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.
52 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
53 side, to decode a TLS 1.3 packet capture.
55 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
56 Previously the default library behaviour applied, sending two, each in
59 JH/03 Debug output for ACL now gives the config file name and line number for
62 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
64 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
66 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
67 buffer overrun for (non-chunking) other transports.
69 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
70 TLS1.3, means that a server rejecting a client certificate is not visible
71 to the client until the first read of encrypted data (typically the
72 response to EHLO). Add detection for that case and treat it as a failed
73 TLS connection attempt, so that the normal retry-in-clear can work (if
76 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
77 and/or domain. Found and fixed by Jason Betts.
79 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
80 configuration). If a CNAME target was not a wellformed name pattern, a
83 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
84 the OS reports them interleaved with other addresses.
86 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
87 used both for input and for a verify callout, both encrypted, SMTP
88 responses being sent by the server could be lost. This resulted in
89 dropped connections and sometimes bounces generated by a peer sending
92 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
93 string. Previously it was possible to cause undefined behaviour in a
94 library routine (usually a crash). Found by "zerons".
96 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
99 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
100 API was removed, so update to use the newer ones.
102 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
103 any timeout set, is taking a long time. Previously we would hang on to a
104 rotated logfile "forever" if the input was arriving with long gaps
105 (a previous attempt to fix addressed lack, for a long time, of initial
108 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
109 shared (NFS) environment. The length of the tempfile name is now
110 4 + 16 ("hdr.$message_exim_id") which might break on file
111 systems which restrict the file name length to lower values.
114 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
115 shared (NFS) environment.
117 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
118 did for all versions <4.90). Notably -M, -m, --invert, -I may be
121 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
122 on some platforms for bit 31.
124 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
125 to changes apparently associated with TLS1.3 handling some of the APIs
126 previously used were either nonfunctional or inappropriate. Strings
127 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
128 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
129 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
130 This affects log line X= elements, the $tls_{in,out}_cipher variables,
131 and the use of specific cipher names in the encrypted= ACL condition.
133 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
135 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
136 verification result was not updated unless hosts_require_ocsp applied.
138 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
139 queue_list_requires_admin set to false, non-admin users were denied the
142 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
143 directory-of-certs mode. Previously they were advertised despite the
146 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
147 A single TCP connection by a client will now hold a TLS connection open
148 for multiple message deliveries, by default. Previoud the default was to
151 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
152 default. If built with the facility, DANE will be used. The facility
153 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
155 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
156 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
157 must be defined and you must still, unless you define DISABLE_TLS, manage
158 the the include-dir and library-file requirements that go with that
159 choice. Non-TLS builds are still supported.
161 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
164 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
165 default. If the platform supports and has the facility enabled, it will
166 be requested on all coneections.
168 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
169 controlled by the build-time option SUPPORT_PIPE_CONNECT.
171 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
173 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
174 success-DSN messages. Previously the From: header was always the default
175 one for these; the option was ignored.
177 JH/28 Fix the timeout on smtp response to apply to the whole response.
178 Previously it was reset for every read, so a teergrubing peer sending
179 single bytes within the time limit could extend the connection for a
180 long time. Credit to Qualsys Security Advisory Team for the discovery.
182 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
183 delivery address, which leaked information of the results of local
184 forwarding. Change to the original envelope recipient address, per
187 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
188 requested. Previously not bounce was generated and a log entry of
189 error ignored was made.
191 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
193 JH/32 Introduce a general tainting mechanism for values read from the input
194 channel, and values derived from them. Refuse to expand any tainted
195 values, to catch one form of exploit.
197 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
198 was unused and the unexpanded text used for the test. Found and
199 fixed by Ruben Jenster.
201 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
202 an attempt to use a TLS library read routine dereffed a nul pointer,
205 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
206 them in our smtp response.
208 JH/36 Have the general router option retry_use_local_part default to true when
209 any of the restrictive preconditions are set (to anything). Previously it
210 was only for check_local user. The change removes one item of manual
211 configuration which is required for proper retries when a remote router
212 handles a subset of addresses for a domain.
214 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
215 link count into consideration.
217 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
218 caused the extension of big_buffer, the following lines were ignored.
220 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
221 accordance with RFC 2308. Previously there was no expiry, so a longlived
222 receive process (eg. due to ACL delays) versus a short SOA value could
225 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
227 JH/39 Promote DMARC support to mainline.
229 JH/40 Bug 2452: Add a References: header to DSNs.
231 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
232 parameters. The relevant library call is documented as "Deprecated: This
233 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
234 3.6.0, DH parameters are negotiated following RFC7919."
236 HS/06 Change the default of dnssec_request_domains to "*"
238 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
239 carried on and emitted a BDAT command, even when PIPELINING was not
242 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
243 buffer was used for the filename, resulting in a trap when tainted
244 arguments (eg. $domain) were used.
246 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
247 recommended to avoid a possible server-load attack. The feature can be
248 re-enabled via the openssl_options main cofiguration option.
250 JH/45 local_scan API: documented the current smtp_printf() call. This changed
251 for version 4.90 - adding a "more data" boolean to the arguments.
252 Bumped the ABI version number also, this having been missed previously;
253 release versions 4.90 to 4.92.3 inclusive were effectively broken in
254 respect of usage of smtp_printf() by either local_scan code or libraries
255 accessed via the ${dlfunc } expansion item. Both will need coding
256 adjustment for any calls to smtp_printf() to match the new function
257 signature; a FALSE value for the new argument is always safe.
259 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
260 the file-offset (which the Linux syscall does, and exim expects); this
261 resulted in an indefinite loop.
263 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
264 to do ARC verification. The Authentication-Results: header line added
265 by the configuration then had no ARC item.
271 JH/01 Remove code calling the customisable local_scan function, unless a new
272 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
274 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
275 non-signal-safe functions being used.
277 JH/03 Bug 2269: When presented with a received message having a stupidly large
278 number of DKIM-Signature headers, disable DKIM verification to avoid
279 a resource-consumption attack. The limit is set at twenty.
281 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
282 report of oldest_pass in ${authres } in consequence, and separate out
283 some descriptions of reasons for verification fail.
285 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
286 files in the spool were present and unlocked. A queue-runner could spot
287 them, resulting in a duplicate delivery. Fix that by doing the unlock
288 after the unlink. Investigation by Tim Stewart. Take the opportunity to
289 add more error-checking on spoolfile handling while that code is being
292 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
293 No known attacks, no CVE, this is defensive hardening.
295 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
296 a queue-runner could start a delivery while other operations were ongoing.
297 Cutthrough delivery was a common victim, resulting in duplicate delivery.
298 Found and investigated by Tim Stewart. Fix by using the open message data
299 file handle rather than opening another, and not locally closing it (which
300 releases a lock) for that case, while creating the temporary .eml format
301 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
303 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
304 $sender_verify_failure/$recipient_verify_failure to "random".
306 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
309 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
310 Previously this would segfault.
312 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
315 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
316 like zero, since the resolver should be doing this for us, But we need one
317 as a CNAME but no MX presence gets the CNAME returned; we need to check
318 that doesn't point to an MX to declare it "no MX returned" rather than
319 "error, loop". A new main option is added so the older capability of
320 following some limited number of chain links is maintained.
322 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
324 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
325 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
326 now-deprecated earlier definitions used only specified the range up to TLS
327 1.2 (in the older-version library docs).
329 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
331 JH/15 Rework TLS client-side context management. Stop using a global, and
332 explicitly pass a context around. This enables future use of TLS for
333 connections to service-daemons (eg. malware scanning) while a client smtp
334 connection is using TLS; with cutthrough connections this is quite likely.
336 JH/16 Fix ARC verification to do AS checks in reverse order.
338 JH/17 Support a "tls" option on the ${readsocket } expansion item.
340 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
341 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
342 Previously the "utf8" would be re-prepended for every additional message.
344 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
345 Previously thery were accepted, resulting in issues when attempting to
346 forward messages to a non-supporting MTA.
348 PP/02 Let -n work with printing macros too, not just options.
350 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
351 one parent address was copied, and bogus data was used at delivery-logging
352 time. Either a crash (after delivery) or bogus log data could result.
353 Discovery and analysis by Tim Stewart.
355 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
356 Previously if the string ended mid-character, we did not insert the
357 promised '?' replacement.
359 PP/04 Documentation: current string operators work on bytes, not codepoints.
361 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
362 should pack well giving a smaller memory footprint so better caching and
363 therefore performance. Group the declarations where this can't be done so
364 that the byte-sized flag variables are not interspersed among pointer
365 variables, giving a better chance of good packing by the compiler.
367 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
368 non-null, to avoid issues with sites running BATV. Previously reports were
369 sent with an empty envelope sender so looked like bounces.
371 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
372 The ignore_error flag wasn't being returned from the filter subprocess so
373 was not set for later routers. Investigation and fix by Matthias Kurz.
375 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
376 and a msg:complete for the whole, when a message is manually removed using
377 -Mrm. Developement by Matthias Kurz, hacked on by JH.
379 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
380 a "Gnu special" function, asprintf() in the DB utility binary builds; I
381 hope that is portable enough.
383 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
384 requiring a known-CA anchor certificate; make it now rely entirely on the
385 TLSA as an anchor. Checking the name on the leaf cert against the name
386 on the A-record for the host is still done for TA (but not for EE mode).
388 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
389 included in delivery lines for non-proxied connections, when compiled with
390 SUPPORT_SOCKS and running with proxy logging enabled.
392 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
393 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
394 move the existing event to fire before the normal logging of message
395 failure so that custom logging is bracketed by normal logging.
397 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
398 msg:fail:internal event. Developement by Matthias Kurz.
400 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
401 far too small for todays use of crypto signatures stored there. Go all
402 the way to the max DNS message size of 64kB, even though this might be
403 overmuch for IOT constrained device use.
405 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
406 copy a string over itself. The library routine is documented as not
407 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
409 JH/32 For main options check_spool_space and check_inode_space, where the
410 platform supports 64b integers, support more than the previous 2^31 kB
411 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
412 the previous G, M, k.
414 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
415 $authenticated_fail_id variable on authentication failure. Previously
418 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
419 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
420 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
421 GNUTLS_SEC_PARAM_MEDIUM.
423 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
424 side. Previously we would continue as if no SNI had been received.
426 JH/36 Harden the handling of string-lists. When a list consisted of a sole
427 "<" character, which should be a list-separator specification, we walked
428 off past the nul-terimation.
430 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
431 causes) even when the retry time is not yet met. Previously they were
432 not, meaning that when (say) an account was over-quota and temp-rejecting,
433 and multiple senders' messages were queued, only one sender would get
434 notified on each configured delay_warning cycle.
436 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
438 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
439 cipher-suites, an error can be left on the stack even for a succeeding
440 accept; this results in impossible error messages when a later operation
443 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
444 return error codes indicating retry. Under TLS1.3 this becomes required.
446 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
447 it only wrote the new authenticators, resulting in a lack of tracking of
448 peer changes of ESMTP extensions until the next cache flush.
450 JH/41 Fix the loop reading a message header line to check for integer overflow,
451 and more-often against header_maxsize. Previously a crafted message could
452 induce a crash of the recive process; now the message is cleanly rejected.
454 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
455 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
461 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
462 When redis_servers is set to a list of > 1 element, and the Redis servers
463 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
464 case of MOVED into a DEFER case instead, thus moving the query onto the
465 next server in the list. For a cluster of N elements, all N servers must
466 be defined in redis_servers.
468 GF/02 Catch and remove uninitialized value warning in exiqsumm
469 Check for existence of @ARGV before looking at $ARGV[0]
471 JH/01 Replace the store_release() internal interface with store_newblock(),
472 which internalises the check required to safely use the old one, plus
473 the allocate and data copy operations duplicated in both (!) of the
474 extant use locations.
476 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
477 modifier. This matches the restriction on the commandline.
479 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
480 Previously only the last row was returned.
482 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
483 we assumed that tags in the header were well-formed, and parsed the
484 element content after inspecting only the first char of the tag.
485 Assumptions at that stage could crash the receive process on malformed
488 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
489 While running the DKIM ACL we operate on the Permanent memory pool so that
490 variables created with "set" persist to the DATA ACL. Also (at any time)
491 DNS lookups that fail create cache records using the Permanent pool. But
492 expansions release any allocations made on the current pool - so a dnsdb
493 lookup expansion done in the DKIM ACL releases the memory used for the
494 DNS negative-cache, and bad things result. Solution is to switch to the
495 Main pool for expansions.
496 While we're in that code, add checks on the DNS cache during store_reset,
497 active in the testsuite.
498 Problem spotted, and debugging aided, by Wolfgang Breyha.
500 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
501 When none of the hosts presented to a transport match an already-open
502 connection, close it and proceed with the list. Previously we would
503 queue the message. Spotted by Lena with Yahoo, probably involving
506 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
507 Previously a spurious "250 OK id=" response was appended to the proper
510 JH/08 The "support for" informational output now, which built with Content
511 Scanning support, has a line for the malware scanner interfaces compiled
512 in. Interface can be individually included or not at build time.
514 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
515 by the template makefile "src/EDITME". The "STREAM" support for an older
516 ClamAV interface method is removed.
518 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
519 rows affected is given instead).
521 JH/11 The runtime Berkeley DB library version is now additionally output by
522 "exim -d -bV". Previously only the compile-time version was shown.
524 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
525 SMTP connection. Previously, when one had more receipients than the
526 first, an abortive onward connection was made. Move to full support for
527 multiple onward connections in sequence, handling cutthrough connection
528 for all multi-message initiating connections.
530 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
531 routers. Previously, a multi-recipient message would fail to match the
532 onward-connection opened for the first recipient, and cause its closure.
534 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
535 a timeout on read on a GnuTLS initiating connection, resulting in the
536 initiating connection being dropped. This mattered most when the callout
537 was marked defer_ok. Fix to keep the two timeout-detection methods
540 JH/15 Relax results from ACL control request to enable cutthrough, in
541 unsupported situations, from error to silently (except under debug)
542 ignoring. This covers use with PRDR, frozen messages, queue-only and
545 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
547 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
548 metadata, resulting in a crash in free().
550 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
551 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
552 Broken also in d185889f4, with init system revamp.
554 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
555 Previously we abruptly closed the connection after reading a malware-
556 found indication; now we go on to read the "scan ok" response line,
559 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
560 ACL. Previously, a crash would result.
562 JH/19 Speed up macro lookups during configuration file read, by skipping non-
563 macro text after a replacement (previously it was only once per line) and
564 by skipping builtin macros when searching for an uppercase lead character.
566 JH/20 DANE support moved from Experimental to mainline. The Makefile control
567 for the build is renamed.
569 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
570 was allocated for every new TLS startup, meaning one per message. Fix
571 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
573 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
574 reported the original. Fix to report (as far as possible) the ACL
575 result replacing the original.
577 JH/23 Fix memory leak during multi-message connections using STARTTLS under
578 OpenSSL. Certificate information is loaded for every new TLS startup,
579 and the resources needed to be freed.
581 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
583 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
586 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
587 DATA response info to the (existing) per-recipient response info for
588 the "C=" log element. It can have useful tracking info from the
589 destination system. Patch from Simon Arlott.
591 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
592 length value. Previously this would segfault.
594 HS/02 Support Avast multiline protoocol, this allows passing flags to
595 newer versions of the scanner.
597 JH/28 Ensure that variables possibly set during message acceptance are marked
598 dead before release of memory in the daemon loop. This stops complaints
599 about them when the debug_store option is enabled. Discovered specifically
600 for sender_rate_period, but applies to a whole set of variables.
601 Do the same for the queue-runner and queue-list loops, for variables set
602 from spool message files. Do the same for the SMTP per-message loop, for
603 certain variables indirectly set in ACL operations.
605 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
606 as a multi-recipient message from a mailinglist manager). The coding had
607 an arbitrary cutoff number of characters while checking for more input;
608 enforced by writing a NUL into the buffer. This corrupted long / fast
609 input. The problem was exposed more widely when more pipelineing of SMTP
610 responses was introduced, and one Exim system was feeding another.
611 The symptom is log complaints of SMTP syntax error (NUL chars) on the
612 receiving system, and refused recipients seen by the sending system
613 (propating to people being dropped from mailing lists).
614 Discovered and pinpointed by David Carter.
616 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
617 replaced by the ${authresults } expansion.
619 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
621 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
622 allows proper process termination in container environments.
624 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
625 Previously the "final dot" had a newline after it; ensure it is CR,LF.
627 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
628 and "err_perm", deprecated since 4.83 when the RFC-defined words
629 "temperror" and "permerror" were introduced.
631 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
632 transport-filters or DKIM-signing. The restriction was lost in the
633 consolidation of verify-callout and delivery SMTP handling.
634 Extend the restriction to also cover ARC-signing.
636 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
637 in defer=pass mode supply a 450 to the initiator. Previously the message
640 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
641 tls_require_ciphers is used as before.
643 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
644 "pass_unscanned". Only tmpfails from the scanner are written to
645 the paniclog, as they may require admin intervention (permission
646 denied, license issues). Other scanner errors (like decompression
647 bombs) do not cause a paniclog entry.
649 JH/36 Fix reinitialisation of DKIM logging variable between messages.
650 Previously it was possible to log spurious information in receive log
653 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
654 triggered odd behaviour from Outlook Express clients.
656 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
659 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
660 since the IETF WG has not yet settled on that versus the original
661 "bare" representation.
663 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
664 Previously the millisecond value corrupted the output.
665 Fix also for syslog_pid=no and log_selector +pid, for which the pid
666 corrupted the output.
672 JH/01 Rework error string handling in TLS interface so that the caller in
673 more cases is responsible for logging. This permits library-sourced
674 string to be attached to addresses during delivery, and collapses
675 pairs of long lines into single ones.
677 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
678 during configuration. Wildcards are allowed and expanded.
680 JH/02 Rework error string handling in DKIM to pass more info back to callers.
681 This permits better logging.
683 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
684 do not close it down and have the child transport start it up again on
685 the passed-on TCP connection. Instead, proxy the child (and any
686 subsequent ones) for TLS via a unix-domain socket channel. Logging is
687 affected: the continued delivery log lines do not have any DNSSEC, TLS
688 Certificate or OCSP information. TLS cipher information is still logged.
690 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
691 identical IP addresses on different listening ports. Will also affect
694 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
695 add noisy ifdef guards to special-case this sillyness.
696 Patch from Bernd Kuhls.
698 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
699 than 255 are no longer allowed.
701 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
702 Disable the session-cache too, which might reduce our load. Since we
703 currrectly use a new context for every connection, both as server and
704 client, there is no benefit for these.
705 GnuTLS appears to not support tickets server-side by default (we don't
706 call gnutls_session_ticket_enable_server()) but client side is enabled
707 by default on recent versions (3.1.3 +) unless the PFS priority string
710 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
711 <https://reproducible-builds.org/specs/source-date-epoch/>.
713 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
714 the check for any unsuccessful recipients did not notice the limit, and
715 erroneously found still-pending ones.
717 JH/08 Pipeline CHUNKING command and data together, on kernels that support
718 MSG_MORE. Only in-clear (not on TLS connections).
720 JH/09 Avoid using a temporary file during transport using dkim. Unless a
721 transport-filter is involved we can buffer the headers in memory for
722 creating the signature, and read the spool data file once for the
723 signature and again for transmission.
725 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
726 4.77 as the kernel support then wasn't solid, having issues in 64bit
727 mode. Now, it's been long enough. Add support for FreeBSD also.
729 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
730 case where the routing stage had gathered several addresses to send to
731 a host before calling the transport for the first, we previously failed
732 to close down TLS in the old transport process before passing the TCP
733 connection to the new process. The new one sent a STARTTLS command
734 which naturally failed, giving a failed delivery and bloating the retry
735 database. Investigation and fix prototype from Wolfgang Breyha.
737 JH/12 Fix check on SMTP command input synchronisation. Previously there were
738 false-negatives in the check that the sender had not preempted a response
739 or prompt from Exim (running as a server), due to that code's lack of
740 awareness of the SMTP input buffering.
742 PP/04 Add commandline_checks_require_admin option.
743 Exim drops privileges sanely, various checks such as -be aren't a
744 security problem, as long as you trust local users with access to their
745 own account. When invoked by services which pass untrusted data to
746 Exim, this might be an issue. Set this option in main configuration
747 AND make fixes to the calling application, such as using `--` to stop
750 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
751 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
752 responses to those, into a single TLS record each way (this usually means
753 a single packet). As a side issue, smtp_enforce_sync now works on TLS
756 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
757 affects you only if you're dancing at the edge of the param size limits.
758 If you are, and this message makes sense to you, then: raise the
759 configured limit or use OpenSSL 1.1. Nothing we can do for older
762 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
763 cmdline element to get the documented default one. Previously it was
766 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
767 are unsuable for non-message-specific reasons - eg. connection timeout,
768 banner-time rejection.
770 JH/16 Fix logging of delivery remote port, when specified by router, under
773 PP/06 Repair manualroute's ability to take options in any order, even if one
774 is the name of a transport.
777 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
779 JH/17 Change the list-building routines interface to use the expanding-string
780 triplet model, for better allocation and copying behaviour.
782 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
783 Previously it was constructed the first time a possibly-matching string
784 was met in the configuration file input during startup; now it is done
787 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
788 compatible one, to avoid the (poorly documented) possibility of a config
789 file in the working directory redirecting the DB files, possibly correpting
790 some existing file. CVE-2017-10140 assigned for BDB.
792 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
793 cache-hot. Previously, although the result was properly cached, the
794 initial verify call returned a defer.
796 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
797 the main verify for receipient in uncached-mode.
799 JH/22 Retire historical build files to an "unsupported" subdir. These are
800 defined as "ones for which we have no current evidence of testing".
802 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
803 if present. Previously it was ignored.
805 JH/24 Start using specified-initialisers in C structure init coding. This is
806 a C99 feature (it's 2017, so now considered safe).
808 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
809 if was a fixed-sized field and bitmask ops via macros; it is now more
812 PP/07 GitHub PR 56: Apply MariaDB build fix.
813 Patch provided by Jaroslav Škarvada.
815 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
816 during Coverity cleanups [4.87 JH/47]
817 Diagnosis and fix provided by Michael Fischer v. Mollard.
819 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
820 the right size to place the terminating semicolon on its own folded
821 line, the header hash was calculated to an incorrect value thanks to
822 the (relaxed) space the fold became.
824 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
825 and confused the parent.
827 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
828 which could crash as a result. This could lead to undeliverable messages.
830 JH/28 Logging: "next input sent too soon" now shows where input was truncated
833 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
834 matters on fast-turnover and PID-randomising systems, which were getting
835 out-of-order delivery.
837 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
838 a possibly-overlapping copy. The symptom was that "Remote host closed
839 connection in response to HELO" was logged instead of the actual 4xx
842 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
843 Previously only that bufferd was discarded, resulting in SYMTP command
846 JH/32 DKIM: when a message has multiple signatures matching an identity given
847 in dkim_verify_signers, run the dkim acl once for each. Previously only
848 one run was done. Bug 2189.
850 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
851 "panic the current process" to "deliberately defer". The panic log is
852 still written with the problem list name; the mail and reject logs now
853 get a temp-reject line for the message that was being handled, saying
854 something like "domains check lookup or other defer". The SMTP 451
855 message is still "Temporary local problem".
857 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
858 A crafted sequence of BDAT commands could result in in-use memory beeing
859 freed. CVE-2017-16943.
861 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
862 from SMTP input. Previously it was always done; now only done for DATA
863 and not BDAT commands. CVE-2017-16944.
865 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
866 to the message (such as an overlong header line). Previously this was
867 not done and we did not exit BDAT mode. Followon from the previous item
868 though a different problem.
874 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
875 than -2003 did; needs libidn2 in addition to libidn.
877 JH/02 The path option on a pipe transport is now expanded before use.
879 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
880 Patch provided by "Björn", documentation fix added too.
882 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
883 missing a wire-to-host endian conversion.
885 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
886 close after a BDAT command line could be taken as a following command,
887 giving a synch failure. Fix by only checking for synch immediately
888 before acknowledging the chunk.
890 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
891 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
892 macro. Patches provided by Josh Soref.
894 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
895 Previously we did not; the RFC seems ambiguous and VRFY is not listed
896 by IANA as a service extension. However, John Klensin suggests that we
899 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
900 the dkim code may be unix-mode line endings rather than smtp wire-format
901 CRLF, so prepend a CR to any bare LF.
903 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
904 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
906 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
907 the first were themselves being wrongly included in the feed into dkim
908 processing; with most chunk sizes in use this resulted in an incorrect
909 body hash calculated value.
911 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
912 DKIM signature block, for verification. Although advised against by
913 standards it is specifically not ruled illegal.
915 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
917 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
918 missing a body hash (the bh= tag).
920 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
921 It seems that HAProxy sends the Proxy Protocol information in clear and
922 only then does a TLS startup, so do the same.
924 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
925 TCP connections (such as for Spamd) unless the daemon successfully set
926 Fast Open mode on its listening sockets. This fixes breakage seen on
927 too-old kernels or those not configured for Fast Open, at the cost of
928 requiring both directions being enabled for TFO, and TFO never being used
929 by non-daemon-related Exim processes.
931 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
932 endings, at least on the first header line. Try to canonify any that get
933 past that check, despite the cost.
935 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
936 now limited to an arbitrary five deep, while parsing addresses with the
937 strip_excess_angle_brackets option enabled.
