated
4160 in the messages that it generates (bounce messages and others, such as
4161 warnings). In the case of bounce messages for non-SMTP messages, there was
4162 also a typo: it was using "Auto_submitted" (underscore instead of
4163 hyphen). Since every message generated by Exim is necessarily in response
4164 to another message, thes have all been changed to:
4166 Auto-Submitted: auto-replied
4168 in accordance with these statements in the RFC:
4170 The auto-replied keyword:
4172 - SHOULD be used on messages sent in direct response to another
4173 message by an automatic process,
4175 - MUST NOT be used on manually-generated messages,
4177 - MAY be used on Delivery Status Notifications (DSNs) and Message
4178 Disposition Notifications (MDNs),
4180 - MUST NOT be used on messages generated by automatic or periodic
4181 processes, except for messages which are automatic responses to
4184 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4185 to the default Received: header definition.
4187 PH/12 Added log selector acl_warn_skipped (default on).
4189 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4190 variables because (a) they are in the wrong storage pool and (b) even if
4191 they were copied, it wouldn't work properly because of the caching.
4193 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4194 checking when decoding. Apparently there are clients that generate
4195 overlong encoded strings. Why am I not surprised?
4197 PH/15 If the first argument of "${if match_address" was not empty, but did not
4198 contain an "@" character, Exim crashed. Now it writes a panic log message
4199 and treats the condition as false.
4201 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4203 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4204 extension "envelope-auth". The code is finished and in agreement with
4205 other implementations, but there is no documentation so far and in fact,
4206 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4207 not changing the active code.
4209 Print executed "if" and "elsif" statements when debugging is used. This
4210 helps a great deal to understand what a filter does.
4212 Document more things not specified clearly in RFC3028. I had all this
4213 sorted out, when out of a sudden new issues came to my mind. Oops."
4215 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4218 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4219 canonical form (as documented). However, after a host name lookup from
4220 the IP address, check_host() was doing a simple string comparison with
4221 addresses acquired from the DNS when checking that the found name did
4222 have the original IP as one of its addresses. Since any found IPv6
4223 addresses are likely to be in abbreviated form, the comparison could
4224 fail. Luckily, there already exists a function for doing the comparison
4225 by converting both addresses to binary, so now that is used instead of
4226 the text comparison.
4228 PH/20 There was another similar case to PH/19, when a complete host name was
4229 given in a host list; looking up its IP address could give an abbreviated
4230 form, whereas the current host's name might or might not be abbreviated.
4231 The same fix has been applied.
4237 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4238 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4241 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4242 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4244 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4246 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4247 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4248 Also, for "Auto-Submitted", treat anything other than "no" as
4249 non-personal, in accordance with RFC 3834. (Previously it treated
4250 anything starting "auto-" as non-personal.)
4252 TF/01 The control=submission/name=... option had a problem with syntax
4253 errors if the name included a slash character. The /name= option
4254 now slurps the rest of the string, so it can include any characters
4255 but it must come last in the list of options (after /sender_retain
4258 PH/05 Some modifications to the interface to the fake nameserver for the new
4266 TK/01 Added the "success_on_redirect" address verification option. See
4267 NewStuff for rationale and an example.
4269 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4271 PH/02 Patch to exigrep to allow it to work on syslog lines.
4273 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4274 fread() to skip over the body file's header line, because in Cygwin the
4275 header line is locked and is inaccessible.
4277 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4278 co-exist for some time) to make it clear that it is the Exim ID that is
4279 referenced, not the Message-ID: header line.
4281 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4282 string_format() function, because snprintf() does not exist on all
4285 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4286 ${stat: expansion item.
4288 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4289 protocol synchronization error", to keep the pedants happy.
4291 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4292 well as for IRIX systems, when gcc is being used. See the host.c source
4295 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4297 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4300 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4301 transport and to the smtp transport in LMTP mode.
4303 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4305 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4306 run a filter in a subprocess. This could lead to confusion in subsequent
4307 lookups in the parent process. There should also be a search_tidyup() at
4308 the end of the subprocess.
4310 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4311 only if the host matched helo_try_verify_hosts, which caused the
4312 verification to occur when the EHLO/HELO command was issued. The ACL just
4313 tested the remembered result. Now, if a previous verification attempt has
4314 not happened, "verify = helo" does it there and then.
4316 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4318 TK/03 Fix log output including CR from clamd.
4320 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4321 exist provoked a memory error which could cause a segfault.
4323 PH/15 Installed PCRE 6.2
4325 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4327 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4328 of the problem. Specifically, suggested +O2 rather than +O1 for the
4331 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4333 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4334 record, the port was not correctly logged when the outgoing_port log
4335 selector was set (it logged the transort's default port).
4337 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4338 fallback_hosts, and "hosts" in the smtp transport.
4340 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4341 host errors such as "Connection refused".
4343 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4344 authentication with radiusclient 0.4.9:
4346 - Error returned from rc_read_config was caught wrongly
4347 - Username/password not passed on to radius server due to wrong length.
4349 The presumption is that some radiusclient API changes for 4.51/PH/17
4350 were not taken care of correctly. The code is still untested by me (my
4351 Linux distribution still has 0.3.2 of radiusclient), but it was
4352 contributed by a Radius user.
4354 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4355 expanding the "port" option of the smtp transport.
4357 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4358 while reading a MIME header. Thanks to Tom Hughes for a patch.
4360 PH/24 Include config.h inside local_scan.h so that configuration settings are
4363 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4364 that in an ACL for RCPT (for example), you can examine exactly what was
4367 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4368 commands, but it was not correctly comparing the address with the actual
4369 client host address. Thus, it would show the EHLO address in Received:
4370 header lines when this was not necessary.
4372 PH/27 Added the % operator to ${eval:}.
4374 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4375 it should be ignoring failures (because with -C, for example, it has lost
4376 privilege). It wasn't ignoring creation failures other than "already
4379 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4382 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4383 because an input line was too long, either on its own, or by virtue of
4384 too many continuations, the temporary file was not being removed, and the
4385 return code was incorrect.
4387 PH/31 Missing "BOOL" in function definition in filtertest.c.
4389 PH/32 Applied Sieve patches from the maintainer.
4391 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4393 PH/33 Added "verify = not_blind".
4395 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4396 Local/Makefile (with some defaults set). These are used in built scripts
4397 such as exicyclog, but they have never been used in the exim_install
4398 script (though there are many overriding facilities there). I have
4399 arranged that the exim_install script now takes note of these two
4402 PH/35 Installed configuration files for Dragonfly.
4404 PH/36 When a locally submitted message by a trusted user did not contain a
4405 From: header, and the sender address was obtained from -f or from an SMTP
4406 MAIL command, and the trusted user did not use -F to supply a sender
4407 name, $originator_name was incorrectly used when constructing a From:
4408 header. Furthermore, $originator_name was used for submission mode
4409 messages from external hosts without From: headers in a similar way,
4410 which is clearly wrong.
4412 PH/37 Added control=suppress_local_fixups.
4414 PH/38 When log_selector = +received_sender was set, and the addition of the
4415 sender made the log line's construction buffer exactly full, or one byte
4416 less than full, an overflow happened when the terminating "\n" was
4419 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4420 when the result of a list match is failure because a DNS lookup failed.
4422 PH/40 RM_COMMAND is now used in the building process.
4424 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4425 the "build-* directories that it finds.
4427 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4428 address in a domain literal was a prefix of an interface address.
4430 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4431 when verifying a sender address, unless rewrite_headers is false.
4433 PH/44 Wrote a long comment about why errors_to addresses are verified as
4434 recipients, not senders.
4436 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4437 the ratelimit ACL was added.
4439 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4441 PH/46 Added extra information about PostgreSQL errors to the error string.
4443 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4444 avoiding the need to install special zones in a real server. This is
4445 backwards compatible; if it can't find the fake resolver, it drops back.
4446 Thus, both old and new test suites can be run.
4448 TF/02 Added util/ratelimit.pl
4450 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4451 clock is set back in time.
4453 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4454 Candler <B.Candler@pobox.com>.
4456 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4457 OK for addresses that are the result of redirections.
4459 PH/48 A number of further additions for the benefit of the new test suite,
4460 including a fake gethostbyname() that interfaces to the fake DNS resolver
4463 TF/06 The fix for widen_domains has also been applied to qualify_single and
4464 search_parents which are the other dnslookup options that can cause
4467 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4470 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4472 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4473 service" if it encountered an MX record with an empty target hostname.
4474 The message is now "an MX or SRV record indicated no SMTP service".
4476 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4477 if the DNS of the sending site is misconfigured. This is quite a
4478 common situation. This change restores the behaviour of treating a
4479 helo verification defer as a failure.
4481 PH/51 If self=fail was set on a router, the bounce message did not include the
4482 actual error message.
4488 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4490 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4491 command itself ended in error, the underlying message about the transport
4492 filter timeout was being overwritten with the pipe command error. Now the
4493 underlying error message should be appended to the second error message.
4495 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4497 PH/02 Reduce the amount of output that "make" produces by default. Full output
4498 can still be requested.
4500 PH/03 The warning log line about a condition test deferring for a "warn" verb
4501 was being output only once per connection, rather than after each
4502 occurrence (because it was using the same function as for successful
4503 "warn" verbs). This seems wrong, so I have changed it.
4505 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4506 it should not have, which might have caused a crash in the right
4507 circumstances, but probably never did.
4509 PH/04 Installed a modified version of Tony Finch's patch to make submission
4510 mode fix the return path as well as the Sender: header line, and to
4511 add a /name= option so that you can make the user's friendly name appear
4514 TF/03 Added the control = fakedefer ACL modifier.
4516 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4517 Mark Lowes for thorough testing.
4519 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4521 TK/03 Merged latest SRS patch from Miles Wilton.
4523 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4524 with the definition in sysexits.h (which is #included earlier).
4525 Fortunately, Exim does not actually use EX_OK. The code used to try to
4526 preserve the sysexits.h value, by assuming that macro definitions were
4527 scanned for macro replacements. I have been disabused of this notion,
4528 so now the code just undefines EX_OK before #including unistd.h.
4530 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4531 in the smtp transport. When a block could not be written in a single
4532 write() function, the timeout was being re-applied to each part-write.
4533 This seems wrong - if the receiver was accepting one byte at a time it
4534 would take for ever. The timeout is now adjusted when this happens. It
4535 doesn't have to be particularly precise.
4537 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4538 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4540 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4541 without a domain if the check to <postmaster@domain> fails.
4543 SC/01 Eximstats: added -xls and the ability to specify output files
4544 (patch written by Frank Heydlauf).
4546 SC/02 Eximstats: use FileHandles for outputting results.
4548 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4550 SC/04 Eximstats: fixed display of large numbers with -nvr option
4552 SC/05 Eximstats: fixed merging of reports with empty tables.
4554 SC/06 Eximstats: added the -include_original_destination flag
4556 SC/07 Eximstats: removed tabs and trailing whitespace.
4558 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4560 TK/06 MBOX spool code: Add real "From " MBOX separator line
4561 so the .eml file is really in mbox format (even though
4562 most programs do not really care). Patch from Alex Miller.
4564 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4565 The latter is generated from $received_to and is only set if the
4566 message has one envelope recipient. SA can use these headers,
4567 obviously out-of-the-box. Patch from Alex Miller.
4569 PH/08 The ${def test on a variable was returning false if the variable's
4570 value was "0", contrary to what the specification has always said!
4571 The result should be true unless the variable is empty.
4573 PH/09 The syntax error of a character other than { following "${if
4574 def:variable_name" (after optional whitespace) was not being diagnosed.
4575 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4576 accidental colon was present, for example, could give incorrect results.
4578 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4579 result is used (not including appendfile, where other changes are about
4582 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4583 This involved changing a lot of size variables from int to off_t. It
4584 should work with maildirs and everything.
4586 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4587 spamd dying while we are connected to it.
4589 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4592 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4593 function for BDB 4.3.
4595 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4597 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4598 the sources for PCRE have been re-arranged and more files are now
4601 PH/15 The code I had for printing potentially long long variables in PH/11
4602 above was not the best (it lost precision). The length of off_t variables
4603 is now inspected at build time, and an appropriate printing format (%ld
4604 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4605 to be "long long int" or "long int". This is needed for the internal
4606 formatting function string_vformat().
4608 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4609 the configuration file to be ":syslog", then the script "guesses" where
4610 the logs files are, rather than using the compiled in default. In our
4611 case the guess is not the same as the compiled default, so the script
4612 suddenly stopped working when I started to use syslog. The patch checks
4613 to see if log_file_path is "". If so, it attempts to read it from exim
4614 with no configuration file to get the compiled in version, before it
4615 falls back to the previous guessing code."
4617 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4618 implementing BATV in an Exim configuration. See NewStuff for the gory
4621 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4622 Makefile that are specific to HP-UX.
4624 PH/18 If the "use_postmaster" option was set for a recipient callout together
4625 with the "random" option, the postmaster address was used as the MAIL
4626 FROM address for the random test, but not for the subsequent recipient
4627 test. It is now used for both.
4629 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4630 patch removes a few documentation additions to RFC 3028, because the
4631 latest draft now contains them. It adds the new en;ascii-case comparator
4632 and a new error check for 8bit text in MIME parts. Comparator and
4633 require names are now matched exactly. I enabled the subaddress
4634 extension, but it is not well tested yet (read: it works for me)."
4636 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4637 rework some of the code of TK/09 above to avoid the hardwired use of
4638 "%lld" and "long long". Replaced the call to snprintf() with a call to
4641 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4642 records point to non-existent hosts", "retry timeout exceeded", and
4643 "retry time not reached for any host after a long failure period".
4645 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4646 experimental DomainKeys support:
4648 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4649 (2) On an error such as an illegally used "control", the wrong name for
4650 the control was given.
4652 These problems did NOT occur unless DomainKeys support was compiled.
4654 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4656 PH/24 Added ${if match_ip condition.
4658 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4659 databases so that it will be absolutely obvious if a crash occurs in the
4660 DB library. This is a regular occurrence (often caused by mis-matched
4663 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4664 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4665 picked up on a user's system that detects such things. There doesn't seem
4666 to be a gcc warning option for this - only an attribute that has to be
4667 put on the function's prototype. It seems that in Fedora Core 4 they have
4668 set this on a number of new functions. No doubt there will be more in due
4671 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4672 specify a transport. However, if an address that was verified by such a
4673 router was the subject of a callout, Exim crashed because it tried to
4674 read the rcpt_include_affixes from the non-existent transport. Now it
4675 just assumes that the setting of that option is false. This bug was
4676 introduced by 4.51/PH/31.