939 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
940 instead leave the unprompted TLS handshake in socket buffer for the
941 TLS library to consume.
943 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
945 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
947 JH/16 Drop variables when they go out of scope. Memory management drops a whole
948 region in one operation, for speed, and this leaves assigned pointers
949 dangling. Add checks run only under the testsuite which checks all
950 variables at a store-reset and panics on a dangling pointer; add code
951 explicitly nulling out all the variables discovered. Fixes one known
952 bug: a transport crash, where a dangling pointer for $sending_ip_address
953 originally assigned in a verify callout, is re-used.
955 PP/06 Drop '.' from @INC in various Perl scripts.
957 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
959 PP/08 Reduce a number of compilation warnings under clang; building with
960 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
961 should be warning-free.
963 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
965 HS/01 Fix portability problems introduced by PP/08 for platforms where
966 realloc(NULL) is not equivalent to malloc() [SunOS et al].
968 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
969 chunk. This allows us to accept broken chunked messages. We need a more
970 general solution here.
972 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
973 already-broken messages in the queue.
975 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
977 JH/19 Fix reference counting bug in routing-generated-address tracking.
983 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
984 supports it and a size is available (ie. the sending peer gave us one).
986 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
987 years of being deprecated). The replacements are the ACLs
988 acl_smtp_mime and acl_not_smtp_mime.
990 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
991 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
992 or even an in-clear connection were permitted. Now, if the host lookup
993 was dnssec and dane was requested then the host is only used if the
994 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
995 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
996 if one fails this test.
997 This means that a poorly-configured remote DNS will make it incommunicado;
998 but it protects against a DNS-interception attack on it.
1000 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1001 non-noisy when a race steals the message being considered.
1003 JH/05 If main configuration option tls_certificate is unset, generate a
1004 self-signed certificate for inbound TLS connections.
1006 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1007 in rewrites and routers.
1009 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1010 and logged a warning sing 4.83; now they are a configuration file error.
1012 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1013 (lacking @domain). Apply the same qualification processing as RCPT.
1015 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1017 JH/10 Support ${sha256:} applied to a string (as well as the previous
1020 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1021 a cutthrough deliver is pending, as we always want to make a connection.
1022 This also avoids re-routing the message when later placing the cutthrough
1023 connection after a verify cache hit.
1024 Do not update it with the verify result either.
1026 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1027 when routing results in more than one destination address.
1029 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1030 signing (which inhibits the cutthrough capability). Previously only
1031 the presence of an option was tested; now an expansion evaluating as
1032 empty is permissible (obviously it should depend only on data available
1033 when the cutthrough connection is made).
1035 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1036 the relevant preceding SMTP command did not note the pipelining mode.
1038 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1039 Previously they were not counted.
1041 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1042 as one having no matching records. Previously we deferred the message
1043 that needed the lookup.
1045 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1046 distinguished as "(=".
1048 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1049 for missing MX records. Previously it only worked for missing A records.
1051 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1053 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1054 after the data-go-ahead and data-ack. Patch from Jason Betts.
1056 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1057 even for a "none" policy. Patch from Tony Meyer.
1059 JH/22 Fix continued use of a connection for further deliveries. If a port was
1060 specified by a router, it must also match for the delivery to be
1063 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1064 When one of the recipients of a message was unsuitable for the connection
1065 (has no matching addresses), we lost track of needing to mark it
1066 deferred. As a result mail would be lost.
1068 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1070 JH/25 Decoding ACL controls is now done using a binary search; the source code
1071 takes up less space and should be simpler to maintain. Merge the ACL
1072 condition decode tables also, with similar effect.
1074 JH/26 Fix problem with one_time used on a redirect router which returned the
1075 parent address unchanged. A retry would see the parent address marked as
1076 delivered, so not attempt the (identical) child. As a result mail would
1079 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1080 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1081 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1084 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1085 Main config options check_{log,spool}_{inodes,space} are now
1086 100 inodes, 10MB unless set otherwise in the configuration.
1088 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1089 Previously it only applied to the main-section connection policy
1092 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1094 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1095 by me. Added RFC7919 DH primes as an alternative.
1097 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1098 are not in the system include path.
1100 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1101 GnuTLS, when a session startup failed (eg because the client disconnected)
1102 Exim did stdio operations after fclose. This was exposed by a recent
1103 change which nulled out the file handle after the fclose.
1105 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1106 signed directly by the cert-signing cert, rather than an intermediate
1107 OCSP-signing cert. This is the model used by LetsEncrypt.
1109 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1111 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1112 an incoming connection.
1114 HS/02 Bug 1802: Do not half-close the connection after sending a request
1117 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1118 fallback to "prime256v1".
1120 JH/34 SECURITY: Use proper copy of DATA command in error message.
1121 Could leak key material. Remotely exploitable. CVE-2016-9963.
1127 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1128 and 3.4.4 - once the server is enabled to respond to an OCSP request
1129 it does even when not requested, resulting in a stapling non-aware
1130 client dropping the TLS connection.
1132 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1133 support variable-length bit vectors. No functional change.
1135 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1136 The I= interface field on outgoing lines is now after the H= remote
1137 host field, same as incoming lines. There is a separate
1138 outgoing_interface log selector which allows you to disable the
1141 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1142 If not running log_selector +smtp_connection the mainlog would be held
1143 open indefinitely after a "too many connections" event, including to a
1144 deleted file after a log rotate. Leave the per net connection logging
1145 leaving it open for efficiency as that will be quickly detected by the
1146 check on the next write.
1148 HS/01 Bug 1671: Fix post transport crash.
1149 Processing the wait-<transport> messages could crash the delivery
1150 process if the message IDs didn't exist for some reason. When
1151 using 'split_spool_directory=yes' the construction of the spool
1152 file name failed already, exposing the same netto behaviour.
1154 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1155 mime_regex ACL conditions.
1157 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1158 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1159 to HELO, local diagnostic string.
1161 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1162 log line to debug. Even when configured with a tls authenticator many
1163 client connections are expected to not authenticate in this way, so
1164 an authenticate fail is not an error.
1166 HS/02 Add the Exim version string to the process info. This way exiwhat
1167 gives some more detail about the running daemon.
1169 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1170 matter for fast-change records such as DNSBLs.
1172 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1173 retry record, even if constant. There may be multiple transports with
1174 different interface settings and the retry behaviour needs to be kept
1177 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1179 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1181 JH/10 Bug 840: fix log_defer_output option of pipe transport
1183 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1184 hosts_randomize. This matters a lot when combined with mua_wrapper.
1186 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1187 ${quote_pgsql:<string>} operator.
1189 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1191 JH/14 Change header file name for freeradius-client. Relevant if compiling
1192 with Radius support; from the Gentoo tree and checked under Fedora.
1194 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1196 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1197 option transport_filter, meaning no filtering.
1199 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1201 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1202 defaults to "*" (all hosts). The variable is now available when not built
1203 with TLS, default unset, mainly to enable keeping the testsuite sane.
1204 If a server certificate is not supplied (via tls_certificate) an error is
1205 logged, and clients will find TLS connections fail on startup. Presumably
1206 they will retry in-clear.
1207 Packagers of Exim are strongly encouraged to create a server certificate
1208 at installation time.
1210 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1211 with the $config_file variable.
1213 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1214 in transport context, after the attempt, and per-recipient. The latter type
1215 is per host attempted. The event data is the error message, and the errno
1216 information encodes the lookup type (A vs. MX) used for the (first) host,
1217 and the trailing two digits of the smtp 4xx response.
1219 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1220 to write to mainlog (or rejectlog, paniclog) in the window between file
1221 creation and permissions/ownership being changed. Particularly affects
1222 installations where exicyclog is run as root, rather than exim user;
1223 result is that the running daemon panics and dies.
1225 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1227 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1228 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1229 "pri" and "weight". Note that the previous implicit priority given by the
1230 list order is no longer honoured.
1232 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1233 for DKIM processing.
1235 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1236 by defining SUPPORT_SOCKS.
1238 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1239 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1240 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1241 variables are renamed to proxy_{local,external}_{address,port}.
1243 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1244 for a build by defining SUPPORT_I18N
1246 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1247 of the query string, and make ${quote_redis:} do that quoting.
1249 JH/29 Move Events support from Experimental to mainline, enabled by default
1250 and removable for a build by defining DISABLE_EVENT.
1252 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1254 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1255 cached by the daemon.
1257 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1258 by defining LOOKUP_REDIS. The libhiredis library is required.
1260 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1261 keys are given for lookup.
1263 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1264 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1265 only supported when built with TLS support. The PolarSSL SHA routines
1266 are still used when the TLS library is too old for convenient support.
1268 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1269 openssl_options), for security. OpenSSL forces this from version 1.1.0
1270 server-side so match that on older versions.
1272 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1273 allocation for $value could be released as the expansion processing
1274 concluded, but leaving the global pointer active for it.
1276 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1277 and to use the domains and local_parts ACL conditions.
1279 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1280 incorrectly not doubled on cutthrough transmission, hence seen as a
1281 body-termination at the receiving system - resulting in truncated mails.
1282 Commonly the sender saw a TCP-level error, and retransmitted the message
1283 via the normal store-and-forward channel. This could result in duplicates
1284 received - but deduplicating mailstores were liable to retain only the
1285 initial truncated version.
1287 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1289 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1291 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1292 we're in there, support oversigning also; bug 1309.
1294 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1296 HS/04 Add support for keep_environment and add_environment options.
1298 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1299 either intentional arithmetic overflow during PRNG, or testing config-
1302 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1303 delivery resulted in actual delivery. Cancel cutthrough before DATA
1306 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1307 rejects a recipient: pass the reject to the originator.
1309 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1310 Many were false-positives and ignorable, but it's worth fixing the
1313 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1314 for the new environment-manipulation done at startup. Move the routines
1315 from being local to tls.c to being global via the os.c file.
1317 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1318 an extract embedded as result-arg for a map, the first arg for extract
1319 is unavailable so we cannot tell if this is a numbered or keyed
1320 extraction. Accept either.
1326 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1329 JH/02 The smtp transport option "multi_domain" is now expanded.
1331 JH/03 The smtp transport now requests PRDR by default, if the server offers
1334 JH/04 Certificate name checking on server certificates, when exim is a client,
1335 is now done by default. The transport option tls_verify_cert_hostnames
1336 can be used to disable this per-host. The build option
1337 EXPERIMENTAL_CERTNAMES is withdrawn.
1339 JH/05 The value of the tls_verify_certificates smtp transport and main options
1340 default to the word "system" to access the system default CA bundle.
1341 For GnuTLS, only version 3.0.20 or later.
1343 JH/06 Verification of the server certificate for a TLS connection is now tried
1344 (but not required) by default. The verification status is now logged by
1345 default, for both outbound TLS and client-certificate supplying inbound
1348 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1351 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1352 Status Notification (bounce) messages are now MIME format per RFC 3464.
1353 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1354 under the control of the dsn_advertise_hosts option, and routers may
1355 have a dsn_lasthop option.
1357 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1358 default, modifiable by a malware= option. The list separator for
1359 the options can now be changed in the usual way. Bug 68.
1361 JH/10 The smtp_receive_timeout main option is now expanded before use.
1363 JH/11 The incoming_interface log option now also enables logging of the
1364 local interface on delivery outgoing connections.
1366 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1367 if the interface and destination host and port all match.
1369 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1372 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1373 Patch from Andrew Lewis.
1375 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1376 now supports optional time-restrictions, weighting, and priority
1377 modifiers per server. Patch originally by <rommer@active.by>.
1379 JH/16 The spamd_address main option now supports a mixed list of local
1380 and remote servers. Remote servers can be IPv6 addresses, and
1381 specify a port-range.
1383 JH/17 Bug 68: The spamd_address main option now supports an optional
1384 timeout value per server.
1386 JH/18 Bug 1581: Router and transport options headers_add/remove can
1387 now have the list separator specified.
1389 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1392 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1395 JH/21 Support for the A6 type of dns record is withdrawn.
1397 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1398 rather than the verbs used.
1400 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1401 from 255 to 1024 chars.
1403 JH/24 Verification callouts now attempt to use TLS by default.
1405 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1406 are generic router options now. The defaults didn't change.
1408 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1409 Original patch from Alexander Shikoff, worked over by JH.
1411 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1412 files if ZCAT_COMMAND is not executable.
1414 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1416 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1418 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1419 Normally benign, it bites when the pair was led to by a CNAME;
1420 modern usage is to not canonicalize the domain to a CNAME target
1421 (and we were inconsistent anyway for A-only vs AAAA+A).
1423 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1425 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1426 when evaluating $sender_host_dnssec.
1428 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1429 $sender_helo_dnssec variable.
1431 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1433 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1435 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1437 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1438 documented as working, but never had. Support all but $spam_report.
1440 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1441 added for tls authenticator.
1443 HS/03 Add perl_taintmode main config option
1449 TL/01 When running the test suite, the README says that variables such as
1450 no_msglog_check are global and can be placed anywhere in a specific
1451 test's script, however it was observed that placement needed to be near
1452 the beginning for it to behave that way. Changed the runtest perl
1453 script to read through the entire script once to detect and set these
1454 variables, reset to the beginning of the script, and then run through
1455 the script parsing/test process like normal.
1457 TL/02 The BSD's have an arc4random API. One of the functions to induce
1458 adding randomness was arc4random_stir(), but it has been removed in
1459 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1460 function when detected.
1462 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1463 cause callback expansion.
1465 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1466 syntax errors in an expansion can be treated as a string instead of
1467 logging or causing an error, due to the internal use of bool_lax
1468 instead of bool when processing it.
1470 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1471 server certificates when making smtp deliveries.
1473 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1475 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1477 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1479 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1480 Merged patch from Sebastian Wiedenroth.
1482 JH/05 Fix results-pipe from transport process. Several recipients, combined
1483 with certificate use, exposed issues where response data items split
1484 over buffer boundaries were not parsed properly. This eventually
1485 resulted in duplicates being sent. This issue only became common enough
1486 to notice due to the introduction of connection certificate information,
1487 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1489 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1490 size buffer was used, resulting in syntax errors when an expansion
1493 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1494 version 3.3.6 or later.
1496 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1497 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1498 both become "event_action", the variables become $event_name, $event_data
1499 and $event_defer_errno. There is a new variable $verify_mode, usable in
1500 routers, transports and related events. The tls:cert event is now also
1501 raised for inbound connections, if the main configuration event_action
1504 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1505 early OCSP support, but no stapling (appears to be less than 1.0.0).
1507 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1508 server certificate names available under the smtp transport option
1509 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1512 JH/10 Time-related extraction expansions from certificates now use the main
1513 option "timezone" setting for output formatting, and are consistent
1514 between OpenSSL and GnuTLS compilations. Bug 1541.
1516 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1517 encoded parameter in the incoming message. Bug 1558.
1519 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1520 include certificate info, eximon was claiming there were spoolfile
1523 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1525 JH/14 Log delivery-related information more consistently, using the sequence
1526 "H=<name> [<ip>]" wherever possible.
1528 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1529 are problematic for Debian distribution, omit them from the release
1532 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1534 JH/16 Fix string representation of time values on 64bit time_t architectures.
1537 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1538 output list separator was used.
1543 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1544 checkers that were complaining about end of non-void function with no
1547 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1548 This was a regression introduced in 4.83 by another bugfix.
1550 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1552 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1553 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1559 TF/01 Correctly close the server side of TLS when forking for delivery.
1561 When a message was received over SMTP with TLS, Exim failed to clear up
1562 the incoming connection properly after forking off the child process to
1563 deliver the message. In some situations the subsequent outgoing
1564 delivery connection happened to have the same fd number as the incoming
1565 connection previously had. Exim would try to use TLS and fail, logging
1566 a "Bad file descriptor" error.
1568 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1569 utilities have not been installed.
1571 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1572 temporary space as the ACL may create new global variables.
1574 TL/01 LDAP support uses per connection or global context settings, depending
1575 upon the detected version of the libraries at build time.
1577 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1578 to extract and use the src ip:port in logging and expansions as if it
1579 were a direct connection from the outside internet. PPv2 support was
1580 updated based on HAProxy spec change in May 2014.
1582 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1584 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1585 Properly escape header and check for NULL return.
1587 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1590 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1592 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1593 characters in header names, implemented as a verify condition.
1594 Contributed by Michael Fischer v. Mollard.
1596 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1597 results permerror and temperror. Previous values are deprecated but
1598 still accepted. In a future release, err_perm and err_temp will be
1599 completely removed, which will be a backward incompatibility if the
1600 ACL tests for either of these two old results. Patch contributed by
1601 user bes-internal on the mailing list.
1603 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1605 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1606 selectors, in both main and reject logs.
1608 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1611 JH/07 Add malware type "sock" for talking to simple daemon.
1613 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1615 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1616 routers/transports under cutthrough routing.
1618 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1619 numbers. Touch up "bool" conditional to keep the same definition.
1621 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1623 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1625 JH/12 Expand items in router/transport headers_add or headers_remove lists
1626 individually rather than the list as a whole. Bug 1452.
1628 Required for reasonable handling of multiple headers_ options when
1629 they may be empty; requires that headers_remove items with embedded
1630 colons must have them doubled (or the list-separator changed).
1632 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1633 view the policy declared in the DMARC record. Currently, $dmarc_status
1634 is a combined value of both the record presence and the result of the
1637 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1639 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1640 dnslookup router and the smtp transport (applying to the forward
1643 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1644 of ldap servers used for a specific lookup. Patch provided by Heiko
1647 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1648 New variable $lookup_dnssec_authenticated for observability.
1650 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1651 Patch submitted by Lars Timman.
1653 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1655 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1656 Requires trusted mode and valid format message id, aborts otherwise.
1657 Patch contributed by Heiko Schlichting.
1659 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1660 certextract with support for various fields. Bug 1358.
1662 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1663 is requested by default, modifiable by smtp transport option
1666 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1667 operate on certificate variables to give certificate fingerprints
1668 Also new ${sha256:cert_variable}.
1670 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1672 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1673 Christian Aistleitner.
1675 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1677 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1678 file. Patch from Wolfgang Breyha.
1680 JH/25 Expand the coverage of the delivery $host and $host_address to
1681 client authenticators run in verify callout. Bug 1476.
1683 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1684 align with daemon_smtp_ports. Bug 72.
1686 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1687 support and error reporting did not work properly.
1689 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1690 and is readable. Patch from Andrew Colin Kissa.
1692 TL/14 Enhance documentation of ${run expansion and how it parses the
1693 commandline after expansion, particularly in the case when an
1694 unquoted variable expansion results in an empty value.
1696 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1698 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1699 by renaming away from it. Addresses GH issue 15, reported by
1702 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1704 TL/15 SECURITY: prevent double expansion in math comparison functions
1705 (can expand unsanitized data). Not remotely exploitable.
1712 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1714 PP/02 Make -n do something, by making it not do something.
1715 When combined with -bP, the name of an option is not output.
1717 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1720 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1721 $sender_host_name and config options to manage this, and basic check
1724 PP/05 DSCP support for outbound connections and control modifier for inbound.
1726 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1727 (Only plugin which currently uses this is kerberos4, which nobody should
1728 be using, but we should make it available and other future plugins might
1729 conceivably use it, even though it would break NAT; stuff *should* be
1730 using channel bindings instead).
1732 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1733 name; added for Sendmail compatibility; requires admin caller.
1734 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1735 just ignore it); requires trusted caller.
1736 Also parse but ignore: -Ac -Am -X<logfile>
1739 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1741 TL/02 Add +smtp_confirmation as a default logging option.
1743 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1744 Patch by Magnus Holmgren from 2007-02-20.
1746 TL/04 Bugzilla 1281 - Spec typo.
1747 Bugzilla 1283 - Spec typo.
1748 Bugzilla 1290 - Spec grammar fixes.
1750 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1752 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1754 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1755 mailing list by by Dmitry Isaikin.
1757 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1759 JH/02 Support "G" suffix to numbers in ${if comparisons.
1761 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1763 NM/01 Bugzilla 1197 - Spec typo
1764 Bugzilla 1196 - Spec examples corrections
1766 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1768 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1769 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1772 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1773 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1775 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1776 "acl {{name}{arg}...}", and optional args on acl condition
1779 JH/05 Permit multiple router/transport headers_add/remove lines.
1781 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1783 JH/07 Avoid using a waiting database for a single-message-only transport.
1784 Performance patch from Paul Fisher. Bugzilla 1262.
1786 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1789 JH/09 Add $headers_added variable, with content from use of ACL modifier
1790 add_header (but not yet added to the message). Bugzilla 199.
1792 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1793 Pulled from Bugzilla 817 by Wolfgang Breyha.
1795 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1797 (nb: this is the same fix as in Exim 4.80.1)
1799 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1802 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1805 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1807 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1808 advertise SMTP AUTH mechanism to us, instead of a generic
1809 protocol violation error. Also, make Exim more robust to bad
1810 data from the Dovecot auth socket.
1812 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1814 When a queue runner is handling a message, Exim first routes the
1815 recipient addresses, during which it prunes them based on the retry
1816 hints database. After that it attempts to deliver the message to
1817 any remaining recipients. It then updates the hints database using
1820 So if a recipient address works intermittently, it can get repeatedly
1821 deferred at routing time. The retry hints record remains fresh so the
1822 address never reaches the final cutoff time.
1824 This is a fairly common occurrence when a user is bumping up against
1825 their storage quota. Exim had some logic in its local delivery code
1826 to deal with this. However it did not apply to per-recipient defers
1827 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1829 This change adds a proper retry rule check during routing so that the
1830 final cutoff time is checked against the message's age. We only do
1831 this check if there is an address retry record and there is not a
1832 domain retry record; this implies that previous attempts to handle
1833 the address had the retry_use_local_parts option turned on. We use
1834 this as an approximation for the destination being like a local
1835 delivery, as in LMTP.
1837 I suspect this new check makes the old local delivery cutoff check
1838 redundant, but I have not verified this so I left the code in place.
1840 TF/02 Correct gecos expansion when From: is a prefix of the username.
1842 Test 0254 submits a message to Exim with the header
1846 When I ran the test suite under the user fanf2, Exim expanded
1847 the header to contain my full name, whereas it should have added
1848 a Resent-Sender: header. It erroneously treats any prefix of the
1849 username as equal to the username.
1851 This change corrects that bug.
1853 GF/01 DCC debug and logging tidyup
1854 Error conditions log to paniclog rather than rejectlog.
1855 Debug lines prefixed by "DCC: " to remove any ambiguity.
1857 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1859 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1860 Bug spotted by Jeremy Harris; was flawed since initial commit.
1861 Would have resulted in OCSP responses post-SNI triggering an Exim
1862 NULL dereference and crash.
1864 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1866 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1867 Bug detection, analysis and fix by Samuel Thibault.
1868 Bugzilla 1331, Debian bug #698092.
1870 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1872 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1873 Server implementation by Todd Lyons, client by JH.
1874 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1875 config variable "prdr_enable" controls whether the server
1876 advertises the facility. If the client requests PRDR a new
1877 acl_data_smtp_prdr ACL is called once for each recipient, after
1878 the body content is received and before the acl_smtp_data ACL.
1879 The client is controlled by both of: a hosts_try_prdr option
1880 on the smtp transport, and the server advertisement.
1881 Default client logging of deliveries and rejections involving
1882 PRDR are flagged with the string "PRDR".
1884 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1885 fclose(). Diagnosis by Todd Lyons.
1887 PP/17 Update configure.default to handle IPv6 localhost better.
1888 Patch by Alain Williams (plus minor tweaks).
1891 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1892 This is now consistent with GnuTLS, and is now documented: the
1893 previous undocumented portable approach to treating the option as
1894 unset was to force an expansion failure. That still works, and
1895 an empty string is now equivalent.
1897 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1898 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1899 not performing validation itself.
1901 PP/20 Added force_command boolean option to pipe transport.
1902 Patch from Nick Koston, of cPanel Inc.
1904 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1907 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1909 PP/21 Fix eximon continuous updating with timestamped log-files.
1910 Broken in a format-string cleanup in 4.80, missed when I repaired the
1911 other false fix of the same issue.
1912 Report and fix from Heiko Schlichting.
1915 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1916 Report from Prashanth Katuri.
1918 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1919 It's SecureTransport, so affects any MacOS clients which use the
1920 system-integrated TLS libraries, including email clients.
1922 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1923 using a MIME ACL for non-SMTP local injection.
1924 Report and assistance in diagnosis by Warren Baker.
1926 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1928 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1930 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1931 last id that failed so it may be referenced in subsequent ACL's.
1933 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1936 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1937 ldap library initialization, allowing self-signed CA's to be
1938 used. Also properly sets require_cert option later in code by
1939 using NULL (global ldap config) instead of ldap handle (per
1940 session). Bug diagnosis and testing by alxgomz.
1942 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1943 the src/util/ subdirectory.
1945 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1946 renamed to Transport Post Delivery Action by Jeremy Harris, as
1949 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1950 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1951 redis_servers = needs to be configured which will be used by the redis
1952 lookup. Patch from Warren Baker, of The Packet Hub.
1954 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1956 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1957 hostname or reverse DNS when processing a host list. Used suggestions
1958 from multiple comments on this bug.
1960 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1962 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1963 Missed a few lines, added it to make the runtest require no keyboard
1966 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1967 contains upper case chars. Make router use caseful_local_part.
1969 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1970 support when GnuTLS has been built with p11-kit.
1976 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1978 This, or similar/improved, will also be change PP/11 of 4.82.