4678 PH/28 Changed -d+all to exclude +memory, because that information is very
4679 rarely of interest, but it makes the output a lot bigger. People tend to
4680 do -d+all out of habit.
4682 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4683 code in os-type was giving problems when libc.so lives in lib64, like on
4686 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4687 aren't the modern standard, and it seems that some systems' include files
4688 don't always have them. Exim was already checking for some of the newer
4689 ones like T_AAAA, and defining it itself. I've added checks for all the
4690 record types that Exim uses.
4692 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4693 not automatically generating a new one, as it is supposed to. This
4694 prevented TLS from working. If the file did exist, but contained invalid
4695 data, a new version was generated, as expected. It was only the case of a
4696 non-existent file that was broken.
4698 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4699 with a change in libdomainkeys > 0.64.
4701 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4702 from DNS. If the selector record carries the flag, it now has
4703 precedence over the domain-wide flag.
4705 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4707 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4708 the use of a port name that isn't defined in /etc/services) occurred, the
4709 message was deferred as in a normal delivery, and thus remained on the
4710 spool, instead of being failed because of the mua_wrapper setting. This
4711 is now fixed, and I tidied up some of the mua_wrapper messages at the
4714 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4715 the messages in a hash of arrays rather than using individual hashes.
4716 This is a bit cleaner and results in dramatic memory savings, albeit
4717 at a slight CPU cost.
4719 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4720 as requested by Marc Sherman.
4722 SC/10 Eximstats: added histograms for user specified patterns as requested
4725 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4727 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4728 fopen() in the content-scanning modules that did not already have it.
4734 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4735 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4737 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4739 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4741 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4742 to test Sieve filters that use "vacation".
4744 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4745 that changes the way the GnuTLS parameters are stored in the cache file.
4746 The new format can be generated externally. For backward compatibility,
4747 if the data in the cache doesn't make sense, Exim assumes it has read an
4748 old-format file, and it generates new data and writes a new file. This
4749 means that you can't go back to an older release without removing the
4752 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4753 work if there are any delivery delays because "one_time" forces the
4754 parent to be marked "delivered", so its unseen clone is never tried
4755 again. For this reason, Exim now forbids the simultaneous setting of
4758 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4759 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4760 entry if you want to know the details. The fix, however, neglected to
4761 consider the case when local delivery batching is involved. The test for
4762 "previously delivered" was not happening when checking to see if an
4763 address could be batched with a previous (undelivered) one; under
4764 certain circumstances this could lead to multiple deliveries to the same
4767 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4768 in its include files, and this causes problems building Exim.
4770 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4771 header_syntax) but Exim was just ignoring anything given after a slash.
4772 In particular, this caused confusion with an attempt to use "verify =
4773 reverse_host_lookup/defer_ok". An error is now given when options are
4774 supplied for verify items that do not have them. (Maybe reverse_host_
4775 lookup should have a defer_ok option, but that's a different point.)
4777 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4778 defined by RFC 821) to 2048, because there were problems with some AUTH
4779 commands, and RFC 1869 says the size should be increased for extended
4780 SMTP commands that take arguments.
4782 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4785 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4786 "unknown" error; now it says that the functionality isn't in the binary.
4788 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4789 an address' error message when a string expansion fails (syntax or
4790 whatever). Otherwise the password may appear in the log. Following change
4791 PH/42 below, there is no longer a chance of it appearing in a bounce
4794 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4796 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4797 of its addresses was ever tried. (Bugzilla bug #2.)
4799 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4800 the result incorrectly in the debug output. (It correctly added a newline
4801 to what was transported.)
4803 TF/01 Added $received_time.
4805 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4806 commented out examples of how to interface to a virus scanner and to
4807 SpamAssassin. Also added commented examples of av_scanner and
4808 spamd_address settings.
4810 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4811 and controls are allowed in which ACLs. There were a couple of minor
4812 errors. Some of the entries in the conditions table (which is a table of
4813 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4814 negation of where the condition IS allowed.
4816 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4818 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4819 header file does not have a version number, so I've had to invent a new
4820 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4821 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4822 radiusclient), but it was contributed by a Radius user.
4824 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4825 files or directories, for OpenSSL.
4827 PH/19 When an Exim process that is running as root has to create an Exim log
4828 file, it does so in a subprocess that runs as exim:exim so as to get the
4829 ownership right at creation (otherwise, other Exim processes might see
4830 the file with the wrong ownership). There was no test for failure of this
4831 fork() call, which would lead to the process getting stuck as it waited
4832 for a non-existent subprocess. Forks do occasionally fail when resources
4833 run out. I reviewed all the other calls to fork(); they all seem to check
4836 PH/20 When checking for unexpected SMTP input at connect time (before writing
4837 the banner), Exim was not dealing correctly with a non-positive return
4838 from the read() function. If the client had disconnected by this time,
4839 the result was a log entry for a synchronization error with an empty
4840 string after "input=" when read() returned zero. If read() returned -1
4841 (an event I could not check), uninitialized data bytes were printed.
4842 There were reports of junk text (parts of files, etc) appearing after
4845 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4847 PH/22 Added support for macro redefinition, and (re)definition in between
4848 driver and ACL definitions.
4850 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4851 forgetting to use the resulting value; it was using the unexpanded value.
4853 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4854 hadn't been configured. The fix is from Juergen Kreileder, who
4855 understands it better than I do:
4857 "Here's what I see happening with three configured cyrus_sasl
4858 authenticators configured (plain, login, cram-md5):
4860 On startup auth_cyrus_sasl_init() gets called for each of these.
4861 This means three calls to sasl_listmech() without a specified mech_list.
4862 => SASL tests which mechs of all available mechs actually work
4863 => three warnings about OTP not working
4864 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4866 With the patch, sasl_listmech() also gets called three times. But now
4867 SASL's mech_list option is set to the server_mech specified in the the
4868 authenticator. Or in other words, the answer from sasl_listmech()
4869 gets limited to just the mech you're testing for (which is different
4871 => the return list contains just 'plain' or 'login', 'cram-md5' or
4872 nothing depending on the value of ob->server_mech.
4874 I've just tested the patch: Authentication still works fine,
4875 unavailable mechs specified in the exim configuration are still
4876 caught, and the auth.log warnings about OTP are gone."
4878 PH/25 When debugging is enabled, the contents of the command line are added
4879 to the debugging output, even when log_selector=+arguments is not
4882 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4883 answer is "GNU", and only if the return is "GNU/something" is the answer
4886 PH/27 $acl_verify_message is now set immediately after the failure of a
4887 verification in an ACL, and so is available in subsequent modifiers. In
4888 particular, the message can be preserved by coding like this:
4890 warn !verify = sender
4891 set acl_m0 = $acl_verify_message
4893 Previously, $acl_verify_message was set only while expanding "message"
4894 and "log_message" when a very denied access.
4896 PH/28 Modified OS/os.c-Linux with
4898 -#ifndef OS_LOAD_AVERAGE
4899 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4901 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4902 nomenclature these days.)
4904 PH/29 Installed patch from the Sieve maintainer that adds the options
4905 sieve_useraddress and sieve_subaddress to the redirect router.
4907 PH/30 In these circumstances:
4908 . Two addresses routed to the same list of hosts;
4909 . First host does not offer TLS;
4910 . First host accepts first address;
4911 . First host gives temporary error to second address;
4912 . Second host offers TLS and a TLS session is established;
4913 . Second host accepts second address.
4914 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4915 and peerdn, if requested) that were in fact used only for the second
4918 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4919 attention to any local part prefix or suffix that was matched by the
4920 router that accepted the address. It now behaves in the same way as it
4921 does for delivery: the affixes are removed from the local part unless
4922 rcpt_include_affixes is set on the transport.
4924 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4925 timeout during the DATA phase of an incoming message.
4927 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4928 applied a patch sanctioned by the Sieve maintainer.
4930 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4931 the uid or gid is negative. A case of a negative gid caused this to be
4932 noticed. The fix allows for either to be negative.
4934 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4935 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4938 PH/36 The change PH/12 above was broken. Fixed it.
4940 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4941 the grounds that routing the same address twice would always produce the
4942 same answer. This might have been true once, but it is certainly no
4943 longer true now. Routing a child address may depend on the previous
4944 routing that produced that child. Some complicated redirection strategies
4945 went wrong when messages had multiple recipients, and made Exim's
4946 behaviour dependent on the order in which the addresses were given.
4948 I have moved the duplicate checking until after the routing is complete.
4949 Exim scans the addresses that are assigned to local and remote
4950 transports, and removes any duplicates. This means that more work will be
4951 done, as duplicates will always all be routed, but duplicates are
4952 presumably rare, so I don't expect this is of any significance.
4954 For deliveries to pipes, files, and autoreplies, the duplicate checking
4955 still happens during the routing process, since they are not going to be
4958 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4959 It corrects a timeout issue with spamd. This is Ian's comment: "The
4960 background is that sometimes spamd either never reads data from a
4961 connection it has accepted, or it never writes response data. The exiscan
4962 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4963 blindly assumes that writes won't block so it may never time out."
4965 PH/39 Allow G after quota size as well as K and M.
4967 PH/40 The value set for $authenticated_id in an authenticator may not contain
4968 binary zeroes or newlines because the value is written to log lines and
4969 to spool files. There was no check on this. Now the value is run through
4970 the string_printing() function so that such characters are converted to
4971 printable escape sequences.
4973 PH/41 $message_linecount is a new variable that contains the total number of
4974 lines in the message. Compare $body_linecount, which is the count for the
4977 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4978 bounce and delay warning messages, except in certain special cases, which
4981 (a) An SMTP error message from a remote host;
4982 (b) A message specified in a :fail: redirection;
4983 (c) A message specified in a "fail" command in a system filter;
4984 (d) A message specified in a FAIL return from the queryprogram router;
4985 (e) A message specified by the cannot_route_message router option.
4987 In these cases only, Exim does include the error details in bounce and
4988 warning messages. There are also a few cases where bland messages such
4989 as "unrouteable address" or "local delivery error" are given.
4991 PH/43 $value is now also set for the "else" part of a ${run expansion.
4993 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4994 being worked on, but at least Exim now implements the latest version to
4997 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4998 process to complete was treated as a delivery failure, a timeout while
4999 writing the message to the pipe was logged, but erroneously treated as a
5000 successful delivery. Such timeouts include transport filter timeouts. For
5001 consistency with the overall process timeout, these timeouts are now
5002 treated as errors, giving rise to delivery failures by default. However,
5003 there is now a new Boolean option for the pipe transport called
5004 timeout_defer, which, if set TRUE, converts the failures into defers for
5005 both kinds of timeout. A transport filter timeout is now identified in
5008 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5009 systems where "make" and "gmake" are different, calling "gmake" at top
5010 level broke things. I've arranged for the value of $(MAKE) to be passed
5011 from the Makefile to this script so that it can call the same version of
5015 A note about Exim versions 4.44 and 4.50
5016 ----------------------------------------
5018 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5019 changes of various kinds. As a consequence, a big documentation update was
5020 needed. This delayed the release for rather longer than seemed good, especially
5021 in the light of a couple of (minor) security issues. Therefore, the changes
5022 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5023 release. So 4.44 and 4.50 are in effect two different branches that both start
5026 I have left the 4.50 change log unchanged; it contains all the changes since
5027 4.43. The change log for 4.44 is below; many of its items are identical to
5028 those for 4.50. This seems to be the most sensible way to preserve the
5029 historical information.
5035 1. Minor wording change to the doc/README.SIEVE file.
5037 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5038 computation of the current number of files was incorrect.
5040 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5041 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5044 4. Give more explanation in the error message when the command for a transport
5045 filter fails to execute.
5047 5. There are several places where Exim runs a non-Exim command in a
5048 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5049 was being done only for the command run by the queryprogram router. It is
5050 now done for all such subprocesses. The other cases are: ${run, transport
5051 filters, and the commands run by the lmtp and pipe transports.
5053 6. Added CONFIGURE_GROUP build-time option.
5055 7. Some older OS have a limit of 256 on the maximum number of file
5056 descriptors. Exim was using setrlimit() to set 1000 as a large value
5057 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5058 systems. I've change it so that if it can't get 1000, it tries for 256.
5060 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5061 was an oversight, and furthermore, ever since the addition of extra
5062 controls (e.g. 4.43/32), the checks on when to allow different forms of
5063 "control" were broken. There should now be diagnostics for all cases when a
5064 control that does not make sense is encountered.
5066 9. Added the /retain_sender option to "control=submission".
5068 10. $recipients is now available in the predata ACL (oversight).
5070 11. Tidy the search cache before the fork to do a delivery from a message
5071 received from the command line. Otherwise the child will trigger a lookup
5072 failure and thereby defer the delivery if it tries to use (for example) a
5073 cached ldap connection that the parent has called unbind on.
5075 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5076 of $address_data from the recipient verification was clobbered by the
5077 sender verification.
5079 13. The value of address_data from a sender verification is now available in
5080 $sender_address_data in subsequent conditions in the ACL statement.
5082 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5084 15. Added a new option "connect=<time>" to callout options, to set a different
5087 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5088 was its contents. (It was OK if the option was not defined at all.)
5090 17. A "Completed" log line is now written for messages that are removed from
5091 the spool by the -Mrm option.
5093 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5094 information about exactly what failed.
5096 19. Added -dd to debug only the daemon process.
5098 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5099 handles timeouts, both on the server side and network timeouts. Renamed the
5100 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5102 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5103 It is now set to "smtps".
5105 22. $host_address is now set to the target address during the checking of
5106 ignore_target_hosts.
5108 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5109 being passed; this would have caused $sender_host_name to have been used if
5110 matching the list had actually called for a host name (not very likely,
5111 since this list is usually IP addresses). A host name is now passed as
5114 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5115 code that specifies a non-restarting handler (typically sigaction() in
5116 modern systems) in an attempt to fix a rare and obscure crash bug.
5118 25. Narrowed the window for a race in the daemon that could cause it to ignore
5119 SIGCHLD signals. This is not a major problem, because they are used only to
5120 wake it up if nothing else does.
5122 26. A malformed maildirsize file could cause Exim to calculate negative values
5123 for the mailbox size or file count. Odd effects could occur as a result.
5124 The maildirsize information is now recalculated if the size or filecount
5127 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5128 support for a long time. Removed HAVE_SYS_VFS_H.
5130 28. Installed the latest version of exipick from John Jetmore.