1984 PP/01 Handle short writes when writing local log-files.
1985 In practice, only affects FreeBSD (8 onwards).
1986 Bugzilla 1053, with thanks to Dmitry Isaikin.
1988 NM/01 Bugzilla 949 - Documentation tweak
1990 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1993 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1995 PP/02 Implemented gsasl authenticator.
1997 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1999 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2000 `pkg-config foo` for cflags/libs.
2002 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2003 with rest of GSASL and with heimdal_gssapi.
2005 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2006 `pkg-config foo` for cflags/libs for the TLS implementation.
2008 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2009 properties get this fed in as external SSF. A number of robustness
2010 and debugging improvements to the cyrus_sasl authenticator.
2012 PP/08 cyrus_sasl server now expands the server_realm option.
2014 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2015 Patch by Jeremy Harris.
2017 PP/10 Added dbmjz lookup type.
2019 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2021 PP/12 MAIL args handles TAB as well as SP, for better interop with
2022 non-compliant senders.
2023 Analysis and variant patch by Todd Lyons.
2025 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2026 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2027 Patch from Dirk Mueller <dmueller@suse.com>
2029 PP/13 tls_peerdn now print-escaped for spool files.
2030 Observed some $tls_peerdn in wild which contained \n, which resulted
2031 in spool file corruption.
2033 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2034 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2035 or write after TLS renegotiation, which otherwise led to messages
2038 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2039 as a tracking header (ie: a signed header comes before the signature).
2040 Patch from Wolfgang Breyha.
2042 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2043 comma-sep list; embedded commas doubled.
2045 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2047 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2049 Report and patch from Dmitry Banschikov.
2051 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2052 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2053 failures. We appear to now support TLS1.1+ with Exim.
2055 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2056 lets Exim select keys and certificates based upon TLS SNI from client.
2057 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2058 before an outbound SMTP session. New log_selector, +tls_sni.
2060 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2061 NULL dereference. Report and patch from Alun Jones.
2063 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2064 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2065 Not seeing resolver debug output on NetBSD, but suspect this is a
2066 resolver implementation change.
2068 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2069 Left warnings. Added "eximon gdb" invocation mode.
2071 PP/21 Defaulting "accept_8bitmime" to true, not false.
2073 PP/22 Added -bw for inetd wait mode support.
2075 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2076 locate the relevant includes and libraries. Made this the default.
2078 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2079 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2081 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2082 This may cause build issues on older platforms.
2084 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2085 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2086 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2087 Added SNI support via GnuTLS too.
2088 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2090 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2092 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2093 Applied second patch from Janne, implementing suggestion to default
2094 multiple-strings-in-record handling to match SPF spec.
2096 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2098 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2099 read-only, out of scope).
2100 Patch from Wolfgang Breyha, report from Stuart Northfield.
2102 PP/29 Fix three issues highlighted by clang analyser static analysis.
2103 Only crash-plausible issue would require the Cambridge-specific
2104 iplookup router and a misconfiguration.
2105 Report from Marcin Mirosław.
2107 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2109 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2110 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2111 As part of this, removing so much warning spew let me fix some minor
2112 real issues in debug logging.
2114 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2115 assignment on my part. Fixed.
2117 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2118 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2119 Janne Snabb (who went above and beyond: thank you).
2121 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2122 string otherwise requires a connection and a bunch more work and it's
2123 relatively easy to get wrong. Should also expose TLS library linkage
2126 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2127 64-bit ${eval} (JH/03).
2129 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2130 GNU libc to support some of the 64-bit stuff, should not lead to
2131 conflicts. Defined before os.h is pulled in, so if a given platform
2132 needs to override this, it can.
2134 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2135 protection layer was required, which is not implemented.
2136 Bugzilla 1254, patch from Wolfgang Breyha.
2138 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2139 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2140 tls_dhparam take prime identifiers. Also unbreak combination of
2141 OpenSSL+DH_params+TLSSNI.
2143 PP/39 Disable SSLv2 by default in OpenSSL support.
2149 PP/01 Solaris build fix for Oracle's LDAP libraries.
2150 Bugzilla 1109, patch from Stephen Usher.
2152 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2154 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2157 TF/02 Fix a couple more cases where we did not log the error message
2158 when unlink() failed. See also change 4.74-TF/03.
2160 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2161 lock up or crash if it happened to be inside a call to libc when it
2162 got a SIGUSR1 from exiwhat.
2164 The SIGUSR1 handler appends the current process status to the process
2165 log which is later printed by exiwhat. It used to use the general
2166 purpose logging code to do this, but several functions it calls are
2167 not safe for signals.
2169 The new output code in the SIGUSR1 handler is specific to the process
2170 log, and simple enough that it's easy to inspect for signal safety.
2171 Removing some special cases also simplifies the general logging code.
2172 Removing the spurious timestamps from the process log simplifies
2175 TF/04 Improved ratelimit ACL condition.
2177 The /noupdate option has been deprecated in favour of /readonly which
2178 has clearer semantics. The /leaky, /strict, and /readonly update modes
2179 are mutually exclusive. The update mode is no longer included in the
2180 database key; it just determines when the database is updated. (This
2181 means that when you upgrade Exim will forget old rate measurements.)
2183 Exim now checks that the per_* options are used with an update mode that
2184 makes sense for the current ACL. For example, when Exim is processing a
2185 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2186 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2187 must specify per_mail/readonly. If you omit the update mode it defaults to
2188 /leaky where that makes sense (as before) or /readonly where required.
2190 The /noupdate option is now undocumented but still supported for
2191 backwards compatibility. It is equivalent to /readonly except that in
2192 ACLs where /readonly is required you may specify /leaky/noupdate or
2193 /strict/noupdate which are treated the same as /readonly.
2195 A useful new feature is the /count= option. This is a generalization
2196 of the per_byte option, so that you can measure the throughput of other
2197 aggregate values. For example, the per_byte option is now equivalent
2198 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2200 The per_rcpt option has been generalized using the /count= mechanism
2201 (though it's more complicated than the per_byte equivalence). When it is
2202 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2203 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2204 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2205 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2206 Note that using per_rcpt with a non-readonly update mode in more than
2207 one ACL will cause the recipients to be double-counted. (The per_mail
2208 and per_byte options don't have this problem.)
2210 The handling of very low rates has changed slightly. If the computed rate
2211 is less than the event's count (usually one) then this event is the first
2212 after a long gap. In this case the rate is set to the same as this event's
2213 count, so that the first message of a spam run is counted properly.
2215 The major new feature is a mechanism for counting the rate of unique
2216 events. The new per_addr option counts the number of different
2217 recipients that someone has sent messages to in the last time period. It
2218 behaves like per_rcpt if all the recipient addresses are different, but
2219 duplicate recipient addresses do not increase the measured rate. Like
2220 the /count= option this is a general mechanism, so the per_addr option
2221 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2222 example, measure the rate that a client uses different sender addresses
2223 with the options per_mail/unique=$sender_address. There are further
2224 details in the main documentation.
2226 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2228 TF/06 Removed a few PCRE remnants.
2230 TF/07 Automatically extract Exim's version number from tags in the git
2231 repository when doing development or release builds.
2233 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2234 Bugzilla 879. Patch from Paul Fisher.
2236 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2237 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2240 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2242 PP/05 Variable $av_failed, true if the AV scanner deferred.
2243 Bugzilla 1078. Patch from John Horne.
2245 PP/06 Stop make process more reliably on build failure.
2246 Bugzilla 1087. Patch from Heiko Schlittermann.
2248 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2249 Bugzilla 1089. Patch from Heiko Schlittermann.
2251 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2252 Bugzilla 1131. Patch from Holger Weiß.
2254 PP/09 Handle IPv6 addresses with SPF.
2255 Bugzilla 860. Patch from Wolfgang Breyha.
2257 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2259 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2262 PP/11 match_* no longer expand right-hand-side by default.
2263 New compile-time build option, EXPAND_LISTMATCH_RHS.
2264 New expansion conditions, "inlist", "inlisti".
2266 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2268 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2270 PP/14 fix log_write() format string regression from TF/03.
2271 Bugzilla 1152. Patch from Dmitry Isaikin.
2277 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2279 PP/02 Harmonised TLS library version reporting; only show if debugging.
2280 Layout now matches that introduced for other libraries in 4.74 PP/03.
2282 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2284 PP/04 New "dns_use_edns0" global option.
2286 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2289 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2290 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2292 TK/01 Updated PolarSSL code to 0.14.2.
2293 Bugzilla 1097. Patch from Andreas Metzler.
2295 PP/07 Catch divide-by-zero in ${eval:...}.
2296 Fixes bugzilla 1102.
2298 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2301 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2302 format-string attack -- SECURITY: remote arbitrary code execution.
2304 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2305 time unintentionally subject to list matching rules, letting the header
2306 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2307 arbitrary string expansion). This allowed for information disclosure.
2309 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2310 INT_MIN/-1 -- value coerced to INT_MAX.
2316 NM/01 Workaround for PCRE version dependency in version reporting
2319 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2320 This fixes portability to compilers other than gcc, notably
2321 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2323 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2324 makefiles for portability to HP-UX and POSIX correctness.
2326 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2327 Also via indented variable definition in the Makefile.
2328 (Debugging by Oliver Heesakkers).
2330 PP/02 Restore caching of spamd results with expanded spamd_address.
2331 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2333 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2334 Improves build reliability. Fix from: Frank Elsner
2336 NM/02 Fix wide character breakage in the rfc2047 coding
2337 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2339 NM/03 Allow underscore in dnslist lookups
2340 Fixes bug 1026. Patch from Graeme Fowler
2342 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2343 Code patches from Adam Ciarcinski of NetBSD.
2345 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2348 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2349 is logged, to avoid truncation. Patch from John Horne.
2351 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2352 Patch from Jakob Hirsch.
2354 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2355 SQL string expansion failure details.
2356 Patch from Andrey Oktyabrski.
2358 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2359 Patch from Simon Arlott.
2361 PP/09 New lookups functionality failed to compile on old gcc which rejects
2362 extern declarations in function scope.
2363 Patch from Oliver Fleischmann
2365 PP/10 Use sig_atomic_t for flags set from signal handlers.
2366 Check getgroups() return and improve debugging.
2367 Fixed developed for diagnosis in bug 927 (which turned out to be
2370 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2371 Patch from Mark Zealey.
2373 PP/12 Bugzilla 1056: Improved spamd server selection.
2374 Patch from Mark Zealey.
2376 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2377 Based on patch from Heiko Schlittermann.
2379 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2380 Patch from Uwe Doering, sign-off by Michael Haardt.
2382 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2383 variable declaration deep within a block. Bug and patch from
2386 PP/15 lookups-Makefile IRIX compatibility coercion.
2388 PP/16 Make DISABLE_DKIM build knob functional.
2390 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2391 Patch by Simon Arlott
2393 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2394 variable argument macros. Our copy now differs from upstream.
2400 TF/01 Failure to get a lock on a hints database can have serious
2401 consequences so log it to the panic log.
2403 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2404 controlled using the smtp_confirmation log selector.
2406 TF/03 Include the error message when we fail to unlink a spool file.
2408 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2409 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2410 for maintaining out-of-tree patches for some time.
2412 PP/01 Bugzilla 139: Documentation and portability issues.
2413 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2414 Handle per-OS dynamic-module compilation flags.
2416 PP/02 Let /dev/null have normal permissions.
2417 The 4.73 fixes were a little too stringent and complained about the
2418 permissions on /dev/null. Exempt it from some checks.
2419 Reported by Andreas M. Kirchwitz.
2421 PP/03 Report version information for many libraries, including
2422 Exim version information for dynamically loaded libraries. Created
2423 version.h, now support a version extension string for distributors
2424 who patch heavily. Dynamic module ABI change.
2426 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2427 privilege escalation vulnerability whereby the Exim run-time user
2428 can cause root to append content of the attacker's choosing to
2431 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2434 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2435 If dropping privileges for untrusted macros, we disabled normal logging
2436 on the basis that it would fail; for the Exim run-time user, this is not
2437 the case, and it resulted in successful deliveries going unlogged.
2438 Fixed. Reported by Andreas Metzler.
2444 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2445 only prepend for the Resent-* case. Fixes regression introduced in
2446 Exim 4.70 by NM/22 for Bugzilla 607.
2448 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2449 increasing numbers of administrators be bitten by this.
2451 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2453 PP/03 Bugzilla 994: added openssl_options main configuration option.
2455 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2457 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2459 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2461 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2462 without a peer certificate, leading to a segfault because of an
2463 assumption that peers always have certificates. Be a little more
2464 paranoid. Problem reported by Martin Tscholak.
2466 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2467 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2468 NB: ClamAV planning to remove STREAM in "middle of 2010".
2469 CL also introduces -bmalware, various -d+acl logging additions and
2470 more caution in buffer sizes.
2472 PP/09 Implemented reverse_ip expansion operator.
2474 PP/10 Bugzilla 937: provide a "debug" ACL control.
2476 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2478 PP/12 Bugzilla 973: Implement --version.
2480 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2482 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2484 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2486 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2487 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2488 ignore trailing whitespace.
2490 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2492 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2495 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2496 Notification from Dr Andrew Aitchison.
2498 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2499 ExtendedDetectionInfo response format.
2500 Notification from John Horne.
2502 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2505 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2506 XSL and documented dependency on system catalogs, with examples of how
2509 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2512 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2513 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2514 configuration file which is writeable by the Exim user or group.
2516 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2517 of configuration files to cover files specified with the -C option if
2518 they are going to be used with root privileges, not just the default
2521 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2522 option (effectively making it always true).
2524 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2525 files to be used while preserving root privileges.
2527 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2528 that rogue child processes cannot use them.
2530 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2531 run-time user, instead of root.
2533 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2534 Exim run-time user without dropping privileges.
2536 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2537 result string, instead of calling string_vformat() twice with the same
2540 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2541 for other users. Others should always drop root privileges if they use
2542 -C on the command line, even for a whitelisted configure file.
2544 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2546 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2552 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2553 $data_path, and $header_path variables; fixed documentation bugs and
2556 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2557 exipick to access non-standard spools, including the "frozen" queue
2560 NM/01 Bugzilla 965: Support mysql stored procedures.
2561 Patch from Alain Williams
2563 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2565 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2566 Patch from Andreas Metzler
2568 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2569 Patch from Kirill Miazine
2571 NM/05 Bugzilla 671: Added umask to procmail example.
2573 JJ/03 installed exipick 20100323.0, fixing doc bug
2575 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2576 directory. Notification and patch from Dan Rosenberg.
2578 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2580 TK/02 Improve log output when DKIM signing operation fails.
2582 MH/01 Treat the transport option dkim_domain as a colon separated
2583 list, not as a single string, and sign the message with each element,
2584 omitting multiple occurences of the same signer.
2586 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2587 Bugzilla 985, 986. Patch by Simon Arlott
2589 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2590 Patch by Simon Arlott
2592 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2593 MBX locking. Notification from Dan Rosenberg.
2599 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2601 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2603 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2605 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2607 NM/04 Bugzilla 915: Flag broken perl installation during build.
2613 TK/01 Added patch by Johannes Berg that expands the main option
2614 "spamd_address" if it starts with a dollar sign.
2616 TK/02 Write list of recipients to X-Envelope-Sender header when building
2617 the mbox-format spool file for content scanning (suggested by Jakob
2620 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2621 (http://www.dcc-servers.net/) support via dccifd. Activated by
2622 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2624 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2625 by Mark Daniel Reidel <mr@df.eu>.
2627 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2628 When building exim an external PCRE library is now needed -
2629 PCRE is a system library on the majority of modern systems.
2630 See entry on PCRE_LIBS in EDITME file.
2632 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2633 conversation. Added nologin parameter to request.
2634 Patch contributed by Kirill Miazine.
2636 TF/01 Do not log submission mode rewrites if they do not change the address.
2638 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2640 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2641 log files in place. Contributed by Roberto Lima.
2643 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2645 TF/03 Bugzilla 615: When checking the local_parts router precondition
2646 after a local_part_suffix or local_part_prefix option, Exim now
2647 does not use the address's named list lookup cache, since this
2648 contains cached lookups for the whole local part.
2650 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2651 Robert Millan. Documentation is in experimental-spec.txt.
2653 TF/04 Bugzilla 668: Fix parallel build (make -j).
2655 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2657 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2658 Patch provided by Jan Srzednicki.
2660 TF/05 Leading white space used to be stripped from $spam_report which
2661 wrecked the formatting. Now it is preserved.
2663 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2664 that they are available at delivery time.
2666 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2668 TF/08 TLS error reporting now respects the incoming_interface and
2669 incoming_port log selectors.
2671 TF/09 Produce a more useful error message if an SMTP transport's hosts
2672 setting expands to an empty string.
2674 NM/06 Bugzilla 744: EXPN did not work under TLS.
2675 Patch provided by Phil Pennock.
2677 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2678 Patch provided by Richard Godbee.
2680 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2681 acl_smtp_notquit, added index entry.
2683 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2684 Patch provided by Eugene Bujak.
2686 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2687 accept(). Patch provided by Maxim Dounin.
2689 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2690 Patch provided by Phil Pennock.
2692 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2694 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2695 Patch provided by Brad "anomie" Jorsch.
2697 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2698 Patch provided by Dean Brooks.
2700 TK/05 Add native DKIM support (does not depend on external libraries).
2702 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2703 Patch provided by Graeme Fowler.
2705 NM/16 Bugzilla 851: Documentation example syntax fix.
2707 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2709 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2712 NM/19 Bugzilla 745: TLS version reporting.
2713 Patch provided by Phil Pennock.
2715 NM/20 Bugzilla 167: bool: condition support.
2716 Patch provided by Phil Pennock.
2718 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2719 clients. Patch provided by Phil Pennock.
2721 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2722 Patch provided by Brad "anomie" Jorsch.
2724 NM/23 Bugzilla 687: Fix misparses in eximstats.
2725 Patch provided by Heiko Schlittermann.
2727 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2728 Patch provided by Heiko Schlittermann.
2730 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2731 plus update to original patch.
2733 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2735 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2736 Patch provided by David Brownlee.
2738 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2740 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2742 NM/30 Bugzilla 888: TLS documentation bugfixes.
2744 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2746 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2747 Unlike the original bugzilla I have changed all shell scripts in src tree.
2749 NM/33 Bugzilla 898: Transport filter timeout fix.
2750 Patch by Todd Rinaldo.
2752 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2753 Patch by Serge Demonchaux.
2755 NM/35 Bugzilla 39: Base64 decode bug fixes.
2756 Patch by Jakob Hirsch.
2758 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2760 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2762 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2764 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2770 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2771 ALT-N's libdkim that I have put here:
2772 http://duncanthrax.net/exim-experimental/
2774 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2775 were called 'true' and it seems that C99 defines that as a reserved
2776 keyword to be used with 'bool' variable types. That means you could
2777 not include C99-style headers which use bools without triggering
2778 build errors in sieve.c.
2780 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2781 as mailq or other aliases. Changed the --help handling significantly
2782 to do whats expected. exim_usage() emits usage/help information.
2784 SC/01 Added the -bylocaldomain option to eximstats.
2786 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2788 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2790 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2796 PH/01 Another patch from the Sieve maintainer.
2798 PH/02 When an IPv6 address is converted to a string for single-key lookup
2799 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2800 dots are used instead of colons so that keys in lsearch files need not
2801 contain colons. This was done some time before quoting was made available
2802 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2803 (notated using the quote facility) so as to distinguish them from IPv4
2804 keys. This meant that lookups for IP addresses in host lists did not work
2805 for iplsearch lookups.
2807 This has been fixed by arranging for IPv6 addresses to be expressed with
2808 colons if the lookup type is iplsearch. This is not incompatible, because
2809 previously such lookups could never work.
2811 The situation is now rather anomalous, since one *can* have colons in
2812 ordinary lsearch keys. However, making the change in all cases is
2813 incompatible and would probably break a number of configurations.
2815 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2818 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2819 conversion specification without a maximum field width, thereby enabling
2820 a rogue spamd server to cause a buffer overflow. While nobody in their
2821 right mind would setup Exim to query an untrusted spamd server, an
2822 attacker that gains access to a server running spamd could potentially
2823 exploit this vulnerability to run arbitrary code as the Exim user.
2825 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2826 $primary_hostname instead of what libspf2 thinks the hosts name is.
2828 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2829 a directory entry by the name of the lookup key. Previously, if a
2830 symlink pointed to a non-existing file or a file in a directory that
2831 Exim lacked permissions to read, a lookup for a key matching that
2832 symlink would fail. Now it is enough that a matching directory entry
2833 exists, symlink or not. (Bugzilla 503.)
2835 PH/03 The body_linecount and body_zerocount variables are now exported in the
2838 PH/04 Added the $dnslist_matched variable.
2840 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2841 This means they are set thereafter only if the connection becomes
2844 PH/06 Added the client_condition to authenticators so that some can be skipped
2845 by clients under certain conditions.
2847 PH/07 The error message for a badly-placed control=no_multiline_responses left
2848 "_responses" off the end of the name.
2850 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2852 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2853 (without spaces) instead of just copying the configuration text.
2855 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2857 PH/11 Added $max_received_linelength.
2859 PH/12 Added +ignore_defer and +include_defer to host lists.
2861 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2862 way in which PCRE > 7.0 is built.
2864 PH/14 Implemented queue_only_load_latch.
2866 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2867 MAIL command. The effect was to mangle the value on 64-bit systems.
2869 PH/16 Another patch from the Sieve maintainer.
2871 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2873 PH/18 If a system quota error occurred while trying to create the file for
2874 a maildir delivery, the message "Mailbox is full" was not appended to the
2875 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2876 only to a quota excession during the actual writing of the file.
2878 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2879 characters?) which causes problems in log lines. The DN values are now
2880 passed through string_printing() before being added to log lines.
2882 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2883 and InterBase are left for another time.)
2885 PH/21 Added message_body_newlines option.
2887 PH/22 Guard against possible overflow in moan_check_errorcopy().
2889 PH/23 POSIX allows open() to be a macro; guard against that.
2891 PH/24 If the recipient of an error message contained an @ in the local part
2892 (suitably quoted, of course), incorrect values were put in $domain and
2893 $local_part during the evaluation of errors_copy.
2899 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2900 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2903 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2904 issue a MAIL command.
2906 PH/02 In an ACL statement such as
2908 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2910 if a client was not listed at all, or was listed with a value other than
2911 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2912 the condition was not true (as it should be), so access was not denied.
2913 The bug was that the ! inversion was incorrectly passed on to the second
2914 item. This has been fixed.
2916 PH/03 Added additional dnslists conditions == and =& which are different from
2917 = and & when the dns lookup returns more than one IP address.
2919 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2920 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2922 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2923 FSYNC, which compiles an option called disable_fsync that allows for
2924 bypassing fsync(). The documentation is heavily laced with warnings.
2926 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2928 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2929 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2930 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2931 including adding "make clean"; (3) Added -fPIC when compiling the test
2932 dynamically loaded module, to get rid of a warning.
2934 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2935 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2936 = 0s. The bug is otherwise harmless.
2938 PH/07 There was a bug in the dovecot authenticator such that the value of
2939 $auth1 could be overwritten, and so not correctly preserved, after a
2940 successful authentication. This usually meant that the value preserved by
2941 the server_setid option was incorrect.
2943 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2945 PH/09 Installed PCRE release 7.0.
2947 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2948 run for batched SMTP input. It is now run at the start of every message
2949 in the batch. While fixing this I discovered that the process information
2950 (output by running exiwhat) was not always getting set for -bs and -bS
2951 input. This is fixed, and it now also says "batched" for BSMTP.
2953 PH/11 Added control=no_pipelining.
2955 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2956 patch, slightly modified), and move the expansion of helo_data till after
2957 the connection is made in the smtp transport (so it can use these
2960 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2962 PH/14 Added log_selector = +pid.
2964 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2966 PH/16 Add ${if forany and ${if forall.
2968 PH/17 Added dsn_from option to vary the From: line in DSNs.
2970 PH/18 Flush SMTP output before performing a callout, unless control =
2971 no_callout_flush is set.
2973 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2974 was true (the default) a successful delivery failed to delete the retry
2975 item, thus causing premature timeout of the address. The bug is now
2978 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2980 PH/21 Long custom messages for fakedefer and fakereject are now split up
2981 into multiline responses in the same way that messages for "deny" and
2982 other ACL rejections are.
2984 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2985 with slight modification.
2987 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2988 draft, changing the syntax and factoring some duplicate code".
2990 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2991 for deliveries of the second and subsequent messages over the same SMTP
2994 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2995 ${reduce, with only minor "tidies".
2997 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2999 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3000 expansion side effects.
3002 PH/27 When a message times out after an over-quota error from an Exim-imposed
3003 quota, the bounce message says "mailbox is full". This message was not
3004 being given when it was a system quota that was exceeded. It now should
3007 MH/03 Made $recipients available in local_scan(). local_scan() already has
3008 better access to the recipient list through recipients_list[], but
3009 $recipients can be useful in postmaster-provided expansion strings.
3011 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3012 in the case of a MAIL command with additional options following the
3013 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3014 were accidentally chopped off.
3016 PH/29 SMTP synchronization checks are implemented when a command is read -
3017 there is a check that no more input is waiting when there shouldn't be
3018 any. However, for some commands, a delay in an ACL can mean that it is
3019 some time before the response is written. In this time, more input might
3020 arrive, invalidly. So now there are extra checks after an ACL has run for
3021 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3022 pipelining has not been advertised.