5132 29. In an address list, if the pattern was not a regular expression, an empty
5133 subject address (from a bounce message) matched only if the pattern was an
5134 empty string. Non-empty patterns were not even tested. This was the wrong
5135 because it is perfectly reasonable to use an empty address as part of a
5136 database query. An empty address is now tested by patterns that are
5137 lookups. However, all the other forms of pattern expect the subject to
5138 contain a local part and a domain, and therefore, for them, an empty
5139 address still always fails if the pattern is not itself empty.
5141 30. Exim went into a mad DNS loop when attempting to do a callout where the
5142 host was specified on an smtp transport, and looking it up yielded more
5143 than one IP address.
5145 31. Re-factored the code for checking spool and log partition space into a
5146 function that finds that data and another that does the check. The former
5147 is then used to implement four new variables: $spool_space, $log_space,
5148 $spool_inodes, and $log_inodes.
5150 32. The RFC2047 encoding function was originally intended for short strings
5151 such as real names; it was not keeping to the 75-character limit for
5152 encoded words that the RFC imposes. It now respects the limit, and
5153 generates multiple encoded words if necessary. To be on the safe side, I
5154 have increased the buffer size for the ${rfc2047: expansion operator from
5157 33. It is now permitted to omit both strings after an "if" condition; if the
5158 condition is true, the result is "true". As before, when the second string
5159 is omitted, a false condition yields an empty string. This makes it less
5160 cumbersome to write custom ACL and router conditions.
5162 34. Failure to deliver a bounce message always caused it to be frozen, even if
5163 there was an errors_to setting on the router. The errors_to setting is now
5166 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5167 canonical form (fully expanded) before being placed in
5168 $sender_host_address.
5170 36. The table in the code that translates DNS record types into text (T_A to
5171 "A" for instance) was missing entries for NS and CNAME. It is just possible
5172 that this could have caused confusion if both these types were looked up
5173 for the same domain, because the text type is used as part of Exim's
5174 per-process caching. But the chance of anyone hitting this buglet seems
5177 37. The dnsdb lookup has been extended in a number of ways.
5179 (1) There is a new type, "zns", which walks up the domain tree until it
5180 finds some nameserver records. It should be used with care.
5182 (2) There is a new type, "mxh", which is like "mx" except that it returns
5183 just the host names, not the priorities.
5185 (3) It is now possible to give a list of domains (or IP addresses) to be
5186 looked up. The behaviour when one of the lookups defers can be
5187 controlled by a keyword.
5189 (4) It is now possible to specify the separator character for use when
5190 multiple records are returned.
5192 38. The dnslists ACL condition has been extended: it is now possible to supply
5193 a list of IP addresses and/or domains to be looked up in a particular DNS
5196 39. Added log_selector=+queue_time_overall.
5198 40. When running the queue in the test harness, wait just a tad after forking a
5199 delivery process, to get repeatability of debugging output.
5201 41. Include certificate and key file names in error message when GnuTLS fails
5202 to set them up, because the GnuTLS error message doesn't include the name
5203 of the failing file when there is a problem reading it.
5205 42. Allow both -bf and -bF in the same test run.
5207 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5209 44. The "Exiscan patch" is now merged into the mainline Exim source.
5211 45. Sometimes the final signoff response after QUIT could fail to get
5212 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5213 < 0 before doing a fflush(). This bug looks as though it goes back to the
5214 introduction of TLS in release 3.20, but "sometimes" must have been rare
5215 because the tests only now provoked it.
5217 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5218 (this can affect the format of dates).
5220 47. exim_tidydb, when checking for the continued existence of a message for
5221 which it has found a message-specific retry record, was not finding
5222 messages that were in split spool directories. Consequently, it was
5223 deleting retry records that should have stayed in existence.
5225 48. Steve fixed some bugs in eximstats.
5227 49. The SPA authentication driver was not abandoning authentication and moving
5228 on to the next authenticator when an expansion was forced to fail,
5229 contradicting the general specification for all authenticators. Instead it
5230 was generating a temporary error. It now behaves as specified.
5232 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5233 (the order specifies the preference for clients). The order is now AES256,
5234 AES128, 3DES, ARCFOUR128.
5236 51. Small patch to Sieve code - explicitly set From: when generating an
5239 52. Exim crashed if a remote delivery caused a very long error message to be
5240 recorded - for instance if somebody sent an entire SpamAssassin report back
5241 as a large number of 550 error lines. This bug was coincidentally fixed by
5242 increasing the size of one of Exim's internal buffers (big_buffer) that
5243 happened as part of the Exiscan merge. However, to be on the safe side, I
5244 have made the code more robust (and fixed the comments that describe what
5247 53. Now that there can be additional text after "Completed" in log lines (if
5248 the queue_time_overall log selector is set), a one-byte patch to exigrep
5249 was needed to allow it to recognize "Completed" as not the last thing in
5252 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5253 patch that reportedly fixes this has been added. I am not expert enough to
5254 create a test for it. This is what the patch creator wrote:
5256 "I found a little strange behaviour of ldap code when working with
5257 Windows 2003 AD Domain, where users was placed in more than one
5258 Organization Units. When I tried to give exim partial DN, the exit code
5259 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5260 But simultaneously result of request was absolutely normal ldap result,
5261 so I produce this patch..."
5263 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5264 REFERENCE, so I have modified the code to exclude the patch when that macro
5267 55. Some experimental protocols are using DNS PTR records for new purposes. The
5268 keys for these records are domain names, not reversed IP addresses. The
5269 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5270 leaves it alone. Component reversal etc. now happens only for IP addresses.
5273 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5275 57. Double the size of the debug message buffer (to 2048) so that more of very
5276 long debug lines gets shown.
5278 58. The exicyclog utility now does better if the number of log files to keep
5279 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5281 59. Two changes related to the smtp_active_hostname option:
5283 (1) $smtp_active_hostname is now available as a variable.
5284 (2) The default for smtp_banner uses $smtp_active_hostname instead
5285 of $primary_hostname.
5287 60. The host_aton() function is supposed to be passed a string that is known
5288 to be a valid IP address. However, in the case of IPv6 addresses, it was
5289 not checking this. This is a hostage to fortune. Exim now panics and dies
5290 if the condition is not met. A case was found where this could be provoked
5291 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5292 components; fortuitously, this particular loophole had already been fixed
5293 by change 4.50/55 above.
5295 If there are any other similar loopholes, the new check in host_aton()
5296 itself should stop them being exploited. The report I received stated that
5297 data on the command line could provoke the exploit when Exim was running as
5298 exim, but did not say which command line option was involved. All I could
5299 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5300 running as the user.
5303 61. There was a buffer overflow vulnerability in the SPA authentication code
5304 (which came originally from the Samba project). I have added a test to the
5305 spa_base64_to_bits() function which I hope fixes it.
5308 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5309 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5311 63. The daemon start-up calls getloadavg() while still root for those OS that
5312 need the first call to be done as root, but it missed one case: when
5313 deliver_queue_load_max is set with deliver_drop_privilege. This is
5314 necessary for the benefit of the queue runner, because there is no re-exec
5315 when deliver_drop_privilege is set.
5317 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5318 This has been fixed.
5320 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5321 was in use, was not putting the data itself into the right store pool;
5322 consequently, it could be overwritten for a subsequent message in the same
5323 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5326 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5328 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5329 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5330 the function were treating the return as a boolean value, which happened to
5331 work because 0=false and not-0=true, but is not correct code.
5333 68. The host_aton() function was not handling scoped IPv6 addresses (those
5334 with, for example, "%eth0" on the end) correctly.
5336 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5337 negated items (that is, ~something) in unsigned ints. Some compilers
5338 apparently mutter when there is no cast.
5340 70. If an address verification called from an ACL failed, and did not produce a
5341 user-specific message (i.e. there was only a "system" message), nothing was
5342 put in $acl_verify_message. In this situation, it now puts the system
5345 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5346 session; change 4.31/43 added the unwanted input to the log line - except
5347 that it did not do this in the start of session case. It now does.
5349 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5350 This is wrong and can cause the other end to generate a synchronization
5351 error if it is another Exim or anything else that does the synchronization
5352 check. A QUIT command is no longer sent after a timeout.
5354 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5355 during host lookups.
5357 74. The defer_ok option of callout verification was not working if it was used
5358 when verifying addresses in header lines, that is, for this case:
5360 verify = header_sender/callout=defer_ok
5362 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5363 those file descriptors could be used for SMTP connections. If anything
5364 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5365 could be sent to the SMTP client, causing chaos. The daemon now opens
5366 stdin, stdout, and stderr to /dev/null when it puts itself into the
5369 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5370 log by default. The user can override this with suitable Perl magic.
5372 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5373 the log message when discard triggers, was not working for the DATA ACL or
5374 for the non-SMTP ACL.
5376 78. Error message wording change in sieve.c.
5378 79. If smtp_accept_max_per_host was set, the number of connections could be
5379 restricted to fewer than expected, because the daemon was trying to set up
5380 a new connection before checking whether the processes handling previous
5381 connections had finished. The check for completed processes is now done
5382 earlier. On busy systems, this bug wouldn't be noticed because something
5383 else would have woken the daemon, and it would have reaped the completed
5386 80. If a message was submitted locally by a user whose login name contained one
5387 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5388 It caused a spool format error. I have fixed the spool reading code. A
5389 related problem was that the "from" clause in the Received: line became
5390 illegal because of the space(s). It is now covered by ${quote_local_part.
5392 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5395 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5397 83. Patch from Sieve maintainer to fix unterminated string problem in
5398 "vacation" handling.
5400 84. Some minor changes to the Linux configuration files to help with other
5401 OS variants using glibc.
5403 85. One more patch for Sieve to update vacation handling to latest spec.
5406 ----------------------------------------------------
5407 See the note above about the 4.44 and 4.50 releases.
5408 ----------------------------------------------------
5414 1. Change 4.43/35 introduced a bug that caused file counts to be
5415 incorrectly computed when quota_filecount was set in an appendfile
5418 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5419 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5422 3. Give more explanation in the error message when the command for a transport
5423 filter fails to execute.
5425 4. There are several places where Exim runs a non-Exim command in a
5426 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5427 was being done only for the command run by the queryprogram router. It is
5428 now done for all such subprocesses. The other cases are: ${run, transport
5429 filters, and the commands run by the lmtp and pipe transports.
5431 5. Some older OS have a limit of 256 on the maximum number of file
5432 descriptors. Exim was using setrlimit() to set 1000 as a large value
5433 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5434 systems. I've change it so that if it can't get 1000, it tries for 256.
5436 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5437 was an oversight, and furthermore, ever since the addition of extra
5438 controls (e.g. 4.43/32), the checks on when to allow different forms of
5439 "control" were broken. There should now be diagnostics for all cases when a
5440 control that does not make sense is encountered.
5442 7. $recipients is now available in the predata ACL (oversight).
5444 8. Tidy the search cache before the fork to do a delivery from a message
5445 received from the command line. Otherwise the child will trigger a lookup
5446 failure and thereby defer the delivery if it tries to use (for example) a
5447 cached ldap connection that the parent has called unbind on.
5449 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5450 of $address_data from the recipient verification was clobbered by the
5451 sender verification.
5453 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5454 was its contents. (It was OK if the option was not defined at all.)
5456 11. A "Completed" log line is now written for messages that are removed from
5457 the spool by the -Mrm option.
5459 12. $host_address is now set to the target address during the checking of
5460 ignore_target_hosts.
5462 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5463 being passed; this would have caused $sender_host_name to have been used if
5464 matching the list had actually called for a host name (not very likely,
5465 since this list is usually IP addresses). A host name is now passed as
5468 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5469 code that specifies a non-restarting handler (typically sigaction() in
5470 modern systems) in an attempt to fix a rare and obscure crash bug.
5472 15. Narrowed the window for a race in the daemon that could cause it to ignore
5473 SIGCHLD signals. This is not a major problem, because they are used only to
5474 wake it up if nothing else does.
5476 16. A malformed maildirsize file could cause Exim to calculate negative values
5477 for the mailbox size or file count. Odd effects could occur as a result.
5478 The maildirsize information is now recalculated if the size or filecount
5481 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5482 support for a long time. Removed HAVE_SYS_VFS_H.
5484 18. Updated exipick to current release from John Jetmore.
5486 19. Allow an empty sender to be matched against a lookup in an address list.
5487 Previously the only cases considered were a regular expression, or an
5490 20. Exim went into a mad DNS lookup loop when doing a callout where the
5491 host was specified on the transport, if the DNS lookup yielded more than
5494 21. The RFC2047 encoding function was originally intended for short strings
5495 such as real names; it was not keeping to the 75-character limit for
5496 encoded words that the RFC imposes. It now respects the limit, and
5497 generates multiple encoded words if necessary. To be on the safe side, I
5498 have increased the buffer size for the ${rfc2047: expansion operator from
5501 22. Failure to deliver a bounce message always caused it to be frozen, even if
5502 there was an errors_to setting on the router. The errors_to setting is now
5505 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5506 canonical form (fully expanded) before being placed in
5507 $sender_host_address.
5509 24. Updated eximstats to version 1.33
5511 25. Include certificate and key file names in error message when GnuTLS fails
5512 to set them up, because the GnuTLS error message doesn't include the name
5513 of the failing file when there is a problem reading it.
5515 26. Expand error message when OpenSSL has problems setting up cert/key files.
5518 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5519 (this can affect the format of dates).
5521 28. exim_tidydb, when checking for the continued existence of a message for
5522 which it has found a message-specific retry record, was not finding
5523 messages that were in split spool directories. Consequently, it was
5524 deleting retry records that should have stayed in existence.
5526 29. eximstats updated to version 1.35
5527 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5528 1.35 - bugfix such that pie charts by volume are generated correctly
5530 30. The SPA authentication driver was not abandoning authentication and moving
5531 on to the next authenticator when an expansion was forced to fail,
5532 contradicting the general specification for all authenticators. Instead it
5533 was generating a temporary error. It now behaves as specified.
5535 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5536 (the order specifies the preference for clients). The order is now AES256,
5537 AES128, 3DES, ARCFOUR128.
5539 31. Small patch to Sieve code - explicitly set From: when generating an
5542 32. Exim crashed if a remote delivery caused a very long error message to be
5543 recorded - for instance if somebody sent an entire SpamAssassin report back
5544 as a large number of 550 error lines. This bug was coincidentally fixed by
5545 increasing the size of one of Exim's internal buffers (big_buffer) that
5546 happened as part of the Exiscan merge. However, to be on the safe side, I
5547 have made the code more robust (and fixed the comments that describe what
5550 33. Some experimental protocols are using DNS PTR records for new purposes. The
5551 keys for these records are domain names, not reversed IP addresses. The
5552 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5553 leaves it alone. Component reversal etc. now happens only for IP addresses.
5556 34. The host_aton() function is supposed to be passed a string that is known
5557 to be a valid IP address. However, in the case of IPv6 addresses, it was
5558 not checking this. This is a hostage to fortune. Exim now panics and dies
5559 if the condition is not met. A case was found where this could be provoked
5560 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5561 components; fortuitously, this particular loophole had already been fixed
5562 by change 4.50/55 or 4.44/33 above.
5564 If there are any other similar loopholes, the new check in host_aton()
5565 itself should stop them being exploited. The report I received stated that
5566 data on the command line could provoke the exploit when Exim was running as
5567 exim, but did not say which command line option was involved. All I could
5568 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5569 running as the user.
5572 35. There was a buffer overflow vulnerability in the SPA authentication code
5573 (which came originally from the Samba project). I have added a test to the
5574 spa_base64_to_bits() function which I hope fixes it.
5577 36. The daemon start-up calls getloadavg() while still root for those OS that
5578 need the first call to be done as root, but it missed one case: when
5579 deliver_queue_load_max is set with deliver_drop_privilege. This is
5580 necessary for the benefit of the queue runner, because there is no re-exec
5581 when deliver_drop_privilege is set.
5583 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5584 was in use, was not putting the data itself into the right store pool;
5585 consequently, it could be overwritten for a subsequent message in the same
5586 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5589 38. Sometimes the final signoff response after QUIT could fail to get
5590 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5591 < 0 before doing a fflush(). This bug looks as though it goes back to the
5592 introduction of TLS in release 3.20, but "sometimes" must have been rare
5593 because the tests only now provoked it.
5599 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5600 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5601 Now it returns an integer. A number of calls were still expecting a T/F
5602 return. Fortuitously, in all cases, the tests worked in OK situations,
5603 which is the norm. However, things would have gone wrong on any write
5604 failures on the smtp file descriptor. This function is used when sending
5605 messages over SMTP and also when doing verify callouts.
5607 2. When Exim is called to do synchronous delivery of a locally submitted
5608 message (the -odf or -odi options), it no longer closes stderr before doing
5611 3. Implemented the mua_wrapper option.
5613 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5615 5. Implemented the functions header_remove(), header_testname(),
5616 header_add_at_position(), and receive_remove_recipient(), and exported them
5619 6. If an ACL "warn" statement specified the addition of headers, Exim already
5620 inserted X-ACL-Warn: at the start if there was no header name. However, it
5621 was not making this test for the second and subsequent header lines if
5622 there were newlines in the string. This meant that an invalid header could
5623 be inserted if Exim was badly configured.
5625 7. Allow an ACL "warn" statement to add header lines at the start or after all
5626 the Received: headers, as well as at the end.
5628 8. Added the rcpt_4xx retry error code.
5630 9. Added postmaster_mailfrom=xxx to callout verification option.
5632 10. Added mailfrom=xxxx to the callout verification option, for verify=
5635 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5636 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5638 12. Inserted some casts to stop certain compilers warning when using pointer
5639 differences as field lengths or precisions in printf-type calls (mostly
5640 affecting debugging statements).
5642 13. Added optional readline() support for -be (dynamically loaded).
5644 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5645 same clock tick as a message's arrival, so that its received time was the
5646 same as the "first fail" time on the retry record, and that message
5647 remained on the queue past the ultimate address timeout, every queue runner
5648 would try a delivery (because it was past the ultimate address timeout) but
5649 after another failure, the ultimate address timeout, which should have then
5650 bounced the address, did not kick in. This was a "< instead of <=" error;
5651 in most cases the first failure would have been in the next clock tick
5652 after the received time, and all would be well.
5654 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5655 being recognized when the domain list was tested by the match_domain
5656 condition in an expansion string.
5658 16. Added the ${str2b64: operator.
5660 17. Exim was always calling setrlimit() to set a large limit for the number of
5661 processes, without checking whether the existing limit was already
5662 adequate. (It did check for the limit on file descriptors.) Furthermore,
5663 errors from getrlimit() and setrlimit() were being ignored. Now they are
5664 logged to the main and panic logs, but Exim does carry on, to try to do its
5665 job under whatever limits there are.
5667 18. Imported PCRE 5.0.
5669 19. Trivial typo in log message " temporarily refused connection" (the leading
5672 20. If the log selector return_path_on_delivery was set and an address was
5673 redirected to /dev/null, the delivery process crashed because it assumed
5674 that a return path would always be set for a "successful" delivery. In this
5675 case, the whole delivery is bypassed as an optimization, and therefore no
5678 21. Internal re-arrangement: the function for sending a challenge and reading
5679 a response while authentication was assuming a zero-terminated challenge
5680 string. It's now changed to take a pointer and a length, to allow for
5681 binary data in such strings.
5683 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5685 23. Exim was not respecting finduser_retries when seeking the login of the
5686 uid under which it was called; it was always trying 10 times. (The default
5687 setting of finduser_retries is zero.) Also, it was sleeping after the final
5688 failure, which is pointless.
5690 24. Implemented tls_on_connect_ports.
5692 25. Implemented acl_smtp_predata.
5694 26. If the domain in control=submission is set empty, Exim assumes that the
5695 authenticated id is a complete email address when it generates From: or
5696 Sender: header lines.
5698 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5699 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5700 chown and chgrp in /bin and hostname in /usr/bin.
5702 28. Exim was keeping the "process log" file open after each use, just as it
5703 does for the main log. This opens the possibility of it remaining open for
5704 long periods when the USR1 signal hits a daemon. Occasional processlog
5705 errors were reported, that could have been caused by this. Anyway, it seems
5706 much more sensible not to leave this file open at all, so that is what now
5709 29. The long-running daemon process does not normally write to the log once it
5710 has entered its main loop, and it closes the log before doing so. This is
5711 so that log files can straightforwardly be renamed and moved. However,
5712 there are a couple of unusual error situations where the daemon does write
5713 log entries, and I had neglected to close the log afterwards.
5715 30. The text of an SMTP error response that was received during a remote
5716 delivery was being truncated at 512 bytes. This is too short for some of
5717 the long messages that one sometimes sees. I've increased the limit to
5720 31. It is now possible to make retry rules that apply only when a message has a
5721 specific sender, in particular, an empty sender.
5723 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5724 it possible to be selective about when SMTP synchronization is enforced.
5726 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5728 32. Implemented hosts_connection_nolog.
5730 33. Added an ACL for QUIT.
5732 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5735 35. Added mailbox_size and mailbox_filecount to appendfile.
5737 36. Added control = no_multiline_responses to ACLs.
5739 37. There was a bug in the logic of the code that waits for the clock to tick
5740 in the case where the clock went backwards by a substantial amount such
5741 that the microsecond fraction of "now" was more than the microsecond
5742 fraction of "then" (but the whole seconds number was less).
5744 38. Added support for the libradius Radius client library this is found on
5745 FreeBSD (previously only the radiusclient library was supported).
5751 1. When certain lookups returned multiple values in the form name=value, the
5752 quoting of the values was not always being done properly. Specifically:
5753 (a) If the value started with a double quote, but contained no whitespace,
5755 (b) If the value contained whitespace other than a space character (i.e.
5756 tabs or newlines or carriage returns) it was not quoted.
5757 This fix has been applied to the mysql and pgsql lookups by writing a
5758 separate quoting function and calling it from the lookup code. The fix
5759 should probably also be applied to nisplus, ibase and oracle lookups, but
5760 since I cannot test any of those, I have not disturbed their existing code.
5762 2. A hit in the callout cache for a specific address caused a log line with no
5763 reason for rejecting RCPT. Now it says "Previous (cached) callout
5764 verification failure".
5766 3. There was an off-by-one bug in the queryprogram router. An over-long
5767 return line was truncated at 256 instead of 255 characters, thereby
5768 overflowing its buffer with the terminating zero. As well as fixing this, I
5769 have increased the buffer size to 1024 (and made a note to document this).
5771 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5772 when Exim is waiting for an SMTP response from a remote server, Exim
5773 restarts its select() call on the socket, thereby resetting its timeout.
5774 This is not a problem when such interrupts are rare. Somebody set up a cron
5775 job to run exiwhat every 2 minutes, which is less than the normal select()
5776 timeout (5 or 10 minutes). This meant that the select() timeout never
5777 kicked in because it was always reset. I have fixed this by comparing the
5778 time when an interrupt arrives with the time at the start of the first call
5779 to select(). If more time than the timeout has elapsed, the interrupt is
5780 treated as a timeout.
5782 5. Some internal re-factoring in preparation for the addition of Sieve
5783 extensions (by MH). In particular, the "personal" test is moved to a
5784 separate function, and given an option for scanning Cc: and Bcc: (which is
5785 not set for Exim filters).
5787 6. When Exim created an email address using the login of the caller as the
5788 local part (e.g. when creating a From: or Sender: header line), it was not
5789 quoting the local part when it contained special characters such as @.
5791 7. Installed new OpenBSD configuration files.
5793 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5794 try to make them clearer.
5796 9. Callout options, other than the timeout value, were being ignored when
5797 verifying sender addresses in header lines. For example, when using
5799 verify = header_sender/callout=no_cache
5801 the cache was (incorrectly) being used.
5803 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5804 only the Cygwin environment.
5806 11. When return_path_on_delivery was set as a log selector, if different remote
5807 addresses in the same message used different return paths and parallel
5808 remote delivery occurred, the wrong values would sometimes be logged.
5809 (Whenever a remote delivery process finished, the return path value from
5810 the most recently started remote delivery process was logged.)
5812 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5813 lines when AUTH and/or TLS are in use. This is the "received protocol"
5814 field. Exim used to use "asmtp" for authenticated SMTP, without any
5815 indication (in the protocol name) for TLS use. Now it follows the RFC and
5816 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5817 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5818 names appear in log lines as well as in Received: header lines.
5820 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5821 extensions, and comparison tests, and to fix some bugs.
5823 14. Changes to the "personal" filter test:
5825 (1) The test was buggy in that it was just doing the equivalent of
5826 "contains" tests on header lines. For example, if a user's address was
5827 anne@some.where, the "personal" test would incorrectly be true for
5829 To: susanne@some.where
5831 This test is now done by extracting each address from the header in turn,
5832 and checking the entire address. Other tests that are part of "personal"
5833 are now done using regular expressions (for example, to check local parts
5834 of addresses in From: header lines).
5836 (2) The list of non-personal local parts in From: addresses has been
5837 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5838 taken from the Sieve specification recommendations.
5840 (3) If the message contains any header line starting with "List-" it is
5841 treated as non-personal.
5843 (4) The test for "circular" in the Subject: header line has been removed
5844 because it now seems ill-conceived.
5846 15. Minor typos in src/EDITME comments corrected.
5848 16. Installed latest exipick from John Jetmore.
5850 17. If headers_add on a router specified a text string that was too long for
5851 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5852 of string_sprintf() is now avoided.
5854 18. $message_body_size was not set (it was always zero) when running the DATA
5855 ACL and the local_scan() function.
5857 19. For the "mail" command in an Exim filter, no default was being set for
5858 the once_repeat time, causing a random time value to be used if "once" was
5859 specified. (If the value happened to be <= 0, no repeat happened.) The
5860 default is now 0s, meaning "never repeat". The "vacation" command was OK
5861 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5862 (I found it when inspecting the code).
5864 20. There is now an overall timeout for performing a callout verification. It
5865 defaults to 4 times the callout timeout, which applies to individual SMTP
5866 commands during the callout. The overall timeout applies when there is more
5867 than one host that can be tried. The timeout is checked before trying the
5868 next host. This prevents very long delays if there are a large number of
5869 hosts and all are timing out (e.g. when the network connections are timing
5870 out). The value of the overall timeout can be changed by specifying an
5871 additional sub-option for "callout", called "maxwait". For example:
5873 verify = sender/callout=5s,maxwait=20s
5875 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5876 to the end before writing, but this should make it even safer).
5878 22. Exim was forgetting that it had advertised PIPELINING for the second and
5879 subsequent messages on an SMTP connection. It was also not resetting its
5880 memory on STARTTLS and an internal HELO.
5882 23. When Exim logs an SMTP synchronization error within a session, it now
5883 records whether PIPELINING has been advertised or not.
5885 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5886 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5887 rather than long int.
5889 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5891 26. Added the never_mail option to autoreply.
5897 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5898 crash if the getsockname() call failed; this can happen if a connection is
5899 closed very soon after it is established. The problem was simply in the
5900 order in which certain operations were done, causing Exim to try to write
5901 to the SMTP stream before it had set up the file descriptor. The bug has
5902 been fixed by making things happen in the correct order.
5908 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5909 before the connection was closed, thus losing the rejection response.
5911 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5912 some early Solaris releases, but causes trouble in current releases where
5913 socklen_t is defined.
5915 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5918 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5921 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5922 that releasing the top of it at the end releases what was used for sub-
5923 expansions (unless the block got too big). However, discard this block if
5924 the first thing is a variable or header, so that we can use its block when
5925 it is dynamic (useful for very large $message_headers, for example).
5927 6. Lookups now cache *every* query, not just the most recent. A new, separate
5928 store pool is used for this. It can be recovered when all lookup caches are
5929 flushed. Lookups now release memory at the end of their result strings.
5930 This has involved some general refactoring of the lookup sources.
5932 7. Some code has been added to the store_xxx() functions to reduce the amount
5933 of flapping under certain conditions.
5935 8. log_incoming_interface used to affect only the <= reception log lines. Now
5936 it causes the local interface and port to be added to several more SMTP log
5937 lines, for example "SMTP connection from", and rejection lines.
5939 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5941 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5943 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5944 could be overwritten at the end of the current message (or the start of a
5945 new message if it was set in a HELO ACL). The value is now preserved for
5946 the duration of the SMTP connection.
5948 12. If a transport had a headers_rewrite setting, and a matching header line
5949 contained an unqualified address, that address was qualified, even if it
5950 did not match any rewriting rules. The underlying bug was that the values
5951 of the flags that permit the existence of unqualified sender and recipient
5952 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5953 non-local messages, and by -bnq for local messages) were not being
5954 preserved with the message after it was received.
5956 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5957 "next input=" as part of the text comprising the host identity instead of
5958 the correct text. The code was using the same buffer for two different
5959 strings. However, depending on which order the printing function evaluated
5960 its arguments, the bug did not always show up. Under Linux, for example, my
5961 test suite worked just fine.