3024 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3026 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3027 returned in the SMTP response when smtp_return_error_details was false.
3028 This has been fixed.
3030 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3031 instead of the C I/O that was originally supplied, because problems were
3032 reported on Solaris.
3034 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3035 Exim which did not show up earlier: it was assuming that a call to
3036 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3037 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3038 was a macro that became an assignment, so it seemed to work. This has
3039 changed to a proper function call with a void return, hence the compile
3040 error. Exim's code has been fixed.
3042 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3045 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3047 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3049 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3050 because it turns out to be too big or there is a timeout) there is no
3051 "Completed" line in the log. When some messages of this type were
3052 selected by exigrep, they were listed as "not completed". Others were
3053 picked up by some special patterns. I have improved the selection
3054 criteria to be more general.
3056 PH/38 The host_find_failed option in the manualroute router can now be set
3057 to "ignore", to completely ignore a host whose IP address cannot be
3058 found. If all hosts are ignored, the behaviour is controlled by the new
3059 host_all_ignored option.
3061 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3062 homing or because of multiple MX records with /mx) generated more than
3063 one IP address, and the following item turned out to be the local host,
3064 all the secondary addresses of the first item were incorrectly removed
3065 from the list, along with the local host and any following hosts (which
3066 is what is supposed to happen).
3068 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3069 whoever called Exim into the -H file. In the case of the daemon it was
3070 behaving confusingly. When first started, it used values for whoever
3071 started the daemon, but after a SIGHUP it used the Exim user (because it
3072 calls itself on a restart). I have changed the code so that it now always
3075 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3076 message are rejected with the same error (e.g. no authentication or bad
3077 sender address), and a DATA command is nevertheless sent (as can happen
3078 with PIPELINING or a stupid MUA), the error message that was given to the
3079 RCPT commands is included in the rejection of the DATA command. This is
3080 intended to be helpful for MUAs that show only the final error to their
3083 PH/42 Another patch from the Sieve maintainer.
3085 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3086 Eximstats - Fixed some broken HTML links and added missing column headers
3088 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3089 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3091 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3093 PH/43 Yet another patch from the Sieve maintainer.
3095 PH/44 I found a way to check for a TCP/IP connection going away before sending
3096 the response to the final '.' that terminates a message, but only in the
3097 case where the client has not sent further data following the '.'
3098 (unfortunately, this is allowed). However, in many cases there won't be
3099 any further data because there won't be any more messages to send. A call
3100 to select() can be used: if it shows that the input is "ready", there is
3101 either input waiting, or the socket has been closed. An attempt to read
3102 the next input character can distinguish the two cases. Previously, Exim
3103 would have sent an OK response which the client would never have see.
3104 This could lead to message repetition. This fix should cure that, at
3105 least in a lot of common cases.
3107 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3108 advertised in response to EHLO.
3114 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3115 fixed by 4.65/MH/01 (is this a record?) are fixed:
3117 (i) An empty string was always treated as zero by the numeric comparison
3118 operators. This behaviour has been restored.
3120 (ii) It is documented that the numeric comparison operators always treat
3121 their arguments as decimal numbers. This was broken in that numbers
3122 starting with 0 were being interpreted as octal.
3124 While fixing these problems I realized that there was another issue that
3125 hadn't been noticed. Values of message_size_limit (both the global option
3126 and the transport option) were treated as octal if they started with 0.
3127 The documentation was vague. These values are now always treated as
3128 decimal, and I will make that clear in the documentation.
3134 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3135 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3138 MH/01 Don't check that the operands of numeric comparison operators are
3139 integers when their expansion is in "skipping" mode (fixes bug
3140 introduced by 4.64-PH/07).
3142 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3143 child addresses, Exim now panics and dies. Previously, because the count
3144 is held in a short int, deliveries were likely to be lost. As such a
3145 large number of recipients for a single message is ridiculous
3146 (performance will be very, very poor), I have chosen to impose a limit
3147 rather than extend the field.
3153 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3154 leftover -K file (the existence of which was triggered by #402).
3155 While we were at it, introduced process PID as part of the -K
3156 filename. This should rule out race conditions when creating
3159 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3160 processing considerably. Previous code took too long for large mails,
3161 triggering a timeout which in turn triggers #401.
3163 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3164 in the DK code in transports.c. sendfile() is not really portable,
3165 hence the _LINUX specificness.
3167 TF/01 In the add_headers option to the mail command in an Exim filter,
3168 there was a bug that Exim would claim a syntax error in any
3169 header after the first one which had an odd number of characters
3172 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3173 callout verification, Exim cached a "reject" for the entire domain. This
3174 is correct for most verifications, but it is not correct for a recipient
3175 verification with use_sender or use_postmaster set, because in that case
3176 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3177 case of MAIL FROM:<> rejection from other early rejections (e.g.
3178 rejection of HELO). When verifying a recipient using a non-null MAIL
3179 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3180 Whatever the result of the callout, the value of the domain cache is
3181 left unchanged (for any other kind of callout, getting as far as trying
3182 RCPT means that the domain itself is ok).
3184 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3187 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3188 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3189 session, the daemon ignores ECONNECT errors and logs others; it now
3190 ignores EPIPE as well.
3192 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3193 (quoted-printable decoding).
3195 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3196 later the small subsequent patch to fix an introduced bug.
3198 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3200 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3202 PH/08 An error is now given if message_size_limit is specified negative.
3204 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3205 to be given (somewhat) arbitrary names.
3207 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3210 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3211 miscellaneous code fixes
3213 PH/10 Added the log_reject_target ACL modifier to specify where to log
3216 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3217 hostname. This is wrong, because it relates to the incoming message (and
3218 probably the interface on which it is arriving) and not to the outgoing
3219 callout (which could be using a different interface). This has been
3220 changed to use the value of the helo_data option from the smtp transport
3221 instead - this is what is used when a message is actually being sent. If
3222 there is no remote transport (possible with a router that sets up host
3223 addresses), $smtp_active_hostname is used.
3225 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3226 tweaks were necessary in order to get it to work (see also 21 below):
3227 (a) The code assumed that strncpy() returns a negative number on buffer
3228 overflow, which isn't the case. Replaced with Exim's string_format()
3230 (b) There were several signed/unsigned issues. I just did the minimum
3231 hacking in of casts. There is scope for a larger refactoring.
3232 (c) The code used strcasecmp() which is not a standard C function.
3233 Replaced with Exim's strcmpic() function.
3234 (d) The code set only $1; it now sets $auth1 as well.
3235 (e) A simple test gave the error "authentication client didn't specify
3236 service in request". It would seem that Dovecot has changed its
3237 interface. Fortunately there's a specification; I followed it and
3238 changed what the client sends and it appears to be working now.
3240 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3243 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3244 address A is aliased to B and C, where B exists and C does not. Without
3245 -v the output is "A verified" because verification stops after a
3246 successful redirection if more than one address is generated. However,
3247 with -v the child addresses are also verified. Exim was outputting "A
3248 failed to verify" and then showing the successful verification for C,
3249 with its parentage. It now outputs "B failed to verify", showing B's
3250 parentage before showing the successful verification of C.
3252 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3253 look up a TXT record in a specific list after matching in a combined
3256 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3257 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3258 they consult the DNS. I had assumed they would set it the way they
3259 wanted; and indeed my experiments on Linux seem to show that in some
3260 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3261 To be on the safe side, however, I have now made the interface to
3262 host_find_byname() similar to host_find_bydns(), with an argument
3263 containing the DNS resolver options. The host_find_byname() function now
3264 sets these options at its start, just as host_find_bydns() does. The smtp
3265 transport options dns_qualify_single and dns_search_parents are passed to
3266 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3267 of host_find_byname() use the default settings of RES_DEFNAMES
3268 (qualify_single) but not RES_DNSRCH (search_parents).
3270 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3271 spool_read_header() do less string testing, by means of a preliminary
3272 switch on the second character of optional "-foo" lines. (This is
3273 overdue, caused by the large number of possibilities that now exist.
3274 Originally there were few.) While I was there, I also converted the
3275 str(n)cmp tests so they don't re-test the leading "-" and the first
3276 character, in the hope this might squeeze out yet more improvement.
3278 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3279 flag allowing group syntax was set by the header_syntax check but not
3280 turned off, possible causing trouble later; (2) The flag was not being
3281 set at all for the header_verify test, causing "group"-style headers to
3282 be rejected. I have now set it in this case, and also caused header_
3283 verify to ignore an empty address taken from a group. While doing this, I
3284 came across some other cases where the code for allowing group syntax
3285 while scanning a header line wasn't quite right (mostly, not resetting
3286 the flag correctly in the right place). These bugs could have caused
3287 trouble for malformed header lines. I hope it is now all correct.
3289 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3290 with the "reply" argument non-NULL. The code, however (which originally
3291 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3292 but it didn't always do it. This confused somebody who was copying the
3293 code for some other use. I have removed all the tests.
3295 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3296 feature that was used to support insecure browsers during the U.S. crypto
3297 embargo. It requires special client support, and Exim is probably the
3298 only MTA that supported it -- and would never use it because real RSA is
3299 always available. This code has been removed, because it had the bad
3300 effect of slowing Exim down by computing (never used) parameters for the
3301 RSA_EXPORT functionality.
3303 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3304 authenticator to fail if there's a tab character in the incoming data
3305 (there should never be unless someone is messing about, as it's supposed
3306 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3307 if the connection is using TLS or if the remote IP is the same as the
3308 local IP, and the "valid-client-cert option" if a client certificate has
3311 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3312 authenticators. This can be used for authorization after authentication
3313 succeeds. (In the case of plaintext, it servers for both authentication
3316 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3317 if any retry times were supplied.
3319 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3320 connection, where there is no client IP address to check. In this
3321 situation, the verify now always succeeds.
3323 PH/25 Applied John Jetmore's -Mset patch.
3325 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3327 PH/27 In a string expansion for a processed (not raw) header when multiple
3328 headers of the same name were present, leading whitespace was being
3329 removed from all of them, but trailing whitespace was being removed only
3330 from the last one. Now trailing whitespace is removed from each header
3331 before concatenation. Completely empty headers in a concatenation (as
3332 before) are ignored.
3334 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3335 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3337 PH/29 [Removed. This was a change that I later backed out, and forgot to
3338 correct the ChangeLog entry (that I had efficiently created) before
3339 committing the later change.]
3341 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3342 address errors (4xx response to RCPT) over the same connection as other
3343 messages routed to the same hosts. Such deliveries are always "forced",
3344 so retry times are not inspected. This resulted in far too many retries
3345 for the affected addresses. The effect occurred only when there were more
3346 hosts than the hosts_max_try setting in the smtp transport when it had
3347 the 4xx errors. Those hosts that it had tried were not added to the list
3348 of hosts for which the message was waiting, so if all were tried, there
3349 was no problem. Two fixes have been applied:
3351 (i) If there are any address or message errors in an SMTP delivery, none
3352 of the hosts (tried or untried) are now added to the list of hosts
3353 for which the message is waiting, so the message should not be a
3354 candidate for sending over the same connection that was used for a
3355 successful delivery of some other message. This seems entirely
3356 reasonable: after all the message is NOT "waiting for some host".
3357 This is so "obvious" that I'm not sure why it wasn't done
3358 previously. Hope I haven't missed anything, but it can't do any
3359 harm, as the worst effect is to miss an optimization.
3361 (ii) If, despite (i), such a delivery is accidentally attempted, the
3362 routing retry time is respected, so at least it doesn't keep
3363 hammering the server.
3365 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3366 in ${readsocket because some servers need this prod.
3368 PH/32 Added some extra debug output when updating a wait-xxx database.
3370 PH/33 The hint "could be header name not terminated by colon", which has been
3371 given for certain expansion errors for a long time, was not being given
3372 for the ${if def:h_colon_omitted{... case.
3374 PH/34 The spec says: "With one important exception, whenever a domain list is
3375 being scanned, $domain contains the subject domain." There was at least
3376 one case where this was not true.
3378 PH/35 The error "getsockname() failed: connection reset by peer" was being
3379 written to the panic log as well as the main log, but it isn't really
3380 panic-worthy as it just means the connection died rather early on. I have
3381 removed the panic log writing for the ECONNRESET error when getsockname()
3384 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3385 runs only) independently of the message's sender address. This meant
3386 that, if the 4xx error was in fact related to the sender, a different
3387 message to the same recipient with a different sender could confuse
3388 things. In particular, this can happen when sending to a greylisting
3389 server, but other circumstances could also provoke similar problems.
3390 I have changed the default so that the retry time for these errors is now
3391 based a combination of the sender and recipient addresses. This change
3392 can be overridden by setting address_retry_include_sender=false in the
3395 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3396 remote server are returned as part of bounce messages. This was not
3397 happening for LMTP over a pipe (the lmtp transport), but now it is the
3398 same for both kinds of LMTP.
3400 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3401 in header lines that were in fact CNAMEs. This is no longer the case.
3403 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3404 and queue runs started by the daemon processed all messages. This has
3405 been fixed so that -R and -S can now usefully be given with -q<time>.
3407 PH/40 Import PCRE release 6.7 (fixes some bugs).
3409 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3411 PH/42 Give an error if -q is specified more than once.
3413 PH/43 Renamed the variables $interface_address and $interface_port as
3414 $received_ip_address and $received_port, to make it clear that these
3415 values apply to message reception, and not to the outgoing interface when
3416 a message is delivered. (The old names remain recognized, of course.)
3418 PH/44 There was no timeout on the connect() call when using a Unix domain
3419 socket in the ${readsocket expansion. There now is.
3421 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3422 be meaningful with "accept".
3424 SC/01 Eximstats V1.43
3425 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3427 SC/02 Eximstats V1.44
3428 Use a glob alias rather than an array ref in the generated
3429 parser. This improves both readability and performance.
3431 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3432 Collect SpamAssassin and rejection statistics.
3433 Don't display local sender or destination tables unless
3434 there is data to show.
3435 Added average volumes into the top table text output.
3437 SC/04 Eximstats V1.46
3438 Collect data on the number of addresses (recipients)
3439 as well as the number of messages.
3441 SC/05 Eximstats V1.47
3442 Added 'Message too big' to the list of mail rejection
3443 reasons (thanks to Marco Gaiarin).
3445 SC/06 Eximstats V1.48
3446 Mainlog lines which have GMT offsets and are too short to
3447 have a flag are now skipped.
3449 SC/07 Eximstats V1.49 (Alain Williams)
3450 Added the -emptyok flag.
3452 SC/08 Eximstats V1.50
3453 Fixes for obtaining the IP address from reject messages.
3455 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3456 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3457 whitespace changes from 4.64-PH/27
3459 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3462 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3463 are found, allow negative numbers in numeric criteria)
3465 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3467 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3468 to match changes made in 4.64-PH/43
3470 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3472 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3473 "domain" is actually an IP address, and force a failure. This locks out
3474 those revolvers/nameservers that support "A-for-A" lookups, in
3475 contravention of the specifications.
3477 PH/48 When a host name was looked up from an IP address, and the subsequent
3478 forward lookup of the name timed out, the host name was left in
3479 $sender_host_name, contrary to the specification.
3481 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3482 restricted to single-key lookups, Exim was not diagnosing an error if
3483 * or *@ was used with a query-style lookup.
3485 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3487 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3488 long ago, but noone interested enough thought of it. Let's just say that
3489 the "1.1" means that there are some new functions that weren't there at
3490 some point in the past.
3492 PH/51 Error processing for expansion failure of helo_data from an smtp
3493 transport during callout processing was broken.
3495 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3496 tested/used via the -bh/-bhc/-bs options.
3498 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3499 bug, fixed in subsequent PCRE releases).
3501 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3502 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3504 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3510 SC/01 Use a glob alias rather than an array ref in eximstats generated
3511 parser. This improves both readability and performance.
3513 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3514 Don't display local sender or destination tables in eximstats unless
3515 there is data to show.
3516 Added average volumes into the eximstats top table text output.
3518 SC/03 Collect data on the number of addresses (recipients) as well
3519 as the number of messages in eximstats.
3521 TF/01 Correct an error in the documentation for the redirect router. Exim
3522 does (usually) call initgroups() when daemonizing.
3524 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3525 with consistent privilege compared to when running as a daemon.
3527 TF/03 Note in the spec that $authenticated_id is not set for local
3528 submissions from trusted users.
3530 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3531 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3533 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3534 by adding some example configuration directives to the default
3535 configuration file. A little bit of work is required to uncomment the
3536 directives and define how usernames and passwords are checked, but
3537 there is now a framework to start from.
3539 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3540 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3541 without this. I don't know how relevant this is to other LDAP libraries.
3543 PH/02 Add the verb name to the "unknown ACL verb" error.
3545 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3547 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3549 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3550 directory not expanded when it should be if an expanded home directory
3551 was set for the address (which is overridden by the transport).
3553 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3556 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3557 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3558 because it is too late at that time, and has no effect.
3560 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3561 security issue with \' (bugzilla #107). I could not use the
3562 PQescapeStringConn() function, because it needs a PGconn value as one of
3565 PH/08 When testing addresses using -bt, indicate those final addresses that
3566 are duplicates that would not cause an additional delivery. At least one
3567 person was confused, thinking that -bt output corresponded to deliveries.
3568 (Suppressing duplicates isn't a good idea as you lose the information
3569 about possibly different redirections that led to the duplicates.)
3571 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3572 systems where poll() doesn't work, in particular OS X.
3574 PH/10 Added more information to debugging output for retry time not reached.
3576 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3577 operations in malware.c.
3579 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3582 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3583 syslog_duplication set false, log lines that would normally be written
3584 both the the main log and to the reject log were not written to syslog at
3587 PH/14 In the default configuration, change the use of "message" in ACL warn
3588 statements to "add_header".
3590 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3591 not followed by a command (e.g. "seen endif").
3593 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3594 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3597 PH/17 Added extra conditions to the default value of delay_warning_condition
3601 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3602 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3603 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3606 The Auto-Submitted: and various List- headers are standardised, whereas I
3607 don't think Precedence: ever was.
3609 PH/18 Refactored debugging code in route_finduser() to show more information,
3610 in particular, the error code if getpwnam() issues one.
3612 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3613 This is apparently needed in addition to the PH/07 change above to avoid
3614 any possible encoding problems.
3616 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3617 but not after initializing Perl.
3619 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3620 output them only if debugging. By default they are written stderr,
3621 apparently, which is not desirable.
3623 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3626 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3629 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3631 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3632 authenticated or an ident call has been made. Suppress the default
3633 values for $authenticated_id and $authenticated_sender (but permit -oMai
3634 and -oMas) when testing with -bh.
3636 PH/24 Re-jigged the order of the tests in the default configuration so that the
3637 tests for valid domains and recipients precede the DNS black list and CSA
3638 tests, on the grounds that those ones are more expensive.
3640 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3641 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3642 This bug exists in every version of Exim that I still have, right back to
3645 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3646 However, an attempt to turn on case-sensitivity in a regex key by
3647 including (?-i) didn't work because the subject string was already
3648 lowercased, and the effects were non-intuitive. It turns out that a
3649 one-line patch can be used to allow (?-i) to work as expected.
3655 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3656 other effects) broke the use of negated acl sub-conditions.
3658 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3661 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3662 "Deny" causes Exim to reject the incoming connection with a 554 error.
3663 Unfortunately, if there is a major crisis, such as a disk failure,
3664 tcp-wrappers gives "deny", whereas what one would like would be some
3665 kind of temporary error. A kludge has been added to help with this.
3666 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3667 554 error is used if errno is still zero or contains ENOENT (which occurs
3668 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3671 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3673 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3674 errors. Otherwise a message that provokes a temporary error (when other
3675 messages do not) can cause a whole host to time out.
3677 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3678 addresses were routed directly to files or pipes from a redirect router.
3679 File deliveries just didn't batch; pipe deliveries might have suffered
3682 PH/06 A failure to get a lock for a hints database would erroneously always say
3683 "Failed to get write lock", even when it was really a read lock.
3685 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3686 of 0600. This has been changed to use the value of the lockfile_mode
3687 option (which defaults to 0600).
3689 PH/08 Applied small patch from the Sieve maintainer.
3691 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3692 folder from quota calculations, a direct delivery into this folder messed
3693 up the contents of the maildirsize file. This was because the regex was
3694 used only to exclude .Trash (or whatever) when the size of the mailbox
3695 was calculated. There was no check that a delivery was happening into an
3696 excluded directory. This bug has been fixed by ignoring all quota
3697 processing for deliveries into excluded directories.
3699 PH/10 Added the maildirfolder_create_regex option to appendfile.
3705 PH/01 The code for finding all the local interface addresses on a FreeBSD
3706 system running IPv6 was broken. This may well have applied to all BSD
3707 systems, as well as to others that have similar system calls. The broken
3708 code found IPv4 interfaces correctly, but gave incorrect values for the
3709 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3710 that it would not match correctly against @[] and not recognize the IPv6
3713 PH/02 The ipliteral router was not recognizing addresses of the form user@
3714 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3716 PH/03 Added disable_ipv6.
3718 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3719 decoded form, because it is most often used to construct To: headers
3720 lines in autoreplies, and the decoded form may well be syntactically
3721 invalid. However, $reply_address has leading white space removed, and all
3722 newlines turned into spaces so that the autoreply transport does not
3725 PH/05 If group was specified without a user on a router, and no group or user
3726 was specified on a transport, the group from the router was ignored.
3728 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3729 for visible compile-time settings that can be used to change these
3730 numbers, for those that want even more. Backwards compatibility with old
3731 spool files has been maintained. However, going back to a previous Exim
3732 release will lost any variables that are in spool files.
3734 PH/07 Two small changes when running in the test harness: increase delay when
3735 passing a TCP/IP connection to a new process, in case the original
3736 process has to generate a bounce, and remove special handling of
3737 127.0.0.2 (sic), which is no longer necessary.
3739 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3740 be the same on different OS.
3742 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3745 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3746 whether --show-vars was specified or not
3748 JJ/02 exipick: Added support for new ACL variable spool format introduced
3751 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3752 syntactically invalid From: or Reply-to: line, and a filter used this to
3753 generate an autoreply, and therefore failed to obtain an address for the
3754 autoreply, Exim could try to deliver to a non-existent relative file
3755 name, causing unrelated and misleading errors. What now happens is that
3756 it logs this as a hard delivery error, but does not attempt to create a
3759 PH/11 The exinext utility has a -C option for testing purposes, but although
3760 the given file was scanned by exinext itself; it wasn't being passed on
3761 when Exim was called.
3763 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3764 an end-of-file indication when reading a command response.
3766 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3767 compiled. In many other places in Exim, IPv6 addresses are always
3768 recognized, so I have changed this. It also means that IPv4 domain
3769 literals of the form [IPV4:n.n.n.n] are now always recognized.
3771 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3772 used if the router is not running as root, for example, when verifying at
3773 ACL time, or when using -bh. The debugging output from this situation was
3774 non-existent - all you got was a failure to exec. I have made two
3777 (a) Failures to set uid/gid, the current directory, or a process leader
3778 in a subprocess such as that created by queryprogram now generate
3779 suitable debugging output when -d is set.
3781 (b) The queryprogram router detects when it is not running as root,
3782 outputs suitable debugging information if -d is set, and then runs
3783 the subprocess without attempting to change uid/gid.
3785 PH/15 Minor change to Makefile for building test_host (undocumented testing
3788 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3789 additional section of a DNS packet that returns MX or SRV records.
3790 Instead, it always explicitly searches for A/AAAA records. This avoids
3791 major problems that occur when a DNS server includes only records of one
3792 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3793 fixed another bug: if SRV records were looked up and the corresponding
3794 address records were *not* found in the additional section, the port
3795 values from the SRV records were lost.
3797 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3798 using the correct key (the original address) when searching the retry
3799 rules in order to find which one to use for generating the retry hint.
3801 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3802 adding the default one. Similarly, if it contains a Reply-To: header, the
3803 errors_reply_to option, if set, is not used.
3805 PH/19 When calculating a retry time, Exim used to measure the "time since
3806 failure" by looking at the "first failed" field in the retry record. Now
3807 it does not use this if it is later than than the arrival time of the
3808 message. Instead it uses the arrival time. This makes for better
3809 behaviour in cases where some deliveries succeed, thus re-setting the
3810 "first failed" field. An example is a quota failure for a huge message
3811 when small messages continue to be delivered. Without this change, the
3812 "time since failure" will always be short, possible causing more frequent
3813 delivery attempts for the huge message than are intended.
3814 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3816 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3817 $1, $2, $3) because the numerical variables can be reset during some
3818 expansion items (e.g. "match"), thereby losing the authentication data.
3820 PH/21 Make -bV show the size of off_t variables so that the test suite can
3821 decide whether to run tests for quotas > 2G.
3823 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3824 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3825 filecount value is greater than 2G or if a quota value is greater than 2G
3826 on a system where the size of off_t is not greater than 4, a panic error
3829 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3830 never match. The debug and -bh output now contains an explicit error
3831 message indicating a malformed IPv4 address or mask.
3833 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3834 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3835 PH/23 above applies.
3837 PH/25 Do not write to syslog when running in the test harness. The only
3838 occasion when this arises is a failure to open the main or panic logs
3839 (for which there is an explicit test).
3841 PH/26 Added the /no_tell option to "control=freeze".
3843 PH/27 If a host name lookup failed very early in a connection, for example, if
3844 the IP address matched host_lookup and the reverse lookup yielded a name
3845 that did not have a forward lookup, an error message of the form "no IP
3846 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3847 could be logged. Now it outputs the IP address instead of "NULL".
3849 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3850 allows the sender and the authenticated sender to be set when
3851 submitting a message from within Exim. Since child_open_exim() is
3852 documented for local_scan(), the new function should be too.
3854 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3855 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3856 results in an empty string is now treated as unset.
3858 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3860 PH/31 Added sender_verify_fail logging option.
3862 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3863 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3864 tidied the source and removed it altogether.
3866 PH/33 When a queue run was abandoned because the load average was too high, a
3867 log line was always written; now it is written only if the queue_run log
3868 selector is set. In addition, the log line for abandonment now contains
3869 information about the queue run such as the pid. This is always present
3870 in "start" and "stop" lines but was omitted from the "abandon" line.
3872 PH/34 Omit spaces between a header name and the colon in the error message that
3873 is given when verify = headers_syntax fails (if there are lots of them,
3874 the message gets confusing).
3876 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3877 names, as there are now some PTR records that contain slashes. This check
3878 is only to protect against broken name servers that fall over on strange
3879 characters, so the fact that it applies to all lookups doesn't matter.
3881 PH/36 Now that the new test suite is complete, we can remove some of the
3882 special code in Exim that was needed for the old test suite. For example,
3883 sorting DNS records because real resolvers return them in an arbitrary
3884 order. The new test suite's fake resolver always returns records in the
3887 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3888 bounces) except when queue_only is set, to avoid logging races between
3889 the different processes.
3891 PH/38 Panic-die if .include specifies a non-absolute path.
3893 PH/39 A tweak to the "H" retry rule from its user.
3895 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3896 a label. They prevented compilation on older perls.
3898 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3899 a warning to be raised on newish perls.
3901 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3902 on queue. Changes to match documented behaviour of showing count of
3903 messages matching specified criteria.
3905 PH/40 Changed the default ident timeout from 30s to 5s.
3907 PH/41 Added support for the use of login_cap features, on those BSD systems
3908 that have them, for controlling the resources used by pipe deliveries.
3910 PH/42 The content-scanning code uses fopen() to create files in which to put
3911 message data. Previously it was not paying any attention to the mode of
3912 the files. Exim runs with umask(0) because the rest of the code creates
3913 files with open(), and sets the required mode explicitly. Thus, these
3914 files were ending up world-writeable. This was not a big issue, because,
3915 being within the spool directory, they were not world-accessible. I have
3916 created a function called modefopen, which takes an additional mode
3917 argument. It sets umask(777), creates the file, chmods it to the required
3918 mode, then resets the umask. All the relevant calls to fopen() in the
3919 content scanning code have been changed to use this function.
3921 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3922 to 24 hours. This avoids potential overflow problems when processing G
3923 and H retry rules. I suspect nobody ever tinkers with this value.
3925 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3927 PH/45 When the plaintext authenticator is running as a client, the server's
3928 challenges are checked to ensure they are valid base64 strings. By
3929 default, the authentication attempt is cancelled if an invalid string is
3930 received. Setting client_ignore_invalid_base64 true ignores these errors.
3931 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3932 they are received. Thus, the responses can be made to depend on the
3933 challenges. If an invalid string is ignored, an empty string is placed in
3936 PH/46 Messages that are created by the autoreply transport now contains a
3937 References: header, in accordance with RFCs 2822 and 3834.
3939 PH/47 Added authenticated_sender_force to the smtp transport.
3941 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3943 PH/49 Installed latest patch from the Sieve maintainer.
3945 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3946 was also set, the appendfile transport was unnecessarily scanning a
3947 directory of message files (e.g. for maildir delivery) to find the count
3948 of files (along with the size), even though it did not need this
3949 information. It now does the scan only if it needs to find either the
3950 size of the count of files.
3952 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3954 PH/52 Two bugs concerned with error handling when the smtp transport is
3957 (i) Exim was not creating retry information for temporary errors given
3958 for individual recipients after the DATA command when the smtp transport
3959 was used in LMTP mode. This meant that they could be retried too
3960 frequently, and not timed out correctly.
3962 (ii) Exim was setting the flag that allows error details to be returned
3963 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3964 recipients that were returned after the DATA command.
3966 PH/53 This is related to PH/52, but is more general: for any failing address,
3967 when detailed error information was permitted to be returned to the
3968 sender, but the error was temporary, then after the final timeout, only
3969 "retry timeout exceeded" was returned. Now it returns the full error as
3970 well as "retry timeout exceeded".
3972 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3973 do this, and (what is worse) MTAs that accept it.
3975 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3976 will now be deprecated.
3978 PH/56 New os.c-cygwin from the Cygwin maintainer.
3980 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3981 formats (previously only available in exim formats via -bpr, -bpru,
3982 and -bpra. Now also available in native and exiqgrep formats)
3984 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3985 with very large, slow to parse queues
3987 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3989 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3991 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3992 responses to authentication challenges, though it was showing the
3993 challenges; (ii) I've removed the CR characters from the debug output for
3996 PH/58 Allow for the insertion of a newline as well as a space when a string
3997 is turned into more than one encoded-word during RFC 2047 encoding. The
3998 Sieve code now uses this.
4000 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4001 data_4xx, lost_connection, tls_required.
4003 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4004 message was being sent as an SMTP response.
4006 PH/61 Add -l and -k options to exicyclog.
4008 PH/62 When verifying, if an address was redirected to one new address, so that
4009 verification continued, and the new address failed or deferred after
4010 having set something in $address_data, the value of $address_data was not
4011 passed back to the ACL. This was different to the case when no
4012 redirection occurred. The value is now passed back in both cases.
4014 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4015 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4016 use login_cap.h, so on its own it isn't the distinguishing feature. The
4017 new name refers directly to the setclassresources() function.
4019 PH/65 Added configuration files for NetBSD3.
4021 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4023 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4024 is preferred over IPv4.
4026 PH/68 The bounce_return_message and bounce_return_body options were not being
4027 honoured for bounces generated during the reception of non-SMTP messages.
4028 In particular, this applied to messages rejected by the ACL. This bug has
4029 been fixed. However, if bounce_return_message is true and bounce_return_
4030 body is false, the headers that are returned for a non-SMTP message
4031 include only those that have been read before the error was detected.
4032 (In the case of an ACL rejection, they have all been read.)
4034 PH/69 The HTML version of the specification is now built in a directory called
4035 spec_html instead of spec.html, because the latter looks like a path with
4036 a MIME-type, and this confuses some software.
4038 PH/70 Catch two compiler warnings in sieve.c.
4040 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4041 function verify_get_ident() calls ip_connect() to connect a socket, but
4042 if the "connect()" function timed out, ip_connect() used to close the
4043 socket. However, verify_get_ident() also closes the socket later, and in
4044 between Exim writes to the log, which may get opened at this point. When
4045 the socket was closed in ip_connect(), the log could get the same file
4046 descriptor number as the socket. This naturally causes chaos. The fix is
4047 not to close the socket in ip_connect(); the socket should be closed by
4048 the function that creates it. There was only one place in the code where
4049 this was missing, in the iplookup router, which I don't think anybody now
4050 uses, but I've fixed it anyway.
4052 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4053 well as to direct DNS lookups. Otherwise the handling of names in host
4054 lists is inconsistent and therefore confusing.
4060 PH/01 Two changes to the default runtime configuration:
4062 (1) Move the checks for relay_from_hosts and authenticated clients from
4063 after to before the (commented out) DNS black list checks.
4065 (2) Add control=submission to the relay_from_hosts and authenticated
4066 clients checks, on the grounds that messages accepted by these
4067 statements are most likely to be submissions.
4069 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4071 (1) Generate an error if the third argument for the ${prvs expansion is
4074 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4077 (3) Reset the variables that are obtained from the first argument of
4078 ${prvscheck and used in the second argument before leaving the code,
4079 because their memory is reclaimed, so using them afterwards may do
4082 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4083 one (it's much easier than Tom thought :-).
4085 (5) Because of (4), we can now allow for the use of $prvscheck_result
4086 inside the third argument.
4088 PH/03 For some reason, the default setting of PATH when running a command from
4089 a pipe transport was just "/usr/bin". I have changed it to
4092 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4093 anything to be listed in the output from -bV.
4095 PH/05 When a filter generated an autoreply, the entire To: header line was
4096 quoted in the delivery log line, like this:
4098 => >A.N.Other <ano@some.domain> <original@ddress> ...
4100 This has been changed so that it extracts the operative address. There
4101 may be more than one such address. If so, they are comma-separated, like
4104 => >ano@some.domain,ona@other.domain <original@ddress> ...
4106 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4107 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4108 not being looked up in the rDNS to get a host name, Exim was showing the
4109 IP address twice in Received: lines, even though the IP addresses were
4110 identical. For example:
4112 Received: from [1.2.3.4] (helo=[1.2.3.4])
4114 However, if the real host name was known, it was omitting the HELO data
4115 if it matched the actual IP address. This has been tidied up so that it
4116 doesn't show the same IP address twice.
4118 PH/07 When both +timestamp and +memory debugging was on, the value given by
4119 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4120 called by the debug printing, thereby overwriting the timestamp buffer.
4121 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4123 PH/08 When the original message was included in an autoreply transport, it
4124 always said "this is a copy of the message, including all the headers",
4125 even if body_only or headers_only was set. It now gives an appropriate
4128 PH/09 Applied a patch from the Sieve maintainer which:
4130 o fixes some comments
4131 o adds the (disabled) notify extension core
4132 o adds some debug output for the result of if/elsif tests
4133 o points to the current vacation draft in the documentation
4134 and documents the missing references header update
4138 o fixes a bug in processing the envelope test (when testing
4139 multiple envelope elements, the last element determined the
4142 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4143 Electronic Mail") by including:
4145 Auto-submitted: auto-generated
4147 in the messages that it generates (bounce messages and others, such as
4148 warnings). In the case of bounce messages for non-SMTP messages, there was
4149 also a typo: it was using "Auto_submitted" (underscore instead of
4150 hyphen). Since every message generated by Exim is necessarily in response
4151 to another message, thes have all been changed to:
4153 Auto-Submitted: auto-replied
4155 in accordance with these statements in the RFC:
4157 The auto-replied keyword:
4159 - SHOULD be used on messages sent in direct response to another
4160 message by an automatic process,
4162 - MUST NOT be used on manually-generated messages,
4164 - MAY be used on Delivery Status Notifications (DSNs) and Message
4165 Disposition Notifications (MDNs),
4167 - MUST NOT be used on messages generated by automatic or periodic
4168 processes, except for messages which are automatic responses to
4171 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4172 to the default Received: header definition.
4174 PH/12 Added log selector acl_warn_skipped (default on).
4176 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4177 variables because (a) they are in the wrong storage pool and (b) even if
4178 they were copied, it wouldn't work properly because of the caching.
4180 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4181 checking when decoding. Apparently there are clients that generate
4182 overlong encoded strings. Why am I not surprised?
4184 PH/15 If the first argument of "${if match_address" was not empty, but did not
4185 contain an "@" character, Exim crashed. Now it writes a panic log message
4186 and treats the condition as false.
4188 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4190 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4191 extension "envelope-auth". The code is finished and in agreement with
4192 other implementations, but there is no documentation so far and in fact,
4193 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4194 not changing the active code.
4196 Print executed "if" and "elsif" statements when debugging is used. This
4197 helps a great deal to understand what a filter does.
4199 Document more things not specified clearly in RFC3028. I had all this
4200 sorted out, when out of a sudden new issues came to my mind. Oops."
4202 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4205 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4206 canonical form (as documented). However, after a host name lookup from
4207 the IP address, check_host() was doing a simple string comparison with
4208 addresses acquired from the DNS when checking that the found name did
4209 have the original IP as one of its addresses. Since any found IPv6
4210 addresses are likely to be in abbreviated form, the comparison could
4211 fail. Luckily, there already exists a function for doing the comparison
4212 by converting both addresses to binary, so now that is used instead of
4213 the text comparison.
4215 PH/20 There was another similar case to PH/19, when a complete host name was
4216 given in a host list; looking up its IP address could give an abbreviated
4217 form, whereas the current host's name might or might not be abbreviated.
4218 The same fix has been applied.
4224 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4225 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4228 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4229 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4231 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4233 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4234 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4235 Also, for "Auto-Submitted", treat anything other than "no" as
4236 non-personal, in accordance with RFC 3834. (Previously it treated
4237 anything starting "auto-" as non-personal.)
4239 TF/01 The control=submission/name=... option had a problem with syntax
4240 errors if the name included a slash character. The /name= option
4241 now slurps the rest of the string, so it can include any characters
4242 but it must come last in the list of options (after /sender_retain
4245 PH/05 Some modifications to the interface to the fake nameserver for the new
4253 TK/01 Added the "success_on_redirect" address verification option. See
4254 NewStuff for rationale and an example.
4256 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4258 PH/02 Patch to exigrep to allow it to work on syslog lines.
4260 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4261 fread() to skip over the body file's header line, because in Cygwin the
4262 header line is locked and is inaccessible.
4264 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4265 co-exist for some time) to make it clear that it is the Exim ID that is
4266 referenced, not the Message-ID: header line.
4268 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4269 string_format() function, because snprintf() does not exist on all
4272 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4273 ${stat: expansion item.
4275 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4276 protocol synchronization error", to keep the pedants happy.
4278 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4279 well as for IRIX systems, when gcc is being used. See the host.c source
4282 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4284 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4287 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4288 transport and to the smtp transport in LMTP mode.
4290 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4292 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4293 run a filter in a subprocess. This could lead to confusion in subsequent
4294 lookups in the parent process. There should also be a search_tidyup() at
4295 the end of the subprocess.
4297 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4298 only if the host matched helo_try_verify_hosts, which caused the
4299 verification to occur when the EHLO/HELO command was issued. The ACL just
4300 tested the remembered result. Now, if a previous verification attempt has
4301 not happened, "verify = helo" does it there and then.
4303 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4305 TK/03 Fix log output including CR from clamd.
4307 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4308 exist provoked a memory error which could cause a segfault.
4310 PH/15 Installed PCRE 6.2
4312 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4314 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4315 of the problem. Specifically, suggested +O2 rather than +O1 for the
4318 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4320 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4321 record, the port was not correctly logged when the outgoing_port log
4322 selector was set (it logged the transort's default port).
4324 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4325 fallback_hosts, and "hosts" in the smtp transport.
4327 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4328 host errors such as "Connection refused".
4330 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4331 authentication with radiusclient 0.4.9:
4333 - Error returned from rc_read_config was caught wrongly
4334 - Username/password not passed on to radius server due to wrong length.
4336 The presumption is that some radiusclient API changes for 4.51/PH/17
4337 were not taken care of correctly. The code is still untested by me (my
4338 Linux distribution still has 0.3.2 of radiusclient), but it was
4339 contributed by a Radius user.
4341 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4342 expanding the "port" option of the smtp transport.
4344 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4345 while reading a MIME header. Thanks to Tom Hughes for a patch.
4347 PH/24 Include config.h inside local_scan.h so that configuration settings are
4350 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4351 that in an ACL for RCPT (for example), you can examine exactly what was
4354 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4355 commands, but it was not correctly comparing the address with the actual
4356 client host address. Thus, it would show the EHLO address in Received:
4357 header lines when this was not necessary.
4359 PH/27 Added the % operator to ${eval:}.
4361 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4362 it should be ignoring failures (because with -C, for example, it has lost
4363 privilege). It wasn't ignoring creation failures other than "already
4366 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4369 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4370 because an input line was too long, either on its own, or by virtue of
4371 too many continuations, the temporary file was not being removed, and the
4372 return code was incorrect.
4374 PH/31 Missing "BOOL" in function definition in filtertest.c.
4376 PH/32 Applied Sieve patches from the maintainer.
4378 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4380 PH/33 Added "verify = not_blind".
4382 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4383 Local/Makefile (with some defaults set). These are used in built scripts
4384 such as exicyclog, but they have never been used in the exim_install
4385 script (though there are many overriding facilities there). I have
4386 arranged that the exim_install script now takes note of these two
4389 PH/35 Installed configuration files for Dragonfly.
4391 PH/36 When a locally submitted message by a trusted user did not contain a
4392 From: header, and the sender address was obtained from -f or from an SMTP
4393 MAIL command, and the trusted user did not use -F to supply a sender
4394 name, $originator_name was incorrectly used when constructing a From:
4395 header. Furthermore, $originator_name was used for submission mode
4396 messages from external hosts without From: headers in a similar way,
4397 which is clearly wrong.
4399 PH/37 Added control=suppress_local_fixups.
4401 PH/38 When log_selector = +received_sender was set, and the addition of the
4402 sender made the log line's construction buffer exactly full, or one byte
4403 less than full, an overflow happened when the terminating "\n" was
4406 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4407 when the result of a list match is failure because a DNS lookup failed.
4409 PH/40 RM_COMMAND is now used in the building process.
4411 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4412 the "build-* directories that it finds.
4414 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4415 address in a domain literal was a prefix of an interface address.
4417 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4418 when verifying a sender address, unless rewrite_headers is false.
4420 PH/44 Wrote a long comment about why errors_to addresses are verified as
4421 recipients, not senders.
4423 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4424 the ratelimit ACL was added.
4426 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4428 PH/46 Added extra information about PostgreSQL errors to the error string.
4430 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4431 avoiding the need to install special zones in a real server. This is
4432 backwards compatible; if it can't find the fake resolver, it drops back.
4433 Thus, both old and new test suites can be run.
4435 TF/02 Added util/ratelimit.pl
4437 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4438 clock is set back in time.
4440 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4441 Candler <B.Candler@pobox.com>.
4443 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4444 OK for addresses that are the result of redirections.
4446 PH/48 A number of further additions for the benefit of the new test suite,
4447 including a fake gethostbyname() that interfaces to the fake DNS resolver
4450 TF/06 The fix for widen_domains has also been applied to qualify_single and
4451 search_parents which are the other dnslookup options that can cause
4454 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4457 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4459 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4460 service" if it encountered an MX record with an empty target hostname.
4461 The message is now "an MX or SRV record indicated no SMTP service".
4463 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4464 if the DNS of the sending site is misconfigured. This is quite a
4465 common situation. This change restores the behaviour of treating a
4466 helo verification defer as a failure.
4468 PH/51 If self=fail was set on a router, the bounce message did not include the
4469 actual error message.
4475 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4477 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4478 command itself ended in error, the underlying message about the transport
4479 filter timeout was being overwritten with the pipe command error. Now the
4480 underlying error message should be appended to the second error message.
4482 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4484 PH/02 Reduce the amount of output that "make" produces by default. Full output
4485 can still be requested.
4487 PH/03 The warning log line about a condition test deferring for a "warn" verb
4488 was being output only once per connection, rather than after each
4489 occurrence (because it was using the same function as for successful
4490 "warn" verbs). This seems wrong, so I have changed it.
4492 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4493 it should not have, which might have caused a crash in the right
4494 circumstances, but probably never did.
4496 PH/04 Installed a modified version of Tony Finch's patch to make submission
4497 mode fix the return path as well as the Sender: header line, and to
4498 add a /name= option so that you can make the user's friendly name appear
4501 TF/03 Added the control = fakedefer ACL modifier.
4503 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4504 Mark Lowes for thorough testing.
4506 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4508 TK/03 Merged latest SRS patch from Miles Wilton.
4510 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4511 with the definition in sysexits.h (which is #included earlier).
4512 Fortunately, Exim does not actually use EX_OK. The code used to try to
4513 preserve the sysexits.h value, by assuming that macro definitions were
4514 scanned for macro replacements. I have been disabused of this notion,
4515 so now the code just undefines EX_OK before #including unistd.h.
4517 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4518 in the smtp transport. When a block could not be written in a single
4519 write() function, the timeout was being re-applied to each part-write.
4520 This seems wrong - if the receiver was accepting one byte at a time it
4521 would take for ever. The timeout is now adjusted when this happens. It
4522 doesn't have to be particularly precise.
4524 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4525 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4527 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4528 without a domain if the check to <postmaster@domain> fails.
4530 SC/01 Eximstats: added -xls and the ability to specify output files
4531 (patch written by Frank Heydlauf).
4533 SC/02 Eximstats: use FileHandles for outputting results.
4535 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4537 SC/04 Eximstats: fixed display of large numbers with -nvr option
4539 SC/05 Eximstats: fixed merging of reports with empty tables.
4541 SC/06 Eximstats: added the -include_original_destination flag
4543 SC/07 Eximstats: removed tabs and trailing whitespace.
4545 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4547 TK/06 MBOX spool code: Add real "From " MBOX separator line
4548 so the .eml file is really in mbox format (even though
4549 most programs do not really care). Patch from Alex Miller.
4551 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4552 The latter is generated from $received_to and is only set if the
4553 message has one envelope recipient. SA can use these headers,
4554 obviously out-of-the-box. Patch from Alex Miller.
4556 PH/08 The ${def test on a variable was returning false if the variable's
4557 value was "0", contrary to what the specification has always said!
4558 The result should be true unless the variable is empty.
4560 PH/09 The syntax error of a character other than { following "${if
4561 def:variable_name" (after optional whitespace) was not being diagnosed.
4562 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4563 accidental colon was present, for example, could give incorrect results.
4565 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4566 result is used (not including appendfile, where other changes are about
4569 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4570 This involved changing a lot of size variables from int to off_t. It
4571 should work with maildirs and everything.
4573 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4574 spamd dying while we are connected to it.
4576 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4579 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4580 function for BDB 4.3.
4582 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4584 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4585 the sources for PCRE have been re-arranged and more files are now
4588 PH/15 The code I had for printing potentially long long variables in PH/11
4589 above was not the best (it lost precision). The length of off_t variables
4590 is now inspected at build time, and an appropriate printing format (%ld
4591 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4592 to be "long long int" or "long int". This is needed for the internal
4593 formatting function string_vformat().
4595 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4596 the configuration file to be ":syslog", then the script "guesses" where
4597 the logs files are, rather than using the compiled in default. In our
4598 case the guess is not the same as the compiled default, so the script
4599 suddenly stopped working when I started to use syslog. The patch checks
4600 to see if log_file_path is "". If so, it attempts to read it from exim
4601 with no configuration file to get the compiled in version, before it
4602 falls back to the previous guessing code."
4604 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4605 implementing BATV in an Exim configuration. See NewStuff for the gory
4608 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4609 Makefile that are specific to HP-UX.
4611 PH/18 If the "use_postmaster" option was set for a recipient callout together
4612 with the "random" option, the postmaster address was used as the MAIL
4613 FROM address for the random test, but not for the subsequent recipient
4614 test. It is now used for both.
4616 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4617 patch removes a few documentation additions to RFC 3028, because the
4618 latest draft now contains them. It adds the new en;ascii-case comparator
4619 and a new error check for 8bit text in MIME parts. Comparator and
4620 require names are now matched exactly. I enabled the subaddress
4621 extension, but it is not well tested yet (read: it works for me)."
4623 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4624 rework some of the code of TK/09 above to avoid the hardwired use of
4625 "%lld" and "long long". Replaced the call to snprintf() with a call to
4628 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4629 records point to non-existent hosts", "retry timeout exceeded", and
4630 "retry time not reached for any host after a long failure period".
4632 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4633 experimental DomainKeys support:
4635 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4636 (2) On an error such as an illegally used "control", the wrong name for
4637 the control was given.
4639 These problems did NOT occur unless DomainKeys support was compiled.
4641 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4643 PH/24 Added ${if match_ip condition.
4645 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4646 databases so that it will be absolutely obvious if a crash occurs in the
4647 DB library. This is a regular occurrence (often caused by mis-matched
4650 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4651 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4652 picked up on a user's system that detects such things. There doesn't seem
4653 to be a gcc warning option for this - only an attribute that has to be
4654 put on the function's prototype. It seems that in Fedora Core 4 they have
4655 set this on a number of new functions. No doubt there will be more in due
4658 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4659 specify a transport. However, if an address that was verified by such a
4660 router was the subject of a callout, Exim crashed because it tried to
4661 read the rcpt_include_affixes from the non-existent transport. Now it
4662 just assumes that the setting of that option is false. This bug was
4663 introduced by 4.51/PH/31.
4665 PH/28 Changed -d+all to exclude +memory, because that information is very
4666 rarely of interest, but it makes the output a lot bigger. People tend to
4667 do -d+all out of habit.
4669 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4670 code in os-type was giving problems when libc.so lives in lib64, like on
4673 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4674 aren't the modern standard, and it seems that some systems' include files
4675 don't always have them. Exim was already checking for some of the newer
4676 ones like T_AAAA, and defining it itself. I've added checks for all the
4677 record types that Exim uses.
4679 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4680 not automatically generating a new one, as it is supposed to. This
4681 prevented TLS from working. If the file did exist, but contained invalid
4682 data, a new version was generated, as expected. It was only the case of a
4683 non-existent file that was broken.
4685 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4686 with a change in libdomainkeys > 0.64.
4688 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4689 from DNS. If the selector record carries the flag, it now has
4690 precedence over the domain-wide flag.
4692 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4694 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4695 the use of a port name that isn't defined in /etc/services) occurred, the
4696 message was deferred as in a normal delivery, and thus remained on the
4697 spool, instead of being failed because of the mua_wrapper setting. This
4698 is now fixed, and I tidied up some of the mua_wrapper messages at the
4701 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4702 the messages in a hash of arrays rather than using individual hashes.
4703 This is a bit cleaner and results in dramatic memory savings, albeit
4704 at a slight CPU cost.