5963 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5964 doesn't work with some older versions of Perl. It has been changed to "my",
5965 which in any case is probably the better facility to use.
5967 15. A really picky compiler found some instances of statements for creating
5968 error messages that either had too many or two few arguments for the format
5971 16. The size of the buffer for calls to the DNS resolver has been increased
5972 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5973 for addresses that have a lot of PTR records. This alleviates a problem; it
5974 does not fully solve it.
5976 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5977 buffer now truncates the list and logs the incident, which is the same
5978 action as happens when Exim is looking up a host name and its aliases.
5979 Previously in this situation something unpredictable would happen;
5980 sometimes it was "internal error: store_reset failed".
5982 18. If a server dropped the connection unexpectedly when an Exim client was
5983 using GnuTLS and trying to read a response, the client delivery process
5984 crashed while trying to generate an error log message.
5986 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5987 string, for example:
5989 warn message = H1: something\nH2: something
5991 the text was added as a single header line from Exim's point of view
5992 though it ended up OK in the delivered message. However, searching for the
5993 second and subsequent header lines using $h_h2: did not work. This has been
5994 fixed. Similarly, if a system filter added multiple headers in this way,
5995 the routers could not see them.
5997 20. Expanded the error message when iplsearch is called with an invalid key to
5998 suggest using net-iplsearch in a host list.
6000 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6001 ACLs are no longer actually imposed (and a message to that effect is
6004 22. If a "gecos" field in a passwd entry contained escaped characters, in
6005 particular, if it contained a \" sequence, Exim got it wrong when building
6006 a From: or a Sender: header from that name. A second bug also caused
6007 incorrect handling when an unquoted " was present following a character
6008 that needed quoting.
6010 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6011 was not being matched caselessly.
6013 24. Arranged for all hyphens in the exim.8 source to be escaped with
6016 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6017 properly. Recipient callout cache records were still being keyed to include
6018 the sender, even when use_sender was set false. This led to far more
6019 callouts that were necessary. The sender is no longer included in the key
6020 when use_sender is false.
6022 26. Added "control = submission" modifier to ACLs.
6024 27. Added the ${base62d: operator to decode base 62 numbers.
6026 28. dnsdb lookups can now access SRV records.
6028 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6029 the configuration file.
6031 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6032 (-v) mode. This makes the output for a verbose queue run more intelligible.
6034 31. Added a use_postmaster feature to recipient callouts.
6036 32. Added the $body_zerocount variable, containing the number of binary zero
6037 bytes in the message body.
6039 33. The time of last modification of the "new" subdirectory is now used as the
6040 "mailbox time last read" when there is a quota error for a maildir
6043 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6045 35. Added +ignore_unknown as a special item in host lists.
6047 36. Code for decoding IPv6 addresses in host lists is now included, even if
6048 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6049 address was recognized as an IP address, but was then not correctly decoded
6050 into binary, causing unexpected and incorrect effects when compared with
6057 1. Very minor rewording of debugging text in manualroute to say "list of
6058 hosts" instead of "hostlist".
6060 2. If verify=header_syntax was set, and a header line with an unqualified
6061 address (no domain) and a large number of spaces between the end of the
6062 name and the colon was received, the reception process suffered a buffer
6063 overflow, and (when I tested it) crashed. This was caused by some obsolete
6064 code that should have been removed. The fix is to remove it!
6066 3. When running in the test harness, delay a bit after writing a bounce
6067 message to get a bit more predictability in the log output.
6069 4. Added a call to search_tidyup() just before forking a reception process. In
6070 theory, someone could use a lookup in the expansion of smtp_accept_max_
6071 per_host which, without the tidyup, could leave open a database connection.
6073 5. Added the variables $recipient_data and $sender_data which get set from a
6074 lookup success in an ACL "recipients" or "senders" condition, or a router
6075 "senders" option, similar to $domain_data and $local_part_data.
6077 6. Moved the writing of debug_print from before to after the "senders" test
6080 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6081 problems for message scanning, either using a data ACL, or using
6082 local_scan() because the Received: header was not generated till after they
6083 were called (in order to set the time as the time of reception completion).
6084 I have revised the way this works. The header is now generated after the
6085 body is received, but before the ACL or local_scan() are called. After they
6086 are run, the timestamp in the header is updated.
6092 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6093 before starting a queue runner without re-exec. This happened only when
6094 deliver_drop_privilege was set or when the Exim user was set to root. The
6095 effect of the bug was that timeouts during subsequent deliveries caused
6096 crashes instead of being properly handled. The handler is now left at its
6097 default (and expected) setting.
6099 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6100 message, again when deliver_drop_privilege is set or Exim is run as root.
6101 The bug described in (1) was not present in this case, but the tidying up
6102 of the other signals was missing. I have made the two cases consistent.
6104 3. The ignore_target_hosts setting on a manualroute router was being ignored
6105 for hosts that were looked up using the /MX notation.
6107 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6110 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6111 operated on the sender address. After changing the $sender_address to <>
6112 for the sender address verify, Exim was re-instated it as the original
6113 (before rewriting) address, but remembering that it had rewritten it, so it
6114 wasn't rewriting it again. This bug also had the effect of breaking the
6115 sender address verification caching when the sender address was rewritten.
6117 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6118 This has been changed so that if the ip literal address matches
6119 ignore_target_hosts, the router declines.
6121 7. Added expansion conditions match_domain, match_address, and match_local_
6122 part (NOT match_host).
6124 8. The placeholder for the Received: header didn't have a length field set.
6126 9. Added code to Exim itself and to exim_lock to test for a specific race
6127 condition that could lead to file corruption when using MBX delivery. The
6128 issue is with the lockfile that is created in /tmp. If this file is removed
6129 after a process has opened it but before that process has acquired a lock,
6130 there is the potential for a second process to recreate the file and also
6131 acquire a lock. This could lead to two Exim processes writing to the file
6132 at the same time. The added code performs the same test as UW imapd; it
6133 checks after acquiring the lock that its file descriptor still refers to
6134 the same named file.
6136 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6137 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6140 11. Added the smtp_active_hostname option. If used, this will typically be made
6141 to depend on the incoming interface address. Because $interface_address is
6142 not set up until the daemon has forked a reception process, error responses
6143 that can happen earlier (such as "too many connections") no longer contain
6146 12. If an expansion in a condition on a "warn" statement fails because a lookup
6147 defers, the "warn" statement is abandoned, and the next ACL statement is
6148 processed. Previously this caused the whole ACL to be aborted.
6150 13. Added the iplsearch lookup type.
6152 14. Added ident_timeout as a log selector.
6154 15. Added tls_certificate_verified as a log selector.
6156 16. Added a global option tls_require_ciphers (compare the smtp transport
6157 option of the same name). This controls incoming TLS connections.
6159 17. I finally figured out how to make tls_require_ciphers do a similar thing
6160 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6161 before starting the TLS session.
6163 18. Tabs are now shown as \t in -bP output.
6165 19. If the log selector return_path_on_delivery was set, Exim crashed when
6166 bouncing a message because it had too many Received: header lines.
6168 20. If two routers both had headers_remove settings, and the first one included
6169 a superfluous trailing colon, the final name in the first list and the
6170 first name in the second list were incorrectly joined into one item (with a
6171 colon in the middle).
6177 1. Added -C and -D options to the exinext utility, mainly to make it easier
6178 to include in the automated testing, but these could be helpful when
6179 multiple configurations are in use.
6181 2. The exinext utility was not formatting the output nicely when there was
6182 an alternate port involved in the retry record key, nor when there was a
6183 message id as well (for retries that were specific to a specific message
6184 and a specific host). It was also confused by IPv6 addresses, because of
6185 the additional colons they contain. I have fixed the IPv4 problem, and
6186 patched it up to do a reasonable job for IPv6.
6188 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6189 delivery, the log line now contains "pipelined" if PIPELINING was used.
6191 4. An SMTP transport process used to panic and die if the bind() call to set
6192 an explicit outgoing interface failed. This has been changed; it is now
6193 treated in the same way as a connect() failure.
6195 5. A reference to $sender_host_name in the part of a conditional expansion
6196 that was being skipped was still causing a DNS lookup. This no longer
6199 6. The def: expansion condition was not recognizing references to header lines
6200 that used bh_ and bheader_.
6202 7. Added the _cache feature to named lists.
6204 8. The code for checking quota_filecount in the appendfile transport was
6205 allowing one more file than it should have been.
6207 9. For compatibility with Sendmail, the command line option
6215 and sets the incoming protocol and host name (for trusted callers). The
6216 host name and its colon can be omitted when only the protocol is to be set.
6217 Note the Exim already has two private options, -pd and -ps, that refer to
6218 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6219 "s", but I don't think that's a major issue.
6221 10. A number of refactoring changes to the code, none of which should affect
6224 (a) The number of logging options was getting close to filling up the
6225 32-bit word that was used as a bit map. I have split them into two classes:
6226 those that are passed in the argument to log_write(), and those that are
6227 only ever tested independently outside of that function. These are now in
6228 separate 32-bit words, so there is plenty of room for expansion again.
6229 There is no change in the user interface or the logging behaviour.
6231 (b) When building, for example, log lines, the code previously used a
6232 macro that called string_cat() twice, in order to add two strings. This is
6233 not really sufficiently general. Furthermore, there was one instance where
6234 it was actually wrong because one of the argument was used twice, and in
6235 one call a function was used. (As it happened, calling the function twice
6236 did not affect the overall behaviour.) The macro has been replaced by a
6237 function that can join an arbitrary number of extra strings onto a growing
6240 (c) The code for expansion conditions now uses a table and a binary chop
6241 instead of a serial search (which was left over from when there were very
6242 few conditions). Also, it now recognizes conditions like "pam" even when
6243 the relevant support is not compiled in: a suitably worded error message is
6244 given if an attempt is made to use such a condition.
6246 11. Added ${time_interval:xxxxx}.
6248 12. A bug was causing one of the ddress fields not to be passed back correctly
6249 from remote delivery subprocesses. The field in question was not being
6250 subsequently used, so this caused to problems in practice.
6252 13. Added new log selectors queue_time and deliver_time.
6254 14. Might have fixed a bug in maildirsizefile handling that threw up
6255 "unexpected character" debug warnings, and recalculated the data
6256 unnecessarily. In any case, I expanded the warning message to give more
6259 15. Added the message "Restricted characters in address" to the statements in
6260 the default ACL that block characters like @ and % in local parts.
6262 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6263 Three changes have been made:
6265 (a) There was a serious bug; a negative response to MAIL caused the whole
6266 recipient domain to be cached as invalid, thereby blocking all messages
6267 to all local parts at the same domain, from all senders. This bug has
6268 been fixed. The domain is no longer cached after a negative response to
6269 MAIL if the sender used is not empty.
6271 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6274 (c) A new callout option, "use_sender" has been added for people who want
6275 the modified behaviour.
6281 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6284 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6285 indeed breaks things for older releases.
6287 3. Added additional logging to the case where there is a problem reading data
6288 from a filter that is running in a subprocess using a pipe, in order to
6289 try to track down a specific problem.
6291 4. Testing facility fudge: when running in the test harness and attempting
6292 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6293 getting "No route to host". Convert this to a timeout.
6295 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6298 6. Some OS don't have socklen_t but use size_t instead. This affects the
6299 fifth argument of getsockopt() amongst other things. This is now
6300 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6301 can be set for individual OS. I have set it for SunOS5, OSF1, and
6302 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6303 some earlier ones do not.
6305 7. Change 4.30/15 was not doing the test caselessly.
6307 8. The standard form for an IPv6 address literal was being rejected by address
6308 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6309 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6310 this, as well as the form without the "IPv6" on the front (but only when
6311 address literals are enabled, of course).
6313 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6315 10. Exim crashed if a message with an empty sender address specified by -f
6316 encountered a router with an errors_to setting. This could be provoked only
6317 by a command such as
6321 where an empty string was supplied; "<>" did not hit this bug.
6323 11. Installed PCRE release 4.5.
6325 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6326 remained set. It is now erased.
6328 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6329 times from message ids (which are base 36 rather than the normal 62).
6331 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6332 were being counted as actual protocol errors, and logged if the log
6333 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6334 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6335 and DATA following a set of rejected RCPTs do not count as protocol errors.
6336 In other words, Exim assumes they were pipelined, though this may not
6337 actually be the case. Of course, in all cases the client gets an
6338 appropriate error code.
6340 15. If a lookup fails in an ACL condition, a message about the failure may
6341 be available; it is used if testing the ACL cannot continue, because most
6342 such messages specify what the cause of the deferral is. However, some
6343 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6344 that caused an old message to be retained and used if a later statement
6345 caused a defer, replacing the real cause of the deferral.
6347 16. If an IP address had so many PTR records that the DNS lookup buffer
6348 was not large enough to hold them, Exim could crash while trying to process
6349 the truncated data. It now detects and logs this case.
6351 17. Further to 4.21/58, another change has been made: if (and only if) the
6352 first line of a message (the first header line) ends with CRLF, a bare LF
6353 in a subsequent header line has a space inserted after it, so as not to
6354 terminate the header.
6356 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6357 unnecessarily, used atoi() instead of strtol(), and didn't check the
6358 termination when getting file sizes from file names by regex.
6360 19. Completely re-implemented the support for maildirsize files, in the light
6361 of a number of problems with the previous contributed implementation
6362 (4.30/29). In particular:
6364 . If the quota is zero, the maildirsize file is maintained, but no quota is
6367 . If the maildir directory does not exist, it is created before any attempt
6368 to write a maildirsize file.
6370 . The quota value in the file is just a cache; if the quota is changed in
6371 the transport, the new value overrides.
6373 . A regular expression is available for excluding directories from the
6376 20. The autoreply transport checks the characters in options that define the
6377 message's headers; it allows continued headers, but it was checking with
6378 isspace() after an embedded newline instead of explicitly looking for a
6381 21. If all the "regular" hosts to which an address was routed had passed their
6382 expiry times, and had not reached their retry times, the address was
6383 bounced, even if fallback hosts were defined. Now Exim should go on to try
6386 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6387 equivalent code in the SMTP transport. Some hosts send humungous responses
6388 to HELO/EHLO, more than 1024 it seems.
6390 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6391 gives compiler warnings in some environments. I've now done it "properly",
6394 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6395 (because of problems with the built-in one) was declared to return uschar *
6396 instead of char *, causing compiler failure.
6398 25. Fixed a file descriptor leak when processing alias/forward files.
6400 26. Fixed a minor format string issue in dbfn.c.