4706 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4707 as requested by Marc Sherman.
4709 SC/10 Eximstats: added histograms for user specified patterns as requested
4712 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4714 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4715 fopen() in the content-scanning modules that did not already have it.
4721 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4722 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4724 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4726 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4728 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4729 to test Sieve filters that use "vacation".
4731 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4732 that changes the way the GnuTLS parameters are stored in the cache file.
4733 The new format can be generated externally. For backward compatibility,
4734 if the data in the cache doesn't make sense, Exim assumes it has read an
4735 old-format file, and it generates new data and writes a new file. This
4736 means that you can't go back to an older release without removing the
4739 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4740 work if there are any delivery delays because "one_time" forces the
4741 parent to be marked "delivered", so its unseen clone is never tried
4742 again. For this reason, Exim now forbids the simultaneous setting of
4745 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4746 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4747 entry if you want to know the details. The fix, however, neglected to
4748 consider the case when local delivery batching is involved. The test for
4749 "previously delivered" was not happening when checking to see if an
4750 address could be batched with a previous (undelivered) one; under
4751 certain circumstances this could lead to multiple deliveries to the same
4754 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4755 in its include files, and this causes problems building Exim.
4757 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4758 header_syntax) but Exim was just ignoring anything given after a slash.
4759 In particular, this caused confusion with an attempt to use "verify =
4760 reverse_host_lookup/defer_ok". An error is now given when options are
4761 supplied for verify items that do not have them. (Maybe reverse_host_
4762 lookup should have a defer_ok option, but that's a different point.)
4764 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4765 defined by RFC 821) to 2048, because there were problems with some AUTH
4766 commands, and RFC 1869 says the size should be increased for extended
4767 SMTP commands that take arguments.
4769 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4772 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4773 "unknown" error; now it says that the functionality isn't in the binary.
4775 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4776 an address' error message when a string expansion fails (syntax or
4777 whatever). Otherwise the password may appear in the log. Following change
4778 PH/42 below, there is no longer a chance of it appearing in a bounce
4781 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4783 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4784 of its addresses was ever tried. (Bugzilla bug #2.)
4786 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4787 the result incorrectly in the debug output. (It correctly added a newline
4788 to what was transported.)
4790 TF/01 Added $received_time.
4792 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4793 commented out examples of how to interface to a virus scanner and to
4794 SpamAssassin. Also added commented examples of av_scanner and
4795 spamd_address settings.
4797 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4798 and controls are allowed in which ACLs. There were a couple of minor
4799 errors. Some of the entries in the conditions table (which is a table of
4800 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4801 negation of where the condition IS allowed.
4803 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4805 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4806 header file does not have a version number, so I've had to invent a new
4807 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4808 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4809 radiusclient), but it was contributed by a Radius user.
4811 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4812 files or directories, for OpenSSL.
4814 PH/19 When an Exim process that is running as root has to create an Exim log
4815 file, it does so in a subprocess that runs as exim:exim so as to get the
4816 ownership right at creation (otherwise, other Exim processes might see
4817 the file with the wrong ownership). There was no test for failure of this
4818 fork() call, which would lead to the process getting stuck as it waited
4819 for a non-existent subprocess. Forks do occasionally fail when resources
4820 run out. I reviewed all the other calls to fork(); they all seem to check
4823 PH/20 When checking for unexpected SMTP input at connect time (before writing
4824 the banner), Exim was not dealing correctly with a non-positive return
4825 from the read() function. If the client had disconnected by this time,
4826 the result was a log entry for a synchronization error with an empty
4827 string after "input=" when read() returned zero. If read() returned -1
4828 (an event I could not check), uninitialized data bytes were printed.
4829 There were reports of junk text (parts of files, etc) appearing after
4832 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4834 PH/22 Added support for macro redefinition, and (re)definition in between
4835 driver and ACL definitions.
4837 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4838 forgetting to use the resulting value; it was using the unexpanded value.
4840 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4841 hadn't been configured. The fix is from Juergen Kreileder, who
4842 understands it better than I do:
4844 "Here's what I see happening with three configured cyrus_sasl
4845 authenticators configured (plain, login, cram-md5):
4847 On startup auth_cyrus_sasl_init() gets called for each of these.
4848 This means three calls to sasl_listmech() without a specified mech_list.
4849 => SASL tests which mechs of all available mechs actually work
4850 => three warnings about OTP not working
4851 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4853 With the patch, sasl_listmech() also gets called three times. But now
4854 SASL's mech_list option is set to the server_mech specified in the the
4855 authenticator. Or in other words, the answer from sasl_listmech()
4856 gets limited to just the mech you're testing for (which is different
4858 => the return list contains just 'plain' or 'login', 'cram-md5' or
4859 nothing depending on the value of ob->server_mech.
4861 I've just tested the patch: Authentication still works fine,
4862 unavailable mechs specified in the exim configuration are still
4863 caught, and the auth.log warnings about OTP are gone."
4865 PH/25 When debugging is enabled, the contents of the command line are added
4866 to the debugging output, even when log_selector=+arguments is not
4869 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4870 answer is "GNU", and only if the return is "GNU/something" is the answer
4873 PH/27 $acl_verify_message is now set immediately after the failure of a
4874 verification in an ACL, and so is available in subsequent modifiers. In
4875 particular, the message can be preserved by coding like this:
4877 warn !verify = sender
4878 set acl_m0 = $acl_verify_message
4880 Previously, $acl_verify_message was set only while expanding "message"
4881 and "log_message" when a very denied access.
4883 PH/28 Modified OS/os.c-Linux with
4885 -#ifndef OS_LOAD_AVERAGE
4886 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4888 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4889 nomenclature these days.)
4891 PH/29 Installed patch from the Sieve maintainer that adds the options
4892 sieve_useraddress and sieve_subaddress to the redirect router.
4894 PH/30 In these circumstances:
4895 . Two addresses routed to the same list of hosts;
4896 . First host does not offer TLS;
4897 . First host accepts first address;
4898 . First host gives temporary error to second address;
4899 . Second host offers TLS and a TLS session is established;
4900 . Second host accepts second address.
4901 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4902 and peerdn, if requested) that were in fact used only for the second
4905 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4906 attention to any local part prefix or suffix that was matched by the
4907 router that accepted the address. It now behaves in the same way as it
4908 does for delivery: the affixes are removed from the local part unless
4909 rcpt_include_affixes is set on the transport.
4911 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4912 timeout during the DATA phase of an incoming message.
4914 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4915 applied a patch sanctioned by the Sieve maintainer.
4917 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4918 the uid or gid is negative. A case of a negative gid caused this to be
4919 noticed. The fix allows for either to be negative.
4921 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4922 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4925 PH/36 The change PH/12 above was broken. Fixed it.
4927 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4928 the grounds that routing the same address twice would always produce the
4929 same answer. This might have been true once, but it is certainly no
4930 longer true now. Routing a child address may depend on the previous
4931 routing that produced that child. Some complicated redirection strategies
4932 went wrong when messages had multiple recipients, and made Exim's
4933 behaviour dependent on the order in which the addresses were given.
4935 I have moved the duplicate checking until after the routing is complete.
4936 Exim scans the addresses that are assigned to local and remote
4937 transports, and removes any duplicates. This means that more work will be
4938 done, as duplicates will always all be routed, but duplicates are
4939 presumably rare, so I don't expect this is of any significance.
4941 For deliveries to pipes, files, and autoreplies, the duplicate checking
4942 still happens during the routing process, since they are not going to be
4945 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4946 It corrects a timeout issue with spamd. This is Ian's comment: "The
4947 background is that sometimes spamd either never reads data from a
4948 connection it has accepted, or it never writes response data. The exiscan
4949 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4950 blindly assumes that writes won't block so it may never time out."
4952 PH/39 Allow G after quota size as well as K and M.
4954 PH/40 The value set for $authenticated_id in an authenticator may not contain
4955 binary zeroes or newlines because the value is written to log lines and
4956 to spool files. There was no check on this. Now the value is run through
4957 the string_printing() function so that such characters are converted to
4958 printable escape sequences.
4960 PH/41 $message_linecount is a new variable that contains the total number of
4961 lines in the message. Compare $body_linecount, which is the count for the
4964 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4965 bounce and delay warning messages, except in certain special cases, which
4968 (a) An SMTP error message from a remote host;
4969 (b) A message specified in a :fail: redirection;
4970 (c) A message specified in a "fail" command in a system filter;
4971 (d) A message specified in a FAIL return from the queryprogram router;
4972 (e) A message specified by the cannot_route_message router option.
4974 In these cases only, Exim does include the error details in bounce and
4975 warning messages. There are also a few cases where bland messages such
4976 as "unrouteable address" or "local delivery error" are given.
4978 PH/43 $value is now also set for the "else" part of a ${run expansion.
4980 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4981 being worked on, but at least Exim now implements the latest version to
4984 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4985 process to complete was treated as a delivery failure, a timeout while
4986 writing the message to the pipe was logged, but erroneously treated as a
4987 successful delivery. Such timeouts include transport filter timeouts. For
4988 consistency with the overall process timeout, these timeouts are now
4989 treated as errors, giving rise to delivery failures by default. However,
4990 there is now a new Boolean option for the pipe transport called
4991 timeout_defer, which, if set TRUE, converts the failures into defers for
4992 both kinds of timeout. A transport filter timeout is now identified in
4995 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4996 systems where "make" and "gmake" are different, calling "gmake" at top
4997 level broke things. I've arranged for the value of $(MAKE) to be passed
4998 from the Makefile to this script so that it can call the same version of
5002 A note about Exim versions 4.44 and 4.50
5003 ----------------------------------------
5005 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5006 changes of various kinds. As a consequence, a big documentation update was
5007 needed. This delayed the release for rather longer than seemed good, especially
5008 in the light of a couple of (minor) security issues. Therefore, the changes
5009 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5010 release. So 4.44 and 4.50 are in effect two different branches that both start
5013 I have left the 4.50 change log unchanged; it contains all the changes since
5014 4.43. The change log for 4.44 is below; many of its items are identical to
5015 those for 4.50. This seems to be the most sensible way to preserve the
5016 historical information.
5022 1. Minor wording change to the doc/README.SIEVE file.
5024 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5025 computation of the current number of files was incorrect.
5027 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5028 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5031 4. Give more explanation in the error message when the command for a transport
5032 filter fails to execute.
5034 5. There are several places where Exim runs a non-Exim command in a
5035 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5036 was being done only for the command run by the queryprogram router. It is
5037 now done for all such subprocesses. The other cases are: ${run, transport
5038 filters, and the commands run by the lmtp and pipe transports.
5040 6. Added CONFIGURE_GROUP build-time option.
5042 7. Some older OS have a limit of 256 on the maximum number of file
5043 descriptors. Exim was using setrlimit() to set 1000 as a large value
5044 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5045 systems. I've change it so that if it can't get 1000, it tries for 256.
5047 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5048 was an oversight, and furthermore, ever since the addition of extra
5049 controls (e.g. 4.43/32), the checks on when to allow different forms of
5050 "control" were broken. There should now be diagnostics for all cases when a
5051 control that does not make sense is encountered.
5053 9. Added the /retain_sender option to "control=submission".
5055 10. $recipients is now available in the predata ACL (oversight).
5057 11. Tidy the search cache before the fork to do a delivery from a message
5058 received from the command line. Otherwise the child will trigger a lookup
5059 failure and thereby defer the delivery if it tries to use (for example) a
5060 cached ldap connection that the parent has called unbind on.
5062 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5063 of $address_data from the recipient verification was clobbered by the
5064 sender verification.
5066 13. The value of address_data from a sender verification is now available in
5067 $sender_address_data in subsequent conditions in the ACL statement.
5069 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5071 15. Added a new option "connect=<time>" to callout options, to set a different
5074 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5075 was its contents. (It was OK if the option was not defined at all.)
5077 17. A "Completed" log line is now written for messages that are removed from
5078 the spool by the -Mrm option.
5080 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5081 information about exactly what failed.
5083 19. Added -dd to debug only the daemon process.
5085 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5086 handles timeouts, both on the server side and network timeouts. Renamed the
5087 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5089 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5090 It is now set to "smtps".
5092 22. $host_address is now set to the target address during the checking of
5093 ignore_target_hosts.
5095 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5096 being passed; this would have caused $sender_host_name to have been used if
5097 matching the list had actually called for a host name (not very likely,
5098 since this list is usually IP addresses). A host name is now passed as
5101 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5102 code that specifies a non-restarting handler (typically sigaction() in
5103 modern systems) in an attempt to fix a rare and obscure crash bug.
5105 25. Narrowed the window for a race in the daemon that could cause it to ignore
5106 SIGCHLD signals. This is not a major problem, because they are used only to
5107 wake it up if nothing else does.
5109 26. A malformed maildirsize file could cause Exim to calculate negative values
5110 for the mailbox size or file count. Odd effects could occur as a result.
5111 The maildirsize information is now recalculated if the size or filecount
5114 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5115 support for a long time. Removed HAVE_SYS_VFS_H.
5117 28. Installed the latest version of exipick from John Jetmore.
5119 29. In an address list, if the pattern was not a regular expression, an empty
5120 subject address (from a bounce message) matched only if the pattern was an
5121 empty string. Non-empty patterns were not even tested. This was the wrong
5122 because it is perfectly reasonable to use an empty address as part of a
5123 database query. An empty address is now tested by patterns that are
5124 lookups. However, all the other forms of pattern expect the subject to
5125 contain a local part and a domain, and therefore, for them, an empty
5126 address still always fails if the pattern is not itself empty.
5128 30. Exim went into a mad DNS loop when attempting to do a callout where the
5129 host was specified on an smtp transport, and looking it up yielded more
5130 than one IP address.
5132 31. Re-factored the code for checking spool and log partition space into a
5133 function that finds that data and another that does the check. The former
5134 is then used to implement four new variables: $spool_space, $log_space,
5135 $spool_inodes, and $log_inodes.
5137 32. The RFC2047 encoding function was originally intended for short strings
5138 such as real names; it was not keeping to the 75-character limit for
5139 encoded words that the RFC imposes. It now respects the limit, and
5140 generates multiple encoded words if necessary. To be on the safe side, I
5141 have increased the buffer size for the ${rfc2047: expansion operator from
5144 33. It is now permitted to omit both strings after an "if" condition; if the
5145 condition is true, the result is "true". As before, when the second string
5146 is omitted, a false condition yields an empty string. This makes it less
5147 cumbersome to write custom ACL and router conditions.
5149 34. Failure to deliver a bounce message always caused it to be frozen, even if
5150 there was an errors_to setting on the router. The errors_to setting is now
5153 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5154 canonical form (fully expanded) before being placed in
5155 $sender_host_address.
5157 36. The table in the code that translates DNS record types into text (T_A to
5158 "A" for instance) was missing entries for NS and CNAME. It is just possible
5159 that this could have caused confusion if both these types were looked up
5160 for the same domain, because the text type is used as part of Exim's
5161 per-process caching. But the chance of anyone hitting this buglet seems
5164 37. The dnsdb lookup has been extended in a number of ways.
5166 (1) There is a new type, "zns", which walks up the domain tree until it
5167 finds some nameserver records. It should be used with care.
5169 (2) There is a new type, "mxh", which is like "mx" except that it returns
5170 just the host names, not the priorities.
5172 (3) It is now possible to give a list of domains (or IP addresses) to be
5173 looked up. The behaviour when one of the lookups defers can be
5174 controlled by a keyword.
5176 (4) It is now possible to specify the separator character for use when
5177 multiple records are returned.
5179 38. The dnslists ACL condition has been extended: it is now possible to supply
5180 a list of IP addresses and/or domains to be looked up in a particular DNS
5183 39. Added log_selector=+queue_time_overall.
5185 40. When running the queue in the test harness, wait just a tad after forking a
5186 delivery process, to get repeatability of debugging output.
5188 41. Include certificate and key file names in error message when GnuTLS fails
5189 to set them up, because the GnuTLS error message doesn't include the name
5190 of the failing file when there is a problem reading it.
5192 42. Allow both -bf and -bF in the same test run.
5194 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5196 44. The "Exiscan patch" is now merged into the mainline Exim source.
5198 45. Sometimes the final signoff response after QUIT could fail to get
5199 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5200 < 0 before doing a fflush(). This bug looks as though it goes back to the
5201 introduction of TLS in release 3.20, but "sometimes" must have been rare
5202 because the tests only now provoked it.
5204 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5205 (this can affect the format of dates).
5207 47. exim_tidydb, when checking for the continued existence of a message for
5208 which it has found a message-specific retry record, was not finding
5209 messages that were in split spool directories. Consequently, it was
5210 deleting retry records that should have stayed in existence.
5212 48. Steve fixed some bugs in eximstats.
5214 49. The SPA authentication driver was not abandoning authentication and moving
5215 on to the next authenticator when an expansion was forced to fail,
5216 contradicting the general specification for all authenticators. Instead it
5217 was generating a temporary error. It now behaves as specified.
5219 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5220 (the order specifies the preference for clients). The order is now AES256,
5221 AES128, 3DES, ARCFOUR128.
5223 51. Small patch to Sieve code - explicitly set From: when generating an
5226 52. Exim crashed if a remote delivery caused a very long error message to be
5227 recorded - for instance if somebody sent an entire SpamAssassin report back
5228 as a large number of 550 error lines. This bug was coincidentally fixed by
5229 increasing the size of one of Exim's internal buffers (big_buffer) that
5230 happened as part of the Exiscan merge. However, to be on the safe side, I
5231 have made the code more robust (and fixed the comments that describe what
5234 53. Now that there can be additional text after "Completed" in log lines (if
5235 the queue_time_overall log selector is set), a one-byte patch to exigrep
5236 was needed to allow it to recognize "Completed" as not the last thing in
5239 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5240 patch that reportedly fixes this has been added. I am not expert enough to
5241 create a test for it. This is what the patch creator wrote:
5243 "I found a little strange behaviour of ldap code when working with
5244 Windows 2003 AD Domain, where users was placed in more than one
5245 Organization Units. When I tried to give exim partial DN, the exit code
5246 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5247 But simultaneously result of request was absolutely normal ldap result,
5248 so I produce this patch..."
5250 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5251 REFERENCE, so I have modified the code to exclude the patch when that macro
5254 55. Some experimental protocols are using DNS PTR records for new purposes. The
5255 keys for these records are domain names, not reversed IP addresses. The
5256 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5257 leaves it alone. Component reversal etc. now happens only for IP addresses.
5260 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5262 57. Double the size of the debug message buffer (to 2048) so that more of very
5263 long debug lines gets shown.
5265 58. The exicyclog utility now does better if the number of log files to keep
5266 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5268 59. Two changes related to the smtp_active_hostname option:
5270 (1) $smtp_active_hostname is now available as a variable.
5271 (2) The default for smtp_banner uses $smtp_active_hostname instead
5272 of $primary_hostname.
5274 60. The host_aton() function is supposed to be passed a string that is known
5275 to be a valid IP address. However, in the case of IPv6 addresses, it was
5276 not checking this. This is a hostage to fortune. Exim now panics and dies
5277 if the condition is not met. A case was found where this could be provoked
5278 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5279 components; fortuitously, this particular loophole had already been fixed
5280 by change 4.50/55 above.
5282 If there are any other similar loopholes, the new check in host_aton()
5283 itself should stop them being exploited. The report I received stated that
5284 data on the command line could provoke the exploit when Exim was running as
5285 exim, but did not say which command line option was involved. All I could
5286 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5287 running as the user.
5290 61. There was a buffer overflow vulnerability in the SPA authentication code
5291 (which came originally from the Samba project). I have added a test to the
5292 spa_base64_to_bits() function which I hope fixes it.
5295 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5296 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5298 63. The daemon start-up calls getloadavg() while still root for those OS that
5299 need the first call to be done as root, but it missed one case: when
5300 deliver_queue_load_max is set with deliver_drop_privilege. This is
5301 necessary for the benefit of the queue runner, because there is no re-exec
5302 when deliver_drop_privilege is set.
5304 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5305 This has been fixed.
5307 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5308 was in use, was not putting the data itself into the right store pool;
5309 consequently, it could be overwritten for a subsequent message in the same
5310 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5313 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5315 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5316 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5317 the function were treating the return as a boolean value, which happened to
5318 work because 0=false and not-0=true, but is not correct code.
5320 68. The host_aton() function was not handling scoped IPv6 addresses (those
5321 with, for example, "%eth0" on the end) correctly.
5323 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5324 negated items (that is, ~something) in unsigned ints. Some compilers
5325 apparently mutter when there is no cast.
5327 70. If an address verification called from an ACL failed, and did not produce a
5328 user-specific message (i.e. there was only a "system" message), nothing was
5329 put in $acl_verify_message. In this situation, it now puts the system
5332 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5333 session; change 4.31/43 added the unwanted input to the log line - except
5334 that it did not do this in the start of session case. It now does.
5336 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5337 This is wrong and can cause the other end to generate a synchronization
5338 error if it is another Exim or anything else that does the synchronization
5339 check. A QUIT command is no longer sent after a timeout.
5341 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5342 during host lookups.
5344 74. The defer_ok option of callout verification was not working if it was used
5345 when verifying addresses in header lines, that is, for this case:
5347 verify = header_sender/callout=defer_ok
5349 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5350 those file descriptors could be used for SMTP connections. If anything
5351 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5352 could be sent to the SMTP client, causing chaos. The daemon now opens
5353 stdin, stdout, and stderr to /dev/null when it puts itself into the
5356 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5357 log by default. The user can override this with suitable Perl magic.
5359 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5360 the log message when discard triggers, was not working for the DATA ACL or
5361 for the non-SMTP ACL.
5363 78. Error message wording change in sieve.c.
5365 79. If smtp_accept_max_per_host was set, the number of connections could be
5366 restricted to fewer than expected, because the daemon was trying to set up
5367 a new connection before checking whether the processes handling previous
5368 connections had finished. The check for completed processes is now done
5369 earlier. On busy systems, this bug wouldn't be noticed because something
5370 else would have woken the daemon, and it would have reaped the completed
5373 80. If a message was submitted locally by a user whose login name contained one
5374 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5375 It caused a spool format error. I have fixed the spool reading code. A
5376 related problem was that the "from" clause in the Received: line became
5377 illegal because of the space(s). It is now covered by ${quote_local_part.
5379 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5382 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5384 83. Patch from Sieve maintainer to fix unterminated string problem in
5385 "vacation" handling.
5387 84. Some minor changes to the Linux configuration files to help with other
5388 OS variants using glibc.
5390 85. One more patch for Sieve to update vacation handling to latest spec.
5393 ----------------------------------------------------
5394 See the note above about the 4.44 and 4.50 releases.
5395 ----------------------------------------------------
5401 1. Change 4.43/35 introduced a bug that caused file counts to be
5402 incorrectly computed when quota_filecount was set in an appendfile
5405 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5406 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5409 3. Give more explanation in the error message when the command for a transport
5410 filter fails to execute.
5412 4. There are several places where Exim runs a non-Exim command in a
5413 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5414 was being done only for the command run by the queryprogram router. It is
5415 now done for all such subprocesses. The other cases are: ${run, transport
5416 filters, and the commands run by the lmtp and pipe transports.
5418 5. Some older OS have a limit of 256 on the maximum number of file
5419 descriptors. Exim was using setrlimit() to set 1000 as a large value
5420 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5421 systems. I've change it so that if it can't get 1000, it tries for 256.
5423 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5424 was an oversight, and furthermore, ever since the addition of extra
5425 controls (e.g. 4.43/32), the checks on when to allow different forms of
5426 "control" were broken. There should now be diagnostics for all cases when a
5427 control that does not make sense is encountered.
5429 7. $recipients is now available in the predata ACL (oversight).
5431 8. Tidy the search cache before the fork to do a delivery from a message
5432 received from the command line. Otherwise the child will trigger a lookup
5433 failure and thereby defer the delivery if it tries to use (for example) a
5434 cached ldap connection that the parent has called unbind on.
5436 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5437 of $address_data from the recipient verification was clobbered by the
5438 sender verification.
5440 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5441 was its contents. (It was OK if the option was not defined at all.)
5443 11. A "Completed" log line is now written for messages that are removed from
5444 the spool by the -Mrm option.
5446 12. $host_address is now set to the target address during the checking of
5447 ignore_target_hosts.
5449 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5450 being passed; this would have caused $sender_host_name to have been used if
5451 matching the list had actually called for a host name (not very likely,
5452 since this list is usually IP addresses). A host name is now passed as
5455 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5456 code that specifies a non-restarting handler (typically sigaction() in
5457 modern systems) in an attempt to fix a rare and obscure crash bug.
5459 15. Narrowed the window for a race in the daemon that could cause it to ignore
5460 SIGCHLD signals. This is not a major problem, because they are used only to
5461 wake it up if nothing else does.
5463 16. A malformed maildirsize file could cause Exim to calculate negative values
5464 for the mailbox size or file count. Odd effects could occur as a result.
5465 The maildirsize information is now recalculated if the size or filecount
5468 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5469 support for a long time. Removed HAVE_SYS_VFS_H.
5471 18. Updated exipick to current release from John Jetmore.
5473 19. Allow an empty sender to be matched against a lookup in an address list.