6402 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6404 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6405 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6408 29. When a sender address is verified, it is cached, to save repeating the test
6409 when there is more than one recipient in a message. However, when the
6410 verification involves a callout, it is possible for different callout
6411 options to be set for different recipients. It is too complicated to keep
6412 track of this in the cache, so now Exim always runs a verification when a
6413 callout is required, relying on the callout cache for the optimization.
6414 The overhead is duplication of the address routing, but this should not be
6417 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6418 to be verified with callout=postmaster, and the main callout worked but the
6419 postmaster check failed, the verification correctly failed. However, if a
6420 subsequent RCPT command asked for sender verification *without* the
6421 postmaster check, incorrect caching caused this verification also to fail,
6424 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6425 it was not caching the DNS options (qualify_single, search_parents) that
6426 were used when the lookup failed. A subsequent lookup with different
6427 options therefore always gave the same answer, though there were cases
6428 where it should not have. (Example: a "domains = !$mx_any" option on a
6429 dnslookup router: the "domains" option is always processed without any
6430 widening, but the router might have qualify_single set.) Now Exim uses the
6431 cached value only when the same options are set.
6433 32. Added John Jetmore's "exipick" utility to the distribution.
6435 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6436 than a timeout (e.g. a certificate is required, and is not provided), an
6437 Exim server now closes the connection immediately. Previously it waited for
6438 the client to close - but if the client is SSL, it seems that they each
6439 wait for each other, leading to a delay before one of them times out.
6441 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6442 maintained 0.8.x compatibility because I don't think many are using it, and
6443 it is clearly obsolete.
6445 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6448 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6449 client certificate was expired. A simple patch fixes this, though I don't
6450 understand the full logic of why the verify callback is called multiple
6453 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6454 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6455 which causes problems with some clients (such as the Certicom SSL Plus
6456 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6457 disables the coutermeasure allowing Eudora to connect."
6459 38. Exim was not checking that a write() to a log file succeeded. This could
6460 lead to Bad Things if a log got too big, in particular if it hit a file
6461 size limit. Exim now panics and dies if it cannot write to a log file, just
6462 as it does if it cannot open a log file.
6464 39. Modified OS/Makefile-Linux so that it now contains
6466 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6468 The two -D definitions ensure that Exim is compiled with large file
6469 support, which makes it possible to handle log files that are bigger than
6472 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6473 instance) a domain was checked against a named list that involved a lookup,
6474 causing $domain_data to be set, then another domain was checked against the
6475 same list, then the first domain was re-checked, the value of $domain_data
6476 after the final check could be wrong. In particular, if the second check
6477 failed, it could be set empty. This bug probably also applied to
6480 41. The strip_trailing_dot option was not being applied to the address given
6481 with the -f command-line option.
6483 42. The code for reading a message's header from the spool was incrementing
6484 $received_count, but never initializing it. This meant that the value was
6485 incorrect (doubled) while delivering a message in the same process in which
6486 it was received. In the most common configuration of Exim, this never
6487 happens - a fresh exec is done - but it can happen when
6488 deliver_drop_privilege is set.
6490 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6491 it now includes up to 150 characters of the unexpected data in the log
6494 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6495 and building data strings. The size of both of these buffers was 10 000
6496 bytes - far larger than anybody would *ever* want, thought I. Needless to
6497 say, somebody hit the limit. I have increased the maximum line length to
6498 20 000 and the maximum data length of concatenated lines to 100 000. I have
6499 also fixed two bugs, because there was no checking on these buffers. Tsk,
6500 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6501 buffer is too small.
6503 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6504 lsearch lookups. Now it does.
6506 46. When parsing a route_list item in a manualroute router, a fixed-length
6507 buffer was used for the list of hosts. I made this 1024 bytes long,
6508 thinking that nobody would ever have a list of hosts that long. Wrong.
6509 Somebody had a whole pile of complicated expansion conditions, and the
6510 string was silently truncated, leading to an expansion error. It turns out
6511 that it is easier to change to an unlimited length (owing to other changes
6512 that have happened since this code was originally written) than to build
6513 structure for giving a limitation error. The length of the item that
6514 expands into the list of hosts is now unlimited.
6516 47. The lsearch lookup could not handle data where the length of text line was
6517 more than 4095 characters. Such lines were truncated, leading to shortened
6518 data being returned. It should now handle lines of any length.
6520 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6521 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6524 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6525 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6526 no longer happens in comment lines. A list of replacements is now placed
6527 at the head of all of the source files, except those whose only change is
6528 to replace PERL_COMMAND in the very first #! line.
6530 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6531 messages on the queue, with a bottom-up merge sort, using code contributed
6532 by Michael Haardt. This should make operations like -bp somewhat faster on
6533 large queues. It won't affect queue runners, except when queue_run_in_order
6536 51. Installed eximstats 1.31 in the distribution.
6538 52. Added support for SRV lookups to the dnslookup router.
6540 53. If an ACL referred to $message_body or $message_body_end, the value was not
6541 reset for any messages that followed in the same SMTP session.
6543 54. The store-handling optimization for building very long strings was not
6544 differentiating between the different store pools. I don't think this
6545 actually made any difference in practice, but I've tidied it.
6547 55. While running the routers to verify a sender address, $sender_address
6548 was still set to the sender address. This is wrong, because when routing to
6549 send a bounce to the sender, it would be empty. Therefore, I have changed
6550 it so that, while verifying a sender address, $sender_address is set to <>.
6551 (There is no change to what happens when verifying a recipient address.)
6553 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6554 target A or AAAA records (if not already returned) without resetting the
6555 qualify_single or search_parents options of the DNS resolver. These are
6556 inappropriate in this case because the targets of MX and SRV records must
6557 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6558 target that, when qualified, matched something in the local domain. These
6559 two options are now turned off when doing these lookups.
6561 57. It seems that at least some releases of Reiserfs (which does not have the
6562 concept of a fixed number of inodes) returns zero and not -1 for the
6563 number of available inodes. This interacted badly with check_spool_inodes,
6564 which assumed that -1 was the "no such thing" setting. What I have done is
6565 to check that the total number of inodes is greater than zero before doing
6566 the test of how many are available.
6568 58. When a "warn" ACL statement has a log_message modifier, the message is
6569 remembered, and not repeated. This is to avoid a lot of repetition when a
6570 message has many recipients that cause the same warning to be written.
6571 However, Exim was preserving the list of already written lines for an
6572 entire SMTP session, which doesn't seem right. The memory is now reset if a
6573 new message is started.
6575 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6576 debugging output unless log_rewrite was also set.
6578 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6579 of (int)(handle) when we know that handle contains (void *)(-1).
6581 61. The Exim daemon panic-logs an error return when it closes the incoming
6582 connection. However "connection reset by peer" seems to be common, and
6583 isn't really an error worthy of noting specially, so that particular error
6586 62. When Exim is trying to find all the local interfaces, it used to panic and
6587 die if the ioctl to get the interface flags failed. However, it seems that
6588 on at least one OS (Solaris 9) it is possible to have an interface that is
6589 included in the list of interfaces, but for which you get a failure error
6590 for this call. This happens when the interface is not "plumbed" into a
6591 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6592 failure of the "get flags" call assumes that the interface is down.
6594 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6595 makes life easier for people who are doing arithmetic on fields extracted
6596 from dates, where you often get leading zeros that should not be
6597 interpreted as octal.
6599 64. Added qualify_domain to the redirect router, to override the global
6602 65. If a pathologically long header line contained very many addresses (the
6603 report of this problem mentioned 10 000) and each of them was rewritten,
6604 Exim could use up a very large amount of memory. (It kept on making new
6605 copies of the header line as it rewrote, and never released the old ones.)
6606 At the expense of a bit more processing, the header rewriting function has
6607 been changed so that it no longer eats memory in this way.
6609 66. The generation of the Received: header has been moved from the time that a
6610 message starts to be received, to the time that it finishes. The timestamp
6611 in the Received: header should now be very close to that of the <= log
6612 line. There are two side-effects of this change:
6614 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6615 logged header lines no longer include the local Received: line, because
6616 it has not yet been created. The same applies to a copy of the message
6617 that is returned to a non-SMTP sender when a message is rejected.
6619 (b) When a filter file is tested using -bf, no additional Received: header
6620 is added to the test message. After some thought, I decided that this
6623 This change does not affect the value of $received_for. It is still set
6624 after address rewriting, but before local_scan() is called.
6626 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6628 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6629 gave an unhelpful panic error message, and a defer error. I have managed to
6630 change this behaviour so that it now rejects any supplied certificate,
6631 which seems right, as the list of acceptable certificates is empty.
6633 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6634 gave an unhelpful defer error. I have not managed to make this reject any
6635 supplied certificates, but the error message it gives is "no certificate
6636 supplied", which is not helpful.
6638 70. exigrep's output now also includes lines that are not associated with any
6639 message, but which match the given pattern. Implemented by a patch from
6640 Martin Sluka, which also tidied up the Perl a bit.
6642 71. Recipient callout verification, like sender verification, was using <> in
6643 the MAIL FROM command. This isn't really the right thing, since the actual
6644 sender may affect whether the remote host accepts the recipient or not. I
6645 have changed it to use the actual sender in the callout; this means that
6646 the cache record is now keyed on a recipient/sender pair, not just the
6647 recipient address. There doesn't seem to be a real danger of callout loops,
6648 since a callout by the remote host to check the sender would use <>.
6649 [SEE ABOVE: changed after hitting problems.]
6651 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6652 temporary errors. However, in the case of such a code being given after
6653 the end of a data transmission (i.e. after ".") Exim was failing to write
6654 a retry record for the message. (Yes, there was some broken host that was
6655 actually sending 8xx at this point.)
6657 73. An unknown lookup type in a host list could cause Exim to panic-die when
6658 the list was checked. (An example that provoked this was putting <; in the
6659 middle of a list instead of at the start.) If this happened during a DATA
6660 ACL check, a -D file could be left lying around. This kind of configuration
6661 error no longer causes Exim to die; instead it causes a defer error. The
6662 incident is still logged to the main and panic logs.
6664 74. Buglet left over from Exim 3 conversion. The message "too many messages
6665 in one connection" was written to the rejectlog but not the mainlog, except
6666 when address rewriting (yes!) was being logged.
6668 75. Added write_rejectlog option.
6670 76. When a system filter was run not as root (that is, when system_filter_user
6671 was set), the values of the $n variables were not being returned to the
6672 main process; thus, they were not subsequently available in the $sn
6675 77. Added +return_path_on_delivery log selector.
6677 78. A connection timeout was being treated differently from recipients deferred
6678 when testing hosts_max_try with a message that was older than the host's
6679 retry timeout. (The host should not be counted, thus allowing all hosts to
6680 be tried at least once before bouncing.) This may have been the cause of an
6681 occasionally reported bug whereby a message would remain on the queue
6682 longer than the retry timeout, but would be bounced if a delivery was
6683 forced. I say "may" because I never totally pinned down the problem;
6684 setting up timeout/retry tests is difficult. See also the next item.
6686 79. The ultimate address timeout was not being applied to errors that involved
6687 a combination of host plus message (for example, a timeout on a MAIL
6688 command). When an address resolved to a number of possible hosts, and they
6689 were not all tried for each delivery (e.g. because of hosts_max_try), a
6690 message could remain on the queue longer than the retry timeout.
6692 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6695 81. Fixed an obscure SMTP outgoing bug which required at least the following
6696 conditions: (a) there was another message waiting for the same server;
6697 (b) the server returned 5xx to all RCPT commands in the first message so
6698 that the message was not completed; (c) the server dropped the connection
6699 or gave a negative response to the RSET that Exim sends to abort the
6700 transaction. The observed case was a dropped connection after DATA that had
6701 been sent in pipelining mode. That is, the server had advertised PIPELINING
6702 but was not implementing it correctly. The effect of the bug was incorrect
6703 behaviour, such as trying another host, and this could lead to a crash.
6709 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6710 and daemon.c were passed as pointers to ints; they should have been
6711 pointers to socklen_t variables (which are typically unsigned ints).
6713 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6716 3. Fixed a really odd bug that affected only the testing scheme; patching a
6717 certain fixed string in the binary changed the value of another string that
6718 happened to be identical to the end of the original first string.
6720 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6721 name", it returns that address as the IP address. On some operating
6722 systems (e.g. Solaris), it also passes back the IP address string as the
6723 "host name". However, on others (e.g. Linux), it passes back an empty
6724 string. Exim wasn't checking for this, and was changing the host name to an
6725 empty string, assuming it had been canonicalized.
6727 5. Although rare, it is permitted to have more than one PTR record for a given
6728 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6729 all the names associated with an address, because they do in Solaris.
6730 However, it seems that they do not in Linux for data that comes from the
6731 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6732 I found this out when I moved to a new Linux workstation and tried to run
6733 the Exim test suite.
6735 To get round this problem I have changed the code so that it now does its
6736 own call to the DNS to look up PTR records when searching for a host name.
6737 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6738 addresses that are only in /etc/hosts are still found.
6740 This behaviour is, however, controlled by an option called host_lookup_
6741 order, which defaults to "bydns:byaddr". If people want to use the other
6742 order, or indeed, just use one or the other means of lookup, they can
6743 specify it in this variable.
6745 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6746 some operating systems, this comes back from gethostbyaddr() as an empty
6747 string, and this is what Exim used to test for. However, it seems that in
6748 other systems, "." is yielded. Exim now tests for this case too.
6750 7. The values of check_spool_space and check_log_space are now held internally
6751 as a number of kilobytes instead of an absolute number of bytes. If a
6752 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6753 kilobyte. This means that much larger values can be stored.
6755 8. Exim monitor: an attempt to get the action menu when not actually pointing
6756 at a message produces an empty menu entitled "No message selected". This
6757 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6758 no entries in it ("Shell widget menu has zero width and/or height"). So I
6759 have added a single, blank menu entry in this case.
6761 9. Added ${quote_local_part.
6763 10. MIME decoding is now applied to the contents of Subject: header lines when
6766 11. Now that a reference to $sender_host_address automatically causes a reverse
6767 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6768 host lookup before query-style lookups in lists that might use this
6769 variable. This has therefore been abolished, and the "net-" prefix is no
6770 longer necessary for query-style lookups.
6772 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6773 have been a typo for LFLAGS, so it has been changed.
6775 13. The install script calls Exim with "-C /dev/null" in order to find the
6776 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6777 to be output. However, since Exim outputs its version number before the
6778 error, it didn't break the script. It just looked ugly. I fixed this by
6779 always allowing "-C /dev/null" if the caller is root.
6781 14. Ignore overlarge ACL variable number when reading spool file - insurance
6782 against a later release with more variables having written the file.
6784 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6785 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6786 this, as well as the form without the "IPv6" on the front.
6788 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6789 OS/Makefile-Darwin file.
6791 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6792 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6794 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6795 variable. However, some OS use ints for inodes. I've added cast to long int
6796 to get rid of the compiler warning.
6798 19. I had forgotten to lock out "/../" in configuration file names when
6799 ALT_CONFIG_PREFIX was set.
6801 20. Routers used for verification do not need to specify transports. However,
6802 if such a router generated a host list, and callout was configured, Exim
6803 crashed, because it could not find a port number from the (non-existent)
6804 transport. It now assumes port 25 in this circumstance.
6806 21. Added the -t option to exigrep.
6808 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6809 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6810 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6811 them is set, LOOKUP_LSEARCH is forced.
6813 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6815 24. Added sender and host information to the "rejected by local_scan()" log
6816 line; previously there was no indication of these.
6818 25. Added .include_if_exists.
6820 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6821 Linux. It turns out that not all file systems support this. Apparently some
6822 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6823 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6824 not supported on the file descriptor, is now ignored when Exim is trying to
6825 sync a directory. This applies only to Linux.
6827 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6829 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6830 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6833 29. Added support for maildirsize files from supplied patch (modified a bit).
6835 30. The use of :fail: followed by an empty string could lead Exim to respond to
6836 sender verification failures with (e.g.):
6838 550 Verification failed for <xxx>
6839 550 Sender verify failed
6841 where the first response line was missing the '-' that indicates it is not
6842 the final line of the response.
6844 31. The loop for finding the name of the user that called Exim had a hardwired
6845 limit of 10; it now uses the value of finduser_retries, which is used for
6846 all other user lookups.
6848 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6851 33. Exim was neglecting to zero errno before one call of strtol() when
6852 expanding a string and expecting an integer value. On some systems this
6853 resulted in spurious "integer overflow" errors. Also, it was casting the
6854 result into an int without checking.
6856 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6857 did not work. The code looks as if it has *never* worked, though it appears
6858 to have been documented since at least release 1.62. I have made it work.
6860 35. The "timeout_DNS" error in retry rules, also documented since at least
6861 1.62, also never worked. As it isn't clear exactly what this means, and
6862 clearly it isn't a major issue, I have abolished the feature by treating it
6863 as "timeout", and writing a warning to the main and panic logs.
6865 36. The display of retry rules for -brt wasn't always showing the error code
6868 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6869 timeout_connect_A, timeout_connect_MX.
6871 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6872 to the empty sender.
6874 39. The daemon was not analysing the content of -oX till after it had closed
6875 stderr and disconnected from the controlling terminal. This meant that any
6876 syntax errors were only noted on the panic log, and the return code from
6877 the command was 0. By re-arranging the code a little, I've made the
6878 decoding happen first, so such errors now appear on stderr, and the return
6879 code is 1. However, the actual setting up of the sockets still happens in
6880 the disconnected process, so errors there are still only recorded on the
6883 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6884 connections (as happens on some IP stacks) was logged at start up time as
6885 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6886 it from "IPv6 and IPv4", which means that two separate sockets are being
6889 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6890 says whether AF_INET or AF_INET6 was passed as an argument.
6892 42. Exiwhat output was messed up when time zones were included in log
6895 43. Exiwhat now gives more information about the daemon's listening ports,
6896 and whether -tls-on-connect was used.
6898 44. The "port" option of the smtp transport is now expanded.
6900 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6901 silently ignored. Now an error message is written to the main and panic
6904 46. There's a new ACL modifier called "logwrite" which writes to a log file
6905 as soon as it is encountered.
6907 47. Added $local_user_uid and $local_user_gid at routing time.
6909 48. Exim crashed when trying to verify a sender address that was being
6912 49. Exim was recognizing only a space character after ".include". It now also
6913 recognizes a tab character.
6915 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6916 extracting the relevant information from the specification. The man page no
6917 longer contains scrambled data for the -d option, and I've added a section
6918 at the front about calling Exim under different names.
6920 51. Added "extra_headers" argument to the "mail" command in filter files.
6922 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6925 53. Installed eximstats 1.29.
6927 54. Added transport_filter_timeout as a generic transport option.
6929 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6930 Cc: header lines. This was required by RFC 822, but it not required by RFC
6933 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6934 incoming messages that did not have them. Now it does so only if the
6935 message originates locally, that is, if there is no associated remote host
6936 address. When Resent- header lines are present, this applies to the Resent-
6937 lines rather than the non-Resent- lines.
6939 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6940 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6942 58. Messages for configuration errors now include the name of the main
6943 configuration file - useful now that there may be more than one file in a
6944 list (.included file names were always shown).
6946 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6947 for those rare installations that do not start the daemon as root or run it
6948 setuid root. I've cut out the call to initgroups() if the daemon is not
6951 60. The Exim user and group can now be bound into the binary as text strings
6952 that are looked up at the start of Exim's processing.
6954 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6956 62. Added $mailstore_basename variable.
6958 63. Installed patch to sieve.c from Michael Haardt.
6960 64. When Exim failed to open the panic log after failing to open the main log,
6961 the original message it was trying to log was written to stderr and debug
6962 output, but if they were not available (the usual case in production), it
6963 was lost. Now it is written to syslog before the two lines that record the
6964 failures to open the logs.
6966 65. Users' Exim filters run in subprocesses under the user's uid. It is
6967 possible for a "deliver" command or an alias in a "personal" command to
6968 provoke an address rewrite. If logging of address rewriting is configured,
6969 this fails because the process is not running as root or exim. There may be
6970 a better way of dealing with this, but for the moment (because 4.30 needs
6971 to be released), I have disabled address rewrite logging when running a
6972 filter in a non-root, non-exim process.
6978 1. The buildconfig auxiliary program wasn't quoting the value set for
6979 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6980 not defined. This bug was masked in 4.22 by the effect that was fixed in
6983 2. Some messages that were rejected after a message id was allocated were
6984 shown as "incomplete" by exigrep. It no longer does this for messages that
6985 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6987 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6988 have allow_domain_literals set, the ID did not get logged in the <= line.
6989 Domain literals are now always recognized in Message-ID: header lines.
6991 4. The first argument for a ${extract expansion item is the key name or field
6992 number. Leading and trailing spaces in this item were not being ignored,
6993 causing some misleading effects.
6995 5. When deliver_drop_privilege was set, single queue runner processes started
6996 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6997 same command in the process it spins off) were not dropping privilege.
6999 6. When the daemon running as "exim" started a queue runner, it always
7000 re-executed Exim in the spun-off process. This is a waste of effort when
7001 deliver_drop_privilege is set. The new process now just calls the
7002 queue-runner function directly.
7008 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7011 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7012 ignored. Though the use of -oP was forcing the writing of a pid file, it
7013 was always written to the default place.
7015 3. If the message "no IP address found for host xxxx" is generated during
7016 incoming verification, it is now followed by identification of the incoming
7017 connection (so you can more easily find what provoked it).
7019 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7021 5. Added some features to "harden" Exim a bit more against certain attacks:
7023 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7024 be put in Local/Makefile. This is like the never_users runtime option,
7025 but it cannot be overridden. The default setting is "root".
7027 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7028 prefix string with which any file named in a -C command line option
7031 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7032 is retained for -C and -D only if the caller of Exim is root. Without
7033 it, the exim user may also use -C and -D and retain privilege.
7035 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7036 command line option is disabled.
7038 6. Macro names set by the -D option must start with an upper case letter, just
7039 like macro names defined in the configuration file.
7041 7. Added "dereference=" facility to LDAP.
7043 8. Two instances of the typo "uknown" in the source files are fixed.
7045 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7046 the Configure-Makefile script screwed up while processing it.
7048 10. Incorporated PCRE 4.4.
7050 11. The SMTP synchronization check was not operating right at the start of an
7051 SMTP session. For example, it could not catch a HELO sent before the client
7052 waited for the greeting. There is now a check for outstanding input at the
7053 point when the greeting is written. Because of the duplex, asynchronous
7054 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7055 way, but not yet received, when the check is performed.
7057 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7058 on TCP/IP sockets, because this apparently causes some broken clients to
7061 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7062 unchanged) from the Cygwin maintainer.
7064 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7065 was supported instead of testing for mbx. Effectively a typo.
7067 15. The spa authenticator server code was not checking that the input it
7068 received was valid base64.
7070 16. The debug output line for the "set" modifier in ACLs was not showing the
7071 name of the variable that was being set.
7073 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7075 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7076 lookup on its own. Something else was needed to trigger the lookup. For
7077 example, a match in host_lookup or the need for a host name in a host list.
7078 Now, if $sender_host_name is referenced and the host name has not yet been
7079 looked up, a lookup is performed. If the lookup fails, the variable remains
7080 empty, and $host_lookup_failed is set to "1".
7082 19. Added "eqi" as a case-independent comparison operator.
7084 20. The saslauthd authentication condition could segfault if neither service
7085 nor realm was specified.
7087 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7088 error message that was logged was misleading, and incoming SMTP
7089 connections were dropped. The message is now more accurate, and temporary
7090 errors are given to SMTP connections.
7092 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7093 (except RSET) with a 421 error, until QUIT is received. However, it was
7094 failing to send a response to QUIT.
7096 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7097 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7098 ACL is now run after the helo_try_verify_hosts code.
7100 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7101 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7102 case-independent, so other case variants are also recognized). Apparently
7103 some systems use these upper case variants.
7105 25. If more than two messages were waiting for the same host, and a transport
7106 filter was specified for the transport, Exim sent two messages over the
7107 same TCP/IP connection, and then failed with "socket operation on non-
7108 socket" when it tried to send the third.
7110 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7112 27. The extern definition of crypt16() in expand.c was not being excluded when
7113 the OS had its own crypt16() function.
7115 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7116 as a preferred synonym for return_size_limit, both as an option and as an
7119 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7121 30. Changed the default configuration ACL to relax the local part checking rule
7122 for addresses that are not in any local domains. For these addresses,
7123 slashes and pipe symbols are allowed within local parts, but the sequence
7124 /../ is explicitly forbidden.
7126 31. SPA server authentication was not clearing the challenge buffer before
7129 32. log_message in a "warn" ACL statement was writing to the reject log as
7130 well as to the main log, which contradicts the documentation and doesn't
7131 seem right (because no rejection is happening). So I have stopped it.
7133 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7134 However, I am unable to do any testing of this.
7136 34. Fixed an infelicity in the appendfile transport. When checking directories
7137 for a mailbox, to see if any needed to be created, it was accidentally
7138 using path names with one or more superfluous leading slashes; tracing
7139 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7141 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7142 contents are added to the log line that is written for every discarded
7143 recipient. (Previously a log_message setting was ignored.)
7145 36. The ${quote: operator now quotes the string if it is empty.
7147 37. The install script runs exim in order to find its version number. If for
7148 some reason other than non-existence or emptiness, which it checks, it
7149 could not run './exim', it was installing it with an empty version number,
7150 i.e. as "exim-". This error state is now caught, and the installation is
7153 38. An argument was missing from the function that creates an error message
7154 when Exim fails to connect to the socket for saslauthd authentication.
7155 This could cause Exim to crash, or give a corrupted message.
7157 39. Added isip, isip4, and isip6 to ${if conditions.
7159 40. The ACL variables $acl_xx are now saved with the message, and can be
7160 accessed later in routers, transports, and filters.
7162 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7163 strings in the file are not string-expanded.
7165 42. If a MAIL command specified a SIZE value that was too large to fit into an
7166 int variable, the check against message_size_limit failed. Such values are
7167 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7168 day this will have to be increased, but I don't think I want to be around
7169 when emails are that large.
7176 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7177 iconv() is not standard in FreeBSD.
7179 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7180 IPv6 enabled. The observed symptom was a segmentation fault on return from
7181 the function os_common_find_running_interfaces() in src/os.c.
7183 3. In the check_special_case() function in daemon.c I had used "errno" as an
7184 argument name, which causes warnings on some systems. This was basically a
7185 typo, since it was named "eno" in the comments!
7187 4. The code that waits for the clock to tick (at a resolution of some fraction
7188 of a second) so as to ensure message-id uniqueness was always waiting for
7189 at least one whole tick, when it could have waited for less. [This is
7190 almost certainly not relevant at current processor speeds, where it is
7191 unlikely to ever wait at all. But we try to future-proof.]
7193 5. The function that sleeps for a time interval that includes fractions of a
7194 second contained a race. It did not block SIGALRM between setting the
7195 timer, and suspending (a couple of lines later). If the interval was short
7196 and the sigsuspend() was delayed until after it had expired, the suspension
7197 never ended. On busy systems this could lead to processes getting stuck for
7200 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7201 process, before it forks any delivery processes. The open lookup caching
7202 mechanism meant that the open file or database connection was passed into
7203 the delivery process. The problem was that delivery processes always tidy
7204 up cached lookup data. This could cause a problem for the next delivery
7205 process started by the queue runner, because the external queue runner
7206 process does not know about the closure. So the next delivery process
7207 still has data in the lookup cache. In the case of a file lookup, there was
7208 no problem because closing a file descriptor in a subprocess doesn't affect
7209 the parent. However, if the lookup was caching a connection to a database,
7210 the connection was closed, and the second delivery process was likely to
7211 see errors such as "PGSQL: query failed: server closed the connection
7212 unexpectedly". The problem has been fixed by closing all cached lookups
7213 in a queue runner before running a delivery process.
7215 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7216 seem to have the "const" qualifier which it has on other OS. I've
7219 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7220 *of the same type* (client or server) with the same public name that an
7221 error should be diagnosed.
7223 9. When Exim looked up a host name for an IP address, but failed to find the
7224 original IP address when looking up the host name (a safety check), it
7225 output the message "<ip address> does not match any IP for NULL", which was
7226 confusing, to say the least. The bug was that the host name should have
7227 appeared instead of "NULL".
7229 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7230 user that is built into the binary, and the -C or -D options is used, root
7231 privilege is dropped before the configuration file is read. In addition,
7232 logging is switched to stderr instead of the normal log files. If the
7233 configuration then re-defines the Exim user, the unprivileged environment
7234 is probably not what is expected, so Exim logs a panic warning message (but
7237 However, if deliver_drop_privilege is set, the unprivileged state may well
7238 be exactly what is intended, so the warning has been cut out in that case,
7239 and Exim is allowed to try to write to its normal log files.
7245 1. smtp_return_error_details was not giving details for temporary sender
7246 or receiver verification errors.