5474 Previously the only cases considered were a regular expression, or an
5477 20. Exim went into a mad DNS lookup loop when doing a callout where the
5478 host was specified on the transport, if the DNS lookup yielded more than
5481 21. The RFC2047 encoding function was originally intended for short strings
5482 such as real names; it was not keeping to the 75-character limit for
5483 encoded words that the RFC imposes. It now respects the limit, and
5484 generates multiple encoded words if necessary. To be on the safe side, I
5485 have increased the buffer size for the ${rfc2047: expansion operator from
5488 22. Failure to deliver a bounce message always caused it to be frozen, even if
5489 there was an errors_to setting on the router. The errors_to setting is now
5492 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5493 canonical form (fully expanded) before being placed in
5494 $sender_host_address.
5496 24. Updated eximstats to version 1.33
5498 25. Include certificate and key file names in error message when GnuTLS fails
5499 to set them up, because the GnuTLS error message doesn't include the name
5500 of the failing file when there is a problem reading it.
5502 26. Expand error message when OpenSSL has problems setting up cert/key files.
5505 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5506 (this can affect the format of dates).
5508 28. exim_tidydb, when checking for the continued existence of a message for
5509 which it has found a message-specific retry record, was not finding
5510 messages that were in split spool directories. Consequently, it was
5511 deleting retry records that should have stayed in existence.
5513 29. eximstats updated to version 1.35
5514 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5515 1.35 - bugfix such that pie charts by volume are generated correctly
5517 30. The SPA authentication driver was not abandoning authentication and moving
5518 on to the next authenticator when an expansion was forced to fail,
5519 contradicting the general specification for all authenticators. Instead it
5520 was generating a temporary error. It now behaves as specified.
5522 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5523 (the order specifies the preference for clients). The order is now AES256,
5524 AES128, 3DES, ARCFOUR128.
5526 31. Small patch to Sieve code - explicitly set From: when generating an
5529 32. Exim crashed if a remote delivery caused a very long error message to be
5530 recorded - for instance if somebody sent an entire SpamAssassin report back
5531 as a large number of 550 error lines. This bug was coincidentally fixed by
5532 increasing the size of one of Exim's internal buffers (big_buffer) that
5533 happened as part of the Exiscan merge. However, to be on the safe side, I
5534 have made the code more robust (and fixed the comments that describe what
5537 33. Some experimental protocols are using DNS PTR records for new purposes. The
5538 keys for these records are domain names, not reversed IP addresses. The
5539 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5540 leaves it alone. Component reversal etc. now happens only for IP addresses.
5543 34. The host_aton() function is supposed to be passed a string that is known
5544 to be a valid IP address. However, in the case of IPv6 addresses, it was
5545 not checking this. This is a hostage to fortune. Exim now panics and dies
5546 if the condition is not met. A case was found where this could be provoked
5547 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5548 components; fortuitously, this particular loophole had already been fixed
5549 by change 4.50/55 or 4.44/33 above.
5551 If there are any other similar loopholes, the new check in host_aton()
5552 itself should stop them being exploited. The report I received stated that
5553 data on the command line could provoke the exploit when Exim was running as
5554 exim, but did not say which command line option was involved. All I could
5555 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5556 running as the user.
5559 35. There was a buffer overflow vulnerability in the SPA authentication code
5560 (which came originally from the Samba project). I have added a test to the
5561 spa_base64_to_bits() function which I hope fixes it.
5564 36. The daemon start-up calls getloadavg() while still root for those OS that
5565 need the first call to be done as root, but it missed one case: when
5566 deliver_queue_load_max is set with deliver_drop_privilege. This is
5567 necessary for the benefit of the queue runner, because there is no re-exec
5568 when deliver_drop_privilege is set.
5570 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5571 was in use, was not putting the data itself into the right store pool;
5572 consequently, it could be overwritten for a subsequent message in the same
5573 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5576 38. Sometimes the final signoff response after QUIT could fail to get
5577 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5578 < 0 before doing a fflush(). This bug looks as though it goes back to the
5579 introduction of TLS in release 3.20, but "sometimes" must have been rare
5580 because the tests only now provoked it.
5586 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5587 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5588 Now it returns an integer. A number of calls were still expecting a T/F
5589 return. Fortuitously, in all cases, the tests worked in OK situations,
5590 which is the norm. However, things would have gone wrong on any write
5591 failures on the smtp file descriptor. This function is used when sending
5592 messages over SMTP and also when doing verify callouts.
5594 2. When Exim is called to do synchronous delivery of a locally submitted
5595 message (the -odf or -odi options), it no longer closes stderr before doing
5598 3. Implemented the mua_wrapper option.
5600 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5602 5. Implemented the functions header_remove(), header_testname(),
5603 header_add_at_position(), and receive_remove_recipient(), and exported them
5606 6. If an ACL "warn" statement specified the addition of headers, Exim already
5607 inserted X-ACL-Warn: at the start if there was no header name. However, it
5608 was not making this test for the second and subsequent header lines if
5609 there were newlines in the string. This meant that an invalid header could
5610 be inserted if Exim was badly configured.
5612 7. Allow an ACL "warn" statement to add header lines at the start or after all
5613 the Received: headers, as well as at the end.
5615 8. Added the rcpt_4xx retry error code.
5617 9. Added postmaster_mailfrom=xxx to callout verification option.
5619 10. Added mailfrom=xxxx to the callout verification option, for verify=
5622 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5623 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5625 12. Inserted some casts to stop certain compilers warning when using pointer
5626 differences as field lengths or precisions in printf-type calls (mostly
5627 affecting debugging statements).
5629 13. Added optional readline() support for -be (dynamically loaded).
5631 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5632 same clock tick as a message's arrival, so that its received time was the
5633 same as the "first fail" time on the retry record, and that message
5634 remained on the queue past the ultimate address timeout, every queue runner
5635 would try a delivery (because it was past the ultimate address timeout) but
5636 after another failure, the ultimate address timeout, which should have then
5637 bounced the address, did not kick in. This was a "< instead of <=" error;
5638 in most cases the first failure would have been in the next clock tick
5639 after the received time, and all would be well.
5641 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5642 being recognized when the domain list was tested by the match_domain
5643 condition in an expansion string.
5645 16. Added the ${str2b64: operator.
5647 17. Exim was always calling setrlimit() to set a large limit for the number of
5648 processes, without checking whether the existing limit was already
5649 adequate. (It did check for the limit on file descriptors.) Furthermore,
5650 errors from getrlimit() and setrlimit() were being ignored. Now they are
5651 logged to the main and panic logs, but Exim does carry on, to try to do its
5652 job under whatever limits there are.
5654 18. Imported PCRE 5.0.
5656 19. Trivial typo in log message " temporarily refused connection" (the leading
5659 20. If the log selector return_path_on_delivery was set and an address was
5660 redirected to /dev/null, the delivery process crashed because it assumed
5661 that a return path would always be set for a "successful" delivery. In this
5662 case, the whole delivery is bypassed as an optimization, and therefore no
5665 21. Internal re-arrangement: the function for sending a challenge and reading
5666 a response while authentication was assuming a zero-terminated challenge
5667 string. It's now changed to take a pointer and a length, to allow for
5668 binary data in such strings.
5670 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5672 23. Exim was not respecting finduser_retries when seeking the login of the
5673 uid under which it was called; it was always trying 10 times. (The default
5674 setting of finduser_retries is zero.) Also, it was sleeping after the final
5675 failure, which is pointless.
5677 24. Implemented tls_on_connect_ports.
5679 25. Implemented acl_smtp_predata.
5681 26. If the domain in control=submission is set empty, Exim assumes that the
5682 authenticated id is a complete email address when it generates From: or
5683 Sender: header lines.
5685 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5686 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5687 chown and chgrp in /bin and hostname in /usr/bin.
5689 28. Exim was keeping the "process log" file open after each use, just as it
5690 does for the main log. This opens the possibility of it remaining open for
5691 long periods when the USR1 signal hits a daemon. Occasional processlog
5692 errors were reported, that could have been caused by this. Anyway, it seems
5693 much more sensible not to leave this file open at all, so that is what now
5696 29. The long-running daemon process does not normally write to the log once it
5697 has entered its main loop, and it closes the log before doing so. This is
5698 so that log files can straightforwardly be renamed and moved. However,
5699 there are a couple of unusual error situations where the daemon does write
5700 log entries, and I had neglected to close the log afterwards.
5702 30. The text of an SMTP error response that was received during a remote
5703 delivery was being truncated at 512 bytes. This is too short for some of
5704 the long messages that one sometimes sees. I've increased the limit to
5707 31. It is now possible to make retry rules that apply only when a message has a
5708 specific sender, in particular, an empty sender.
5710 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5711 it possible to be selective about when SMTP synchronization is enforced.
5713 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5715 32. Implemented hosts_connection_nolog.
5717 33. Added an ACL for QUIT.
5719 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5722 35. Added mailbox_size and mailbox_filecount to appendfile.
5724 36. Added control = no_multiline_responses to ACLs.
5726 37. There was a bug in the logic of the code that waits for the clock to tick
5727 in the case where the clock went backwards by a substantial amount such
5728 that the microsecond fraction of "now" was more than the microsecond
5729 fraction of "then" (but the whole seconds number was less).
5731 38. Added support for the libradius Radius client library this is found on
5732 FreeBSD (previously only the radiusclient library was supported).
5738 1. When certain lookups returned multiple values in the form name=value, the
5739 quoting of the values was not always being done properly. Specifically:
5740 (a) If the value started with a double quote, but contained no whitespace,
5742 (b) If the value contained whitespace other than a space character (i.e.
5743 tabs or newlines or carriage returns) it was not quoted.
5744 This fix has been applied to the mysql and pgsql lookups by writing a
5745 separate quoting function and calling it from the lookup code. The fix
5746 should probably also be applied to nisplus, ibase and oracle lookups, but
5747 since I cannot test any of those, I have not disturbed their existing code.
5749 2. A hit in the callout cache for a specific address caused a log line with no
5750 reason for rejecting RCPT. Now it says "Previous (cached) callout
5751 verification failure".
5753 3. There was an off-by-one bug in the queryprogram router. An over-long
5754 return line was truncated at 256 instead of 255 characters, thereby
5755 overflowing its buffer with the terminating zero. As well as fixing this, I
5756 have increased the buffer size to 1024 (and made a note to document this).
5758 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5759 when Exim is waiting for an SMTP response from a remote server, Exim
5760 restarts its select() call on the socket, thereby resetting its timeout.
5761 This is not a problem when such interrupts are rare. Somebody set up a cron
5762 job to run exiwhat every 2 minutes, which is less than the normal select()
5763 timeout (5 or 10 minutes). This meant that the select() timeout never
5764 kicked in because it was always reset. I have fixed this by comparing the
5765 time when an interrupt arrives with the time at the start of the first call
5766 to select(). If more time than the timeout has elapsed, the interrupt is
5767 treated as a timeout.
5769 5. Some internal re-factoring in preparation for the addition of Sieve
5770 extensions (by MH). In particular, the "personal" test is moved to a
5771 separate function, and given an option for scanning Cc: and Bcc: (which is
5772 not set for Exim filters).
5774 6. When Exim created an email address using the login of the caller as the
5775 local part (e.g. when creating a From: or Sender: header line), it was not
5776 quoting the local part when it contained special characters such as @.
5778 7. Installed new OpenBSD configuration files.
5780 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5781 try to make them clearer.
5783 9. Callout options, other than the timeout value, were being ignored when
5784 verifying sender addresses in header lines. For example, when using
5786 verify = header_sender/callout=no_cache
5788 the cache was (incorrectly) being used.
5790 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5791 only the Cygwin environment.
5793 11. When return_path_on_delivery was set as a log selector, if different remote
5794 addresses in the same message used different return paths and parallel
5795 remote delivery occurred, the wrong values would sometimes be logged.
5796 (Whenever a remote delivery process finished, the return path value from
5797 the most recently started remote delivery process was logged.)
5799 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5800 lines when AUTH and/or TLS are in use. This is the "received protocol"
5801 field. Exim used to use "asmtp" for authenticated SMTP, without any
5802 indication (in the protocol name) for TLS use. Now it follows the RFC and
5803 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5804 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5805 names appear in log lines as well as in Received: header lines.
5807 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5808 extensions, and comparison tests, and to fix some bugs.
5810 14. Changes to the "personal" filter test:
5812 (1) The test was buggy in that it was just doing the equivalent of
5813 "contains" tests on header lines. For example, if a user's address was
5814 anne@some.where, the "personal" test would incorrectly be true for
5816 To: susanne@some.where
5818 This test is now done by extracting each address from the header in turn,
5819 and checking the entire address. Other tests that are part of "personal"
5820 are now done using regular expressions (for example, to check local parts
5821 of addresses in From: header lines).
5823 (2) The list of non-personal local parts in From: addresses has been
5824 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5825 taken from the Sieve specification recommendations.
5827 (3) If the message contains any header line starting with "List-" it is
5828 treated as non-personal.
5830 (4) The test for "circular" in the Subject: header line has been removed
5831 because it now seems ill-conceived.
5833 15. Minor typos in src/EDITME comments corrected.
5835 16. Installed latest exipick from John Jetmore.
5837 17. If headers_add on a router specified a text string that was too long for
5838 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5839 of string_sprintf() is now avoided.
5841 18. $message_body_size was not set (it was always zero) when running the DATA
5842 ACL and the local_scan() function.
5844 19. For the "mail" command in an Exim filter, no default was being set for
5845 the once_repeat time, causing a random time value to be used if "once" was
5846 specified. (If the value happened to be <= 0, no repeat happened.) The
5847 default is now 0s, meaning "never repeat". The "vacation" command was OK
5848 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5849 (I found it when inspecting the code).
5851 20. There is now an overall timeout for performing a callout verification. It
5852 defaults to 4 times the callout timeout, which applies to individual SMTP
5853 commands during the callout. The overall timeout applies when there is more
5854 than one host that can be tried. The timeout is checked before trying the
5855 next host. This prevents very long delays if there are a large number of
5856 hosts and all are timing out (e.g. when the network connections are timing
5857 out). The value of the overall timeout can be changed by specifying an
5858 additional sub-option for "callout", called "maxwait". For example:
5860 verify = sender/callout=5s,maxwait=20s
5862 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5863 to the end before writing, but this should make it even safer).
5865 22. Exim was forgetting that it had advertised PIPELINING for the second and
5866 subsequent messages on an SMTP connection. It was also not resetting its
5867 memory on STARTTLS and an internal HELO.
5869 23. When Exim logs an SMTP synchronization error within a session, it now
5870 records whether PIPELINING has been advertised or not.
5872 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5873 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5874 rather than long int.
5876 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5878 26. Added the never_mail option to autoreply.
5884 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5885 crash if the getsockname() call failed; this can happen if a connection is
5886 closed very soon after it is established. The problem was simply in the
5887 order in which certain operations were done, causing Exim to try to write
5888 to the SMTP stream before it had set up the file descriptor. The bug has
5889 been fixed by making things happen in the correct order.
5895 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5896 before the connection was closed, thus losing the rejection response.
5898 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5899 some early Solaris releases, but causes trouble in current releases where
5900 socklen_t is defined.
5902 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5905 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5908 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5909 that releasing the top of it at the end releases what was used for sub-
5910 expansions (unless the block got too big). However, discard this block if
5911 the first thing is a variable or header, so that we can use its block when
5912 it is dynamic (useful for very large $message_headers, for example).
5914 6. Lookups now cache *every* query, not just the most recent. A new, separate
5915 store pool is used for this. It can be recovered when all lookup caches are
5916 flushed. Lookups now release memory at the end of their result strings.
5917 This has involved some general refactoring of the lookup sources.
5919 7. Some code has been added to the store_xxx() functions to reduce the amount
5920 of flapping under certain conditions.
5922 8. log_incoming_interface used to affect only the <= reception log lines. Now
5923 it causes the local interface and port to be added to several more SMTP log
5924 lines, for example "SMTP connection from", and rejection lines.
5926 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5928 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5930 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5931 could be overwritten at the end of the current message (or the start of a
5932 new message if it was set in a HELO ACL). The value is now preserved for
5933 the duration of the SMTP connection.
5935 12. If a transport had a headers_rewrite setting, and a matching header line
5936 contained an unqualified address, that address was qualified, even if it
5937 did not match any rewriting rules. The underlying bug was that the values
5938 of the flags that permit the existence of unqualified sender and recipient
5939 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5940 non-local messages, and by -bnq for local messages) were not being
5941 preserved with the message after it was received.
5943 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5944 "next input=" as part of the text comprising the host identity instead of
5945 the correct text. The code was using the same buffer for two different
5946 strings. However, depending on which order the printing function evaluated
5947 its arguments, the bug did not always show up. Under Linux, for example, my
5948 test suite worked just fine.
5950 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5951 doesn't work with some older versions of Perl. It has been changed to "my",
5952 which in any case is probably the better facility to use.
5954 15. A really picky compiler found some instances of statements for creating
5955 error messages that either had too many or two few arguments for the format
5958 16. The size of the buffer for calls to the DNS resolver has been increased
5959 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5960 for addresses that have a lot of PTR records. This alleviates a problem; it
5961 does not fully solve it.
5963 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5964 buffer now truncates the list and logs the incident, which is the same
5965 action as happens when Exim is looking up a host name and its aliases.
5966 Previously in this situation something unpredictable would happen;
5967 sometimes it was "internal error: store_reset failed".
5969 18. If a server dropped the connection unexpectedly when an Exim client was
5970 using GnuTLS and trying to read a response, the client delivery process
5971 crashed while trying to generate an error log message.
5973 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5974 string, for example:
5976 warn message = H1: something\nH2: something
5978 the text was added as a single header line from Exim's point of view
5979 though it ended up OK in the delivered message. However, searching for the
5980 second and subsequent header lines using $h_h2: did not work. This has been
5981 fixed. Similarly, if a system filter added multiple headers in this way,
5982 the routers could not see them.
5984 20. Expanded the error message when iplsearch is called with an invalid key to
5985 suggest using net-iplsearch in a host list.
5987 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5988 ACLs are no longer actually imposed (and a message to that effect is
5991 22. If a "gecos" field in a passwd entry contained escaped characters, in
5992 particular, if it contained a \" sequence, Exim got it wrong when building
5993 a From: or a Sender: header from that name. A second bug also caused
5994 incorrect handling when an unquoted " was present following a character
5995 that needed quoting.
5997 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5998 was not being matched caselessly.
6000 24. Arranged for all hyphens in the exim.8 source to be escaped with
6003 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6004 properly. Recipient callout cache records were still being keyed to include
6005 the sender, even when use_sender was set false. This led to far more
6006 callouts that were necessary. The sender is no longer included in the key
6007 when use_sender is false.
6009 26. Added "control = submission" modifier to ACLs.
6011 27. Added the ${base62d: operator to decode base 62 numbers.
6013 28. dnsdb lookups can now access SRV records.
6015 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6016 the configuration file.
6018 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6019 (-v) mode. This makes the output for a verbose queue run more intelligible.
6021 31. Added a use_postmaster feature to recipient callouts.
6023 32. Added the $body_zerocount variable, containing the number of binary zero
6024 bytes in the message body.
6026 33. The time of last modification of the "new" subdirectory is now used as the
6027 "mailbox time last read" when there is a quota error for a maildir
6030 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6032 35. Added +ignore_unknown as a special item in host lists.
6034 36. Code for decoding IPv6 addresses in host lists is now included, even if
6035 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6036 address was recognized as an IP address, but was then not correctly decoded
6037 into binary, causing unexpected and incorrect effects when compared with
6044 1. Very minor rewording of debugging text in manualroute to say "list of
6045 hosts" instead of "hostlist".
6047 2. If verify=header_syntax was set, and a header line with an unqualified
6048 address (no domain) and a large number of spaces between the end of the
6049 name and the colon was received, the reception process suffered a buffer
6050 overflow, and (when I tested it) crashed. This was caused by some obsolete
6051 code that should have been removed. The fix is to remove it!
6053 3. When running in the test harness, delay a bit after writing a bounce
6054 message to get a bit more predictability in the log output.
6056 4. Added a call to search_tidyup() just before forking a reception process. In
6057 theory, someone could use a lookup in the expansion of smtp_accept_max_
6058 per_host which, without the tidyup, could leave open a database connection.
6060 5. Added the variables $recipient_data and $sender_data which get set from a
6061 lookup success in an ACL "recipients" or "senders" condition, or a router
6062 "senders" option, similar to $domain_data and $local_part_data.
6064 6. Moved the writing of debug_print from before to after the "senders" test
6067 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6068 problems for message scanning, either using a data ACL, or using
6069 local_scan() because the Received: header was not generated till after they
6070 were called (in order to set the time as the time of reception completion).
6071 I have revised the way this works. The header is now generated after the
6072 body is received, but before the ACL or local_scan() are called. After they
6073 are run, the timestamp in the header is updated.
6079 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6080 before starting a queue runner without re-exec. This happened only when
6081 deliver_drop_privilege was set or when the Exim user was set to root. The
6082 effect of the bug was that timeouts during subsequent deliveries caused
6083 crashes instead of being properly handled. The handler is now left at its
6084 default (and expected) setting.
6086 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6087 message, again when deliver_drop_privilege is set or Exim is run as root.
6088 The bug described in (1) was not present in this case, but the tidying up
6089 of the other signals was missing. I have made the two cases consistent.
6091 3. The ignore_target_hosts setting on a manualroute router was being ignored
6092 for hosts that were looked up using the /MX notation.
6094 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6097 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6098 operated on the sender address. After changing the $sender_address to <>
6099 for the sender address verify, Exim was re-instated it as the original
6100 (before rewriting) address, but remembering that it had rewritten it, so it
6101 wasn't rewriting it again. This bug also had the effect of breaking the
6102 sender address verification caching when the sender address was rewritten.
6104 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6105 This has been changed so that if the ip literal address matches
6106 ignore_target_hosts, the router declines.
6108 7. Added expansion conditions match_domain, match_address, and match_local_
6109 part (NOT match_host).
6111 8. The placeholder for the Received: header didn't have a length field set.
6113 9. Added code to Exim itself and to exim_lock to test for a specific race
6114 condition that could lead to file corruption when using MBX delivery. The
6115 issue is with the lockfile that is created in /tmp. If this file is removed
6116 after a process has opened it but before that process has acquired a lock,
6117 there is the potential for a second process to recreate the file and also
6118 acquire a lock. This could lead to two Exim processes writing to the file
6119 at the same time. The added code performs the same test as UW imapd; it
6120 checks after acquiring the lock that its file descriptor still refers to
6121 the same named file.
6123 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6124 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6127 11. Added the smtp_active_hostname option. If used, this will typically be made
6128 to depend on the incoming interface address. Because $interface_address is
6129 not set up until the daemon has forked a reception process, error responses
6130 that can happen earlier (such as "too many connections") no longer contain
6133 12. If an expansion in a condition on a "warn" statement fails because a lookup
6134 defers, the "warn" statement is abandoned, and the next ACL statement is
6135 processed. Previously this caused the whole ACL to be aborted.
6137 13. Added the iplsearch lookup type.
6139 14. Added ident_timeout as a log selector.
6141 15. Added tls_certificate_verified as a log selector.
6143 16. Added a global option tls_require_ciphers (compare the smtp transport
6144 option of the same name). This controls incoming TLS connections.
6146 17. I finally figured out how to make tls_require_ciphers do a similar thing
6147 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6148 before starting the TLS session.
6150 18. Tabs are now shown as \t in -bP output.
6152 19. If the log selector return_path_on_delivery was set, Exim crashed when
6153 bouncing a message because it had too many Received: header lines.
6155 20. If two routers both had headers_remove settings, and the first one included
6156 a superfluous trailing colon, the final name in the first list and the
6157 first name in the second list were incorrectly joined into one item (with a
6158 colon in the middle).
6164 1. Added -C and -D options to the exinext utility, mainly to make it easier
6165 to include in the automated testing, but these could be helpful when
6166 multiple configurations are in use.
6168 2. The exinext utility was not formatting the output nicely when there was
6169 an alternate port involved in the retry record key, nor when there was a
6170 message id as well (for retries that were specific to a specific message
6171 and a specific host). It was also confused by IPv6 addresses, because of
6172 the additional colons they contain. I have fixed the IPv4 problem, and
6173 patched it up to do a reasonable job for IPv6.
6175 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6176 delivery, the log line now contains "pipelined" if PIPELINING was used.
6178 4. An SMTP transport process used to panic and die if the bind() call to set
6179 an explicit outgoing interface failed. This has been changed; it is now
6180 treated in the same way as a connect() failure.
6182 5. A reference to $sender_host_name in the part of a conditional expansion
6183 that was being skipped was still causing a DNS lookup. This no longer
6186 6. The def: expansion condition was not recognizing references to header lines
6187 that used bh_ and bheader_.
6189 7. Added the _cache feature to named lists.
6191 8. The code for checking quota_filecount in the appendfile transport was
6192 allowing one more file than it should have been.
6194 9. For compatibility with Sendmail, the command line option
6202 and sets the incoming protocol and host name (for trusted callers). The
6203 host name and its colon can be omitted when only the protocol is to be set.
6204 Note the Exim already has two private options, -pd and -ps, that refer to
6205 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6206 "s", but I don't think that's a major issue.
6208 10. A number of refactoring changes to the code, none of which should affect
6211 (a) The number of logging options was getting close to filling up the
6212 32-bit word that was used as a bit map. I have split them into two classes:
6213 those that are passed in the argument to log_write(), and those that are
6214 only ever tested independently outside of that function. These are now in
6215 separate 32-bit words, so there is plenty of room for expansion again.
6216 There is no change in the user interface or the logging behaviour.
6218 (b) When building, for example, log lines, the code previously used a
6219 macro that called string_cat() twice, in order to add two strings. This is
6220 not really sufficiently general. Furthermore, there was one instance where
6221 it was actually wrong because one of the argument was used twice, and in
6222 one call a function was used. (As it happened, calling the function twice
6223 did not affect the overall behaviour.) The macro has been replaced by a
6224 function that can join an arbitrary number of extra strings onto a growing
6227 (c) The code for expansion conditions now uses a table and a binary chop
6228 instead of a serial search (which was left over from when there were very
6229 few conditions). Also, it now recognizes conditions like "pam" even when
6230 the relevant support is not compiled in: a suitably worded error message is
6231 given if an attempt is made to use such a condition.
6233 11. Added ${time_interval:xxxxx}.
6235 12. A bug was causing one of the ddress fields not to be passed back correctly
6236 from remote delivery subprocesses. The field in question was not being
6237 subsequently used, so this caused to problems in practice.
6239 13. Added new log selectors queue_time and deliver_time.
6241 14. Might have fixed a bug in maildirsizefile handling that threw up
6242 "unexpected character" debug warnings, and recalculated the data
6243 unnecessarily. In any case, I expanded the warning message to give more
6246 15. Added the message "Restricted characters in address" to the statements in
6247 the default ACL that block characters like @ and % in local parts.
6249 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6250 Three changes have been made:
6252 (a) There was a serious bug; a negative response to MAIL caused the whole
6253 recipient domain to be cached as invalid, thereby blocking all messages
6254 to all local parts at the same domain, from all senders. This bug has
6255 been fixed. The domain is no longer cached after a negative response to
6256 MAIL if the sender used is not empty.
6258 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6261 (c) A new callout option, "use_sender" has been added for people who want
6262 the modified behaviour.
6268 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6271 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6272 indeed breaks things for older releases.
6274 3. Added additional logging to the case where there is a problem reading data
6275 from a filter that is running in a subprocess using a pipe, in order to
6276 try to track down a specific problem.
6278 4. Testing facility fudge: when running in the test harness and attempting
6279 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6280 getting "No route to host". Convert this to a timeout.
6282 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6285 6. Some OS don't have socklen_t but use size_t instead. This affects the
6286 fifth argument of getsockopt() amongst other things. This is now
6287 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6288 can be set for individual OS. I have set it for SunOS5, OSF1, and
6289 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6290 some earlier ones do not.
6292 7. Change 4.30/15 was not doing the test caselessly.
6294 8. The standard form for an IPv6 address literal was being rejected by address
6295 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6296 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6297 this, as well as the form without the "IPv6" on the front (but only when
6298 address literals are enabled, of course).
6300 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6302 10. Exim crashed if a message with an empty sender address specified by -f
6303 encountered a router with an errors_to setting. This could be provoked only
6304 by a command such as
6308 where an empty string was supplied; "<>" did not hit this bug.
6310 11. Installed PCRE release 4.5.
6312 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6313 remained set. It is now erased.
6315 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6316 times from message ids (which are base 36 rather than the normal 62).
6318 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6319 were being counted as actual protocol errors, and logged if the log
6320 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6321 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6322 and DATA following a set of rejected RCPTs do not count as protocol errors.
6323 In other words, Exim assumes they were pipelined, though this may not
6324 actually be the case. Of course, in all cases the client gets an
6325 appropriate error code.
6327 15. If a lookup fails in an ACL condition, a message about the failure may
6328 be available; it is used if testing the ACL cannot continue, because most
6329 such messages specify what the cause of the deferral is. However, some
6330 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6331 that caused an old message to be retained and used if a later statement
6332 caused a defer, replacing the real cause of the deferral.
6334 16. If an IP address had so many PTR records that the DNS lookup buffer
6335 was not large enough to hold them, Exim could crash while trying to process
6336 the truncated data. It now detects and logs this case.
6338 17. Further to 4.21/58, another change has been made: if (and only if) the
6339 first line of a message (the first header line) ends with CRLF, a bare LF
6340 in a subsequent header line has a space inserted after it, so as not to
6341 terminate the header.
6343 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6344 unnecessarily, used atoi() instead of strtol(), and didn't check the
6345 termination when getting file sizes from file names by regex.
6347 19. Completely re-implemented the support for maildirsize files, in the light
6348 of a number of problems with the previous contributed implementation
6349 (4.30/29). In particular:
6351 . If the quota is zero, the maildirsize file is maintained, but no quota is
6354 . If the maildir directory does not exist, it is created before any attempt
6355 to write a maildirsize file.
6357 . The quota value in the file is just a cache; if the quota is changed in
6358 the transport, the new value overrides.
6360 . A regular expression is available for excluding directories from the
6363 20. The autoreply transport checks the characters in options that define the
6364 message's headers; it allows continued headers, but it was checking with
6365 isspace() after an embedded newline instead of explicitly looking for a
6368 21. If all the "regular" hosts to which an address was routed had passed their
6369 expiry times, and had not reached their retry times, the address was
6370 bounced, even if fallback hosts were defined. Now Exim should go on to try
6373 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6374 equivalent code in the SMTP transport. Some hosts send humungous responses
6375 to HELO/EHLO, more than 1024 it seems.
6377 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6378 gives compiler warnings in some environments. I've now done it "properly",
6381 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6382 (because of problems with the built-in one) was declared to return uschar *
6383 instead of char *, causing compiler failure.
6385 25. Fixed a file descriptor leak when processing alias/forward files.
6387 26. Fixed a minor format string issue in dbfn.c.
6389 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6391 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6392 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6395 29. When a sender address is verified, it is cached, to save repeating the test
6396 when there is more than one recipient in a message. However, when the
6397 verification involves a callout, it is possible for different callout
6398 options to be set for different recipients. It is too complicated to keep
6399 track of this in the cache, so now Exim always runs a verification when a
6400 callout is required, relying on the callout cache for the optimization.
6401 The overhead is duplication of the address routing, but this should not be
6404 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6405 to be verified with callout=postmaster, and the main callout worked but the
6406 postmaster check failed, the verification correctly failed. However, if a
6407 subsequent RCPT command asked for sender verification *without* the
6408 postmaster check, incorrect caching caused this verification also to fail,
6411 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6412 it was not caching the DNS options (qualify_single, search_parents) that
6413 were used when the lookup failed. A subsequent lookup with different
6414 options therefore always gave the same answer, though there were cases
6415 where it should not have. (Example: a "domains = !$mx_any" option on a
6416 dnslookup router: the "domains" option is always processed without any
6417 widening, but the router might have qualify_single set.) Now Exim uses the
6418 cached value only when the same options are set.
6420 32. Added John Jetmore's "exipick" utility to the distribution.
6422 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6423 than a timeout (e.g. a certificate is required, and is not provided), an
6424 Exim server now closes the connection immediately. Previously it waited for
6425 the client to close - but if the client is SSL, it seems that they each
6426 wait for each other, leading to a delay before one of them times out.
6428 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6429 maintained 0.8.x compatibility because I don't think many are using it, and
6430 it is clearly obsolete.
6432 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6435 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6436 client certificate was expired. A simple patch fixes this, though I don't
6437 understand the full logic of why the verify callback is called multiple
6440 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6441 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6442 which causes problems with some clients (such as the Certicom SSL Plus
6443 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6444 disables the coutermeasure allowing Eudora to connect."
6446 38. Exim was not checking that a write() to a log file succeeded. This could
6447 lead to Bad Things if a log got too big, in particular if it hit a file
6448 size limit. Exim now panics and dies if it cannot write to a log file, just
6449 as it does if it cannot open a log file.
6451 39. Modified OS/Makefile-Linux so that it now contains
6453 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6455 The two -D definitions ensure that Exim is compiled with large file
6456 support, which makes it possible to handle log files that are bigger than
6459 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6460 instance) a domain was checked against a named list that involved a lookup,
6461 causing $domain_data to be set, then another domain was checked against the
6462 same list, then the first domain was re-checked, the value of $domain_data
6463 after the final check could be wrong. In particular, if the second check
6464 failed, it could be set empty. This bug probably also applied to
6467 41. The strip_trailing_dot option was not being applied to the address given
6468 with the -f command-line option.
6470 42. The code for reading a message's header from the spool was incrementing
6471 $received_count, but never initializing it. This meant that the value was
6472 incorrect (doubled) while delivering a message in the same process in which
6473 it was received. In the most common configuration of Exim, this never
6474 happens - a fresh exec is done - but it can happen when
6475 deliver_drop_privilege is set.
6477 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6478 it now includes up to 150 characters of the unexpected data in the log
6481 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6482 and building data strings. The size of both of these buffers was 10 000
6483 bytes - far larger than anybody would *ever* want, thought I. Needless to
6484 say, somebody hit the limit. I have increased the maximum line length to
6485 20 000 and the maximum data length of concatenated lines to 100 000. I have
6486 also fixed two bugs, because there was no checking on these buffers. Tsk,
6487 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6488 buffer is too small.
6490 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6491 lsearch lookups. Now it does.
6493 46. When parsing a route_list item in a manualroute router, a fixed-length
6494 buffer was used for the list of hosts. I made this 1024 bytes long,
6495 thinking that nobody would ever have a list of hosts that long. Wrong.
6496 Somebody had a whole pile of complicated expansion conditions, and the
6497 string was silently truncated, leading to an expansion error. It turns out
6498 that it is easier to change to an unlimited length (owing to other changes
6499 that have happened since this code was originally written) than to build
6500 structure for giving a limitation error. The length of the item that
6501 expands into the list of hosts is now unlimited.
6503 47. The lsearch lookup could not handle data where the length of text line was
6504 more than 4095 characters. Such lines were truncated, leading to shortened
6505 data being returned. It should now handle lines of any length.
6507 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6508 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6511 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6512 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6513 no longer happens in comment lines. A list of replacements is now placed
6514 at the head of all of the source files, except those whose only change is
6515 to replace PERL_COMMAND in the very first #! line.
6517 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6518 messages on the queue, with a bottom-up merge sort, using code contributed
6519 by Michael Haardt. This should make operations like -bp somewhat faster on
6520 large queues. It won't affect queue runners, except when queue_run_in_order
6523 51. Installed eximstats 1.31 in the distribution.
6525 52. Added support for SRV lookups to the dnslookup router.
6527 53. If an ACL referred to $message_body or $message_body_end, the value was not
6528 reset for any messages that followed in the same SMTP session.
6530 54. The store-handling optimization for building very long strings was not
6531 differentiating between the different store pools. I don't think this
6532 actually made any difference in practice, but I've tidied it.
6534 55. While running the routers to verify a sender address, $sender_address
6535 was still set to the sender address. This is wrong, because when routing to
6536 send a bounce to the sender, it would be empty. Therefore, I have changed
6537 it so that, while verifying a sender address, $sender_address is set to <>.
6538 (There is no change to what happens when verifying a recipient address.)
6540 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6541 target A or AAAA records (if not already returned) without resetting the
6542 qualify_single or search_parents options of the DNS resolver. These are
6543 inappropriate in this case because the targets of MX and SRV records must
6544 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6545 target that, when qualified, matched something in the local domain. These
6546 two options are now turned off when doing these lookups.
6548 57. It seems that at least some releases of Reiserfs (which does not have the
6549 concept of a fixed number of inodes) returns zero and not -1 for the
6550 number of available inodes. This interacted badly with check_spool_inodes,
6551 which assumed that -1 was the "no such thing" setting. What I have done is
6552 to check that the total number of inodes is greater than zero before doing
6553 the test of how many are available.
6555 58. When a "warn" ACL statement has a log_message modifier, the message is
6556 remembered, and not repeated. This is to avoid a lot of repetition when a
6557 message has many recipients that cause the same warning to be written.
6558 However, Exim was preserving the list of already written lines for an
6559 entire SMTP session, which doesn't seem right. The memory is now reset if a
6560 new message is started.
6562 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6563 debugging output unless log_rewrite was also set.
6565 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6566 of (int)(handle) when we know that handle contains (void *)(-1).
6568 61. The Exim daemon panic-logs an error return when it closes the incoming
6569 connection. However "connection reset by peer" seems to be common, and
6570 isn't really an error worthy of noting specially, so that particular error
6573 62. When Exim is trying to find all the local interfaces, it used to panic and
6574 die if the ioctl to get the interface flags failed. However, it seems that
6575 on at least one OS (Solaris 9) it is possible to have an interface that is
6576 included in the list of interfaces, but for which you get a failure error
6577 for this call. This happens when the interface is not "plumbed" into a
6578 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6579 failure of the "get flags" call assumes that the interface is down.
6581 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6582 makes life easier for people who are doing arithmetic on fields extracted
6583 from dates, where you often get leading zeros that should not be
6584 interpreted as octal.
6586 64. Added qualify_domain to the redirect router, to override the global
6589 65. If a pathologically long header line contained very many addresses (the
6590 report of this problem mentioned 10 000) and each of them was rewritten,
6591 Exim could use up a very large amount of memory. (It kept on making new
6592 copies of the header line as it rewrote, and never released the old ones.)
6593 At the expense of a bit more processing, the header rewriting function has
6594 been changed so that it no longer eats memory in this way.
6596 66. The generation of the Received: header has been moved from the time that a
6597 message starts to be received, to the time that it finishes. The timestamp
6598 in the Received: header should now be very close to that of the <= log
6599 line. There are two side-effects of this change:
6601 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6602 logged header lines no longer include the local Received: line, because
6603 it has not yet been created. The same applies to a copy of the message
6604 that is returned to a non-SMTP sender when a message is rejected.
6606 (b) When a filter file is tested using -bf, no additional Received: header
6607 is added to the test message. After some thought, I decided that this
6610 This change does not affect the value of $received_for. It is still set
6611 after address rewriting, but before local_scan() is called.
6613 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6615 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6616 gave an unhelpful panic error message, and a defer error. I have managed to
6617 change this behaviour so that it now rejects any supplied certificate,
6618 which seems right, as the list of acceptable certificates is empty.
6620 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6621 gave an unhelpful defer error. I have not managed to make this reject any
6622 supplied certificates, but the error message it gives is "no certificate
6623 supplied", which is not helpful.
6625 70. exigrep's output now also includes lines that are not associated with any
6626 message, but which match the given pattern. Implemented by a patch from
6627 Martin Sluka, which also tidied up the Perl a bit.
6629 71. Recipient callout verification, like sender verification, was using <> in
6630 the MAIL FROM command. This isn't really the right thing, since the actual
6631 sender may affect whether the remote host accepts the recipient or not. I
6632 have changed it to use the actual sender in the callout; this means that
6633 the cache record is now keyed on a recipient/sender pair, not just the
6634 recipient address. There doesn't seem to be a real danger of callout loops,
6635 since a callout by the remote host to check the sender would use <>.
6636 [SEE ABOVE: changed after hitting problems.]
6638 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6639 temporary errors. However, in the case of such a code being given after
6640 the end of a data transmission (i.e. after ".") Exim was failing to write
6641 a retry record for the message. (Yes, there was some broken host that was
6642 actually sending 8xx at this point.)
6644 73. An unknown lookup type in a host list could cause Exim to panic-die when
6645 the list was checked. (An example that provoked this was putting <; in the
6646 middle of a list instead of at the start.) If this happened during a DATA
6647 ACL check, a -D file could be left lying around. This kind of configuration
6648 error no longer causes Exim to die; instead it causes a defer error. The
6649 incident is still logged to the main and panic logs.
6651 74. Buglet left over from Exim 3 conversion. The message "too many messages
6652 in one connection" was written to the rejectlog but not the mainlog, except
6653 when address rewriting (yes!) was being logged.
6655 75. Added write_rejectlog option.
6657 76. When a system filter was run not as root (that is, when system_filter_user
6658 was set), the values of the $n variables were not being returned to the
6659 main process; thus, they were not subsequently available in the $sn
6662 77. Added +return_path_on_delivery log selector.
6664 78. A connection timeout was being treated differently from recipients deferred
6665 when testing hosts_max_try with a message that was older than the host's
6666 retry timeout. (The host should not be counted, thus allowing all hosts to
6667 be tried at least once before bouncing.) This may have been the cause of an
6668 occasionally reported bug whereby a message would remain on the queue
6669 longer than the retry timeout, but would be bounced if a delivery was
6670 forced. I say "may" because I never totally pinned down the problem;
6671 setting up timeout/retry tests is difficult. See also the next item.
6673 79. The ultimate address timeout was not being applied to errors that involved
6674 a combination of host plus message (for example, a timeout on a MAIL
6675 command). When an address resolved to a number of possible hosts, and they
6676 were not all tried for each delivery (e.g. because of hosts_max_try), a
6677 message could remain on the queue longer than the retry timeout.
6679 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6682 81. Fixed an obscure SMTP outgoing bug which required at least the following
6683 conditions: (a) there was another message waiting for the same server;
6684 (b) the server returned 5xx to all RCPT commands in the first message so
6685 that the message was not completed; (c) the server dropped the connection
6686 or gave a negative response to the RSET that Exim sends to abort the
6687 transaction. The observed case was a dropped connection after DATA that had
6688 been sent in pipelining mode. That is, the server had advertised PIPELINING
6689 but was not implementing it correctly. The effect of the bug was incorrect
6690 behaviour, such as trying another host, and this could lead to a crash.
6696 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6697 and daemon.c were passed as pointers to ints; they should have been
6698 pointers to socklen_t variables (which are typically unsigned ints).
6700 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6703 3. Fixed a really odd bug that affected only the testing scheme; patching a
6704 certain fixed string in the binary changed the value of another string that
6705 happened to be identical to the end of the original first string.
6707 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6708 name", it returns that address as the IP address. On some operating
6709 systems (e.g. Solaris), it also passes back the IP address string as the
6710 "host name". However, on others (e.g. Linux), it passes back an empty
6711 string. Exim wasn't checking for this, and was changing the host name to an
6712 empty string, assuming it had been canonicalized.
6714 5. Although rare, it is permitted to have more than one PTR record for a given
6715 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6716 all the names associated with an address, because they do in Solaris.
6717 However, it seems that they do not in Linux for data that comes from the
6718 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6719 I found this out when I moved to a new Linux workstation and tried to run
6720 the Exim test suite.
6722 To get round this problem I have changed the code so that it now does its
6723 own call to the DNS to look up PTR records when searching for a host name.
6724 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6725 addresses that are only in /etc/hosts are still found.
6727 This behaviour is, however, controlled by an option called host_lookup_
6728 order, which defaults to "bydns:byaddr". If people want to use the other
6729 order, or indeed, just use one or the other means of lookup, they can
6730 specify it in this variable.
6732 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6733 some operating systems, this comes back from gethostbyaddr() as an empty
6734 string, and this is what Exim used to test for. However, it seems that in
6735 other systems, "." is yielded. Exim now tests for this case too.
6737 7. The values of check_spool_space and check_log_space are now held internally
6738 as a number of kilobytes instead of an absolute number of bytes. If a
6739 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6740 kilobyte. This means that much larger values can be stored.
6742 8. Exim monitor: an attempt to get the action menu when not actually pointing
6743 at a message produces an empty menu entitled "No message selected". This
6744 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6745 no entries in it ("Shell widget menu has zero width and/or height"). So I
6746 have added a single, blank menu entry in this case.
6748 9. Added ${quote_local_part.
6750 10. MIME decoding is now applied to the contents of Subject: header lines when
6753 11. Now that a reference to $sender_host_address automatically causes a reverse
6754 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6755 host lookup before query-style lookups in lists that might use this
6756 variable. This has therefore been abolished, and the "net-" prefix is no
6757 longer necessary for query-style lookups.
6759 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6760 have been a typo for LFLAGS, so it has been changed.
6762 13. The install script calls Exim with "-C /dev/null" in order to find the
6763 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6764 to be output. However, since Exim outputs its version number before the
6765 error, it didn't break the script. It just looked ugly. I fixed this by
6766 always allowing "-C /dev/null" if the caller is root.
6768 14. Ignore overlarge ACL variable number when reading spool file - insurance
6769 against a later release with more variables having written the file.
6771 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6772 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6773 this, as well as the form without the "IPv6" on the front.
6775 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6776 OS/Makefile-Darwin file.
6778 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6779 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6781 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6782 variable. However, some OS use ints for inodes. I've added cast to long int
6783 to get rid of the compiler warning.
6785 19. I had forgotten to lock out "/../" in configuration file names when
6786 ALT_CONFIG_PREFIX was set.
6788 20. Routers used for verification do not need to specify transports. However,
6789 if such a router generated a host list, and callout was configured, Exim
6790 crashed, because it could not find a port number from the (non-existent)
6791 transport. It now assumes port 25 in this circumstance.
6793 21. Added the -t option to exigrep.
6795 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6796 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6797 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6798 them is set, LOOKUP_LSEARCH is forced.
6800 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6802 24. Added sender and host information to the "rejected by local_scan()" log
6803 line; previously there was no indication of these.
6805 25. Added .include_if_exists.
6807 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6808 Linux. It turns out that not all file systems support this. Apparently some
6809 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6810 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6811 not supported on the file descriptor, is now ignored when Exim is trying to
6812 sync a directory. This applies only to Linux.
6814 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6816 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6817 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6820 29. Added support for maildirsize files from supplied patch (modified a bit).
6822 30. The use of :fail: followed by an empty string could lead Exim to respond to
6823 sender verification failures with (e.g.):
6825 550 Verification failed for <xxx>
6826 550 Sender verify failed
6828 where the first response line was missing the '-' that indicates it is not
6829 the final line of the response.
6831 31. The loop for finding the name of the user that called Exim had a hardwired
6832 limit of 10; it now uses the value of finduser_retries, which is used for
6833 all other user lookups.
6835 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6838 33. Exim was neglecting to zero errno before one call of strtol() when
6839 expanding a string and expecting an integer value. On some systems this
6840 resulted in spurious "integer overflow" errors. Also, it was casting the
6841 result into an int without checking.
6843 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6844 did not work. The code looks as if it has *never* worked, though it appears
6845 to have been documented since at least release 1.62. I have made it work.
6847 35. The "timeout_DNS" error in retry rules, also documented since at least
6848 1.62, also never worked. As it isn't clear exactly what this means, and
6849 clearly it isn't a major issue, I have abolished the feature by treating it
6850 as "timeout", and writing a warning to the main and panic logs.
6852 36. The display of retry rules for -brt wasn't always showing the error code
6855 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6856 timeout_connect_A, timeout_connect_MX.
6858 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6859 to the empty sender.
6861 39. The daemon was not analysing the content of -oX till after it had closed
6862 stderr and disconnected from the controlling terminal. This meant that any
6863 syntax errors were only noted on the panic log, and the return code from
6864 the command was 0. By re-arranging the code a little, I've made the
6865 decoding happen first, so such errors now appear on stderr, and the return
6866 code is 1. However, the actual setting up of the sockets still happens in
6867 the disconnected process, so errors there are still only recorded on the
6870 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6871 connections (as happens on some IP stacks) was logged at start up time as
6872 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6873 it from "IPv6 and IPv4", which means that two separate sockets are being
6876 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6877 says whether AF_INET or AF_INET6 was passed as an argument.
6879 42. Exiwhat output was messed up when time zones were included in log
6882 43. Exiwhat now gives more information about the daemon's listening ports,
6883 and whether -tls-on-connect was used.
6885 44. The "port" option of the smtp transport is now expanded.
6887 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6888 silently ignored. Now an error message is written to the main and panic
6891 46. There's a new ACL modifier called "logwrite" which writes to a log file
6892 as soon as it is encountered.
6894 47. Added $local_user_uid and $local_user_gid at routing time.
6896 48. Exim crashed when trying to verify a sender address that was being
6899 49. Exim was recognizing only a space character after ".include". It now also
6900 recognizes a tab character.
6902 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6903 extracting the relevant information from the specification. The man page no
6904 longer contains scrambled data for the -d option, and I've added a section
6905 at the front about calling Exim under different names.
6907 51. Added "extra_headers" argument to the "mail" command in filter files.
6909 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6912 53. Installed eximstats 1.29.
6914 54. Added transport_filter_timeout as a generic transport option.
6916 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6917 Cc: header lines. This was required by RFC 822, but it not required by RFC
6920 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6921 incoming messages that did not have them. Now it does so only if the
6922 message originates locally, that is, if there is no associated remote host
6923 address. When Resent- header lines are present, this applies to the Resent-
6924 lines rather than the non-Resent- lines.
6926 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6927 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6929 58. Messages for configuration errors now include the name of the main
6930 configuration file - useful now that there may be more than one file in a
6931 list (.included file names were always shown).
6933 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6934 for those rare installations that do not start the daemon as root or run it
6935 setuid root. I've cut out the call to initgroups() if the daemon is not
6938 60. The Exim user and group can now be bound into the binary as text strings
6939 that are looked up at the start of Exim's processing.
6941 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6943 62. Added $mailstore_basename variable.
6945 63. Installed patch to sieve.c from Michael Haardt.
6947 64. When Exim failed to open the panic log after failing to open the main log,
6948 the original message it was trying to log was written to stderr and debug
6949 output, but if they were not available (the usual case in production), it
6950 was lost. Now it is written to syslog before the two lines that record the
6951 failures to open the logs.
6953 65. Users' Exim filters run in subprocesses under the user's uid. It is
6954 possible for a "deliver" command or an alias in a "personal" command to
6955 provoke an address rewrite. If logging of address rewriting is configured,
6956 this fails because the process is not running as root or exim. There may be
6957 a better way of dealing with this