7248 2. Diagnose a configuration error if two authenticators have the same public
7251 3. Exim used not to create the message log file for a message until the first
7252 delivery attempt. This could be confusing when incoming messages were held
7253 for policy or load reasons. The message log file is now created at the time
7254 the message is received, and an initial "Received" line is written to it.
7256 4. The automatically generated man page for command line options had a minor
7257 bug that caused no ill effects; however, a more serious problem was that
7258 the procedure for building the man page automatically didn't always
7259 operate. Consequently, release 4.20 contains an out-of-date version. This
7260 shouldn't happen again.
7262 5. When building Exim with embedded Perl support, the script that builds the
7263 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7264 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7266 6. The freeze_tell option was not being used for messages that were frozen on
7267 arrival, either by an ACL or by local_scan().
7269 7. Added the smtp_incomplete_transaction log selector.
7271 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7272 was accepting AUTH without a new EHLO.
7274 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7275 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7278 10. Logging of TCP/IP connections (when configured) now happens in the main
7279 daemon process instead of the child process, so that the TCP/IP connection
7280 count is more accurate (but it can never be perfect).
7282 11. The use of "drop" in a nested ACL was not being handled correctly in the
7283 outer ACL. Now, if condition failure induced by the nested "drop" causes
7284 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7285 or "require"), the connection is dropped.
7287 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7288 that yield "discard" can now be used with an "accept" or a "discard" verb,
7289 but an error is generated for any others (because I can't see a useful way
7290 to define what should happen).
7292 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7293 are now processed in the same way as lines in the Exim configuration file.
7294 In particular, continuation lines are supported.
7296 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7298 15. Added -ti meaning -t -i.
7300 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7301 domains, and warn by logging if others are found.
7303 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7304 returned from GIFCONF when Exim is trying to find the list of interfaces on
7305 a host. The code in os.c has been modified to copy each ifreq to an aligned
7306 structure in all cases.
7308 Also, in some cases, the returned ifreq's were being copied to a 'struct
7309 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7310 means the last couple of bytes of an IPv6 address could be chopped if the
7311 ifreq contained only a normal sockaddr (14 bytes storage).
7313 18. Named domain lists were not supported in the hosts_treat_as_local option.
7314 An entry such as +xxxx was not recognized, and was treated as a literal
7317 19. Ensure that header lines added by a DATA ACL are included in the reject log
7318 if the ACL subsequently rejects the message.
7320 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7321 MD5 (which is deprecated).
7323 21. When testing a filter file using -bf, Exim was writing a message when it
7324 took the sender from a "From " line in the message, but it was not doing so
7325 when it took $return_path from a Return-Path: header line. It now does.
7327 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7328 with a valid header line field name (a series of printing characters
7329 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7331 23. Changed "disc" in the source to "disk" to conform to the documentation and
7332 the book and for uniformity.
7334 24. Ignore Sendmail's -Ooption=value command line item.
7336 25. When execve() failed while trying to run a command in a pipe transport,
7337 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7338 could be confused with a return value of 69 from the command itself. This
7339 has been changed to 127, the value the shell returns if it is asked to run
7340 a non-existent command. The wording for the related log line suggests a
7341 non-existent command as the problem.
7343 26. If received_header_text expands to an empty string, do not add a Received:
7344 header line to the message. (Well, it adds a token one on the spool, but
7345 marks it "old" so that it doesn't get used or transmitted.)
7347 27. Installed eximstats 1.28 (addition of -nt option).
7349 28. There was no check for failure on the call to getsockname() in the daemon
7350 code. This can fail if there is a shortage of resources on the system, with
7351 ENOMEM, for example. A temporary error is now given on failure.
7353 29. Contrary to the C standard, it seems that in some environments, the
7354 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7355 program. Exim now does this explicitly; it affects the formatting of
7356 timestamps using strftime().
7358 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7359 complaints. I've now initialized all the variables, to avoid this.
7361 32. Header lines added by a system filter were not being "seen" during
7362 transport-time rewrites.
7364 33. The info_callback() function passed to OpenSSL is set up with type void
7365 (*)(SSL *, int, int), as described somewhere. However, when calling the
7366 function (actually a macro) that sets it up, the type void(*)() is
7367 expected. I've put in a cast to prevent warnings from picky compilers.
7369 34. If a DNS black list lookup found a CNAME record, but there were no A
7370 records associated with the domain it pointed at, Exim crashed.
7372 35. If a DNS black list lookup returned more than one A record, Exim ignored
7373 all but the first. It now scans all returned addresses if a particular IP
7374 value is being sought. In this situation, the contents of the
7375 $dnslist_value variable are a list of all the addresses, separated by a
7378 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7379 to accept a host name and all its aliases if the forward lookup for any of
7380 them yielded the IP address of the incoming connection. Now it accepts only
7381 those names whose forward lookup yields the correct IP address. Any other
7382 names are discarded. This closes a loophole whereby a rogue DNS
7383 administrator could create reverse DNS records to break through a
7384 wildcarded host restriction in an ACL.
7386 37. If a user filter or a system filter that ran in a subprocess used any of
7387 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7388 the wrong values were passed to the pipe command ($thisaddress had the
7389 value of $0, $0 had the value of $1, etc). This bug was introduced by
7390 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7392 38. Improved the line breaking for long SMTP error messages from ACLs.
7393 Previously, if there was no break point between 40 and 75 characters, Exim
7394 left the rest of the message alone. Two changes have been made: (a) I've
7395 reduced the minimum length to 35 characters; (b) if it can't find a break
7396 point between 35 and 75 characters, it looks ahead and uses the first one
7397 that it finds. This may give the occasional overlong line, but at least the
7398 remaining text gets split now.
7400 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7401 file descriptors might be low, and that setting 1000 would always raise it.
7402 It turns out that in some environments, the limit is already over 1000 and
7403 that lowering it causes trouble. So now Exim takes care not to decrease it.
7405 40. When delivering a message, the value of $return_path is set to $sender_
7406 address at the start of routing (routers may change the value). By an
7407 oversight, this default was not being set up when an address was tested by
7408 -bt or -bv, which affected the outcome if any router or filter referred to
7411 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7412 log only when "message" or "log_message" are set. However, if one of the
7413 conditions was an address verification, or a call to a nested ACL, the
7414 messages generated by the underlying test were being passed through. This
7415 no longer happens. The underlying message is available in $acl_verify_
7416 message for both "message" and "log_message" expansions, so it can be
7417 passed through if needed.
7419 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7420 new expansion $bh_ to give the encoded byte string without charset
7421 translation. Translation happens only if iconv() is available; HAVE_ICONV
7422 indicates this at build time. HEADERS_CHARSET gives the charset to
7423 translate to; headers_charset can change it in the configuration, and
7424 "headers charset" can change it in an individual filter file.
7426 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7427 that creates RFC 2047 encoded "words" labels them as that charset instead
7428 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7429 expansion operator; (ii) when Exim creates a From: line for a local
7430 message; (iii) when a header line is rewritten to include a "phrase" part.
7432 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7433 buggy, causing it to skip the first lines of messages whose message ID
7434 ended in 'D'. This would not have bitten before Exim release 4.14, because
7435 message IDs were unlikely to end in 'D' before then. The effect was to have
7436 incorrect size information for certain domains.
7438 45. #include "config.h" was missing at the start of the crypt16.c module. This
7439 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7442 46. If there was a timeout during a "random" callout check, Exim treated it as
7443 a failure of the random address, and carried on sending RSET and the real
7444 address. If the delay was just some slowness somewhere, the response to the
7445 original RCPT would be taken as a response to RSET and so on, causing
7446 mayhem of various kinds.
7448 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7449 when I implemented it. It didn't allow for the fact that some option values
7450 may legitimately be negative (e.g. size_addition), and it didn't even do
7451 the right test for positive values.
7453 48. Domain names in DNS records are case-independent. Exim always looks them up
7454 in lower case. Some resolvers return domain names in exactly the case they
7455 appear in the zone file, that is, they may contain uppercase letters. Not
7456 all resolvers do this - some return always lower case. Exim was treating a
7457 change of case by a resolver as a change of domain, similar to a widening
7458 of a domain abbreviation. This triggered its re-routing code and so it was
7459 trying to route what was effectively the same domain again. This normally
7460 caused routing to fail (because the router wouldn't handle the domain
7461 twice). Now Exim checks for this case specially, and just changes the
7462 casing of the domain that it ultimately uses when it transmits the message
7465 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7468 50. If a filter generated a file delivery with a non-absolute name (possible if
7469 no home directory exists for the router), the forbid_file option was not
7472 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7473 the existing equality matching.
7475 52. Exim was using ints instead of ino_t variables in some places where it was
7476 dealing with inode numbers.
7478 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7479 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7480 called TMPDIR, and if it finds it is different, it changes its value.
7482 54. The smtp_printf() function is now made available to local_scan() so
7483 additional output lines can be written before returning. There is also an
7484 smtp_fflush() function to enable the detection of a dropped connection.
7485 The variables smtp_input and smtp_batched_input are exported to
7488 55. Changed the default runtime configuration: the message "Unknown user"
7489 has been removed from the ACL, and instead placed on the localuser router,
7490 using the cannot_route_message feature. This means that any verification
7491 failures that generate their own messages won't get overridden. Similarly,
7492 the "Unrouteable address" message that was in the ACL for unverifiable
7493 relay addresses has also been removed.
7495 56. Added hosts_avoid_esmtp to the smtp transport.
7497 57. The exicyclog script was not checking for the esoteric option
7498 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7499 will work only if exicyclog is run under the appropriate euid.
7501 58. Following a discussion on the list, the rules by which Exim recognises line
7502 endings on incoming messages have been changed. The -dropcr and drop_cr
7503 options are now no-ops, retained only for backwards compatibility. The
7504 following line terminators are recognized: LF CRLF CR. However, special
7505 processing applies to CR:
7507 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7508 nor a local message in the state where . is a terminator.
7510 (ii) If a bare CR is encountered in a header line, an extra space is added
7511 after the line terminator so as not to end the header. The reasoning
7512 behind this is that bare CRs in header lines are most likely either
7513 to be mistakes, or people trying to play silly games.
7515 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7516 was being incorrectly given as 18 bytes larger than it should have been.
7517 This is a VOB (very old bug).
7519 60. This may never have affected anything current, but just in case it has:
7520 When the local host is found other than at the start of a list of hosts,
7521 the local host, those with the same MX, and any that follow, are discarded.
7522 When the list in question was part of a longer list of hosts, the following
7523 hosts (not currently being processed) were also being discarded. This no
7524 longer happens. I'm not sure if this situation could ever has previously
7527 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7530 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7531 header. This is something that is recommended in a new Internet Draft, and
7532 is something that is documented as being done by Sendmail. There are two
7533 possible values. For messages generated by the autoreply transport, Exim
7536 Auto-Submitted: auto-replied
7538 whereas for all other generated messages (e.g. bounces) it adds
7540 Auto-Submitted: auto-generated
7542 63. The "personal" condition in filters now includes a test for the
7543 Auto-Submitted: header. If it contains the string "auto-" the message it
7544 not considered personal.
7546 64. Added rcpt_include_affixes as a generic transport option.
7548 65. Added queue_only_override (default true).
7550 66. Added the syslog_duplication option.
7552 67. If what should have been the first header line of a message consisted of
7553 a space followed by a colon, Exim was mis-interpreting it as a header line.
7554 It isn't of course - it is syntactically invalid and should therefore be
7555 treated as the start of the message body. The misbehaviour could have
7556 caused a number of strange effects, including loss of data in subsequent
7557 header lines, and spool format errors.
7559 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7560 client host had authenticated. This control can now be exercised by an ACL
7561 for more flexibility.
7563 69. By default, callouts do not happen when testing with -bh. There is now a
7564 variant, -bhc, which does actually run the callout code, including
7565 consulting and updating the callout cache.
7567 70. Added support for saslauthd authentication, courtesy of Alexander
7570 71. If statvfs() failed on the spool or log directories while checking their
7571 size for availability, Exim confusingly gave the error "space shortage".
7572 Furthermore, in debugging mode it crashed with a floating point exception.
7573 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7574 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7575 serious problem, Exim now writes to the main and panic logs when this
7576 happens, with details of the failure. It then refuses to accept the
7577 incoming message, giving the message "spool directory problem" or "log
7578 directory problem" with a 421 code for SMTP messages.
7580 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7581 0, 1, and 2 exist, because some OS complain for execs without them (see
7582 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7583 descriptors. However, the code omitted to check that the open succeeded,
7584 causing mysterious errors if for some reason the permissions on /dev/null
7585 got screwed. Now Exim writes a message to the main and panic logs, and
7586 bombs out if it can't open /dev/null.
7588 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7589 interact so that it is all more flexible. It is supposed to remain
7590 backwards compatible. Also added extra_local_interfaces.
7592 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7593 to bomb out with an assertion failure - to the client this appears as a
7594 connection drop. This problem occurs in the part of the code that was taken
7595 from the Samba project. Fortunately, the assertion is in a very simple
7596 function, so I have fixed this by reproducing the function inline in the
7597 one place where it is called, and arranging for authentication to fail
7598 instead of killing the process with assert().
7600 75. The SPA client code was not working when the server requested OEM rather
7601 than Unicode encoding.
7603 76. Added code to make require_files with a specific uid setting more usable in
7604 the case where statting the file as root fails - usually a non-root-mounted
7605 NFS file system. When this happens and the failure is EACCES, Exim now
7606 forks a subprocess and does the per-uid checking as the relevant uid.
7608 77. Added process_log_path.
7610 78. If log_file_path was not explicitly set, a setting of check_log_space or
7611 check_log_inodes was ignored.
7613 79. If a space check for the spool or log partitions fails, the incident is now
7614 logged. Of course, in the latter case the data may get lost...
7616 80. Added the %p formatting code to string_format() so that it can be used to
7617 print addresses in debug_print(). Adjusted all the address printing in the
7618 debugging in store.c to use %p rather than %d.
7620 81. There was a concern that a line of code in smtp_in.c could overflow a
7621 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7622 initially expressed, the concern was not well-founded, because trailing
7623 spaces are removed early. However, if the trailing spaces were followed by
7624 a NULL, they did not get removed, so the overflow was possible. Two fixes
7627 (a) I re-wrote the offending code in a cleaner fashion.
7628 (b) If an incoming SMTP command contains a NULL character, it is rejected
7631 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7632 runs initgroups(), so that if the Exim user is in any additional groups,
7633 they will be used during message reception.
7639 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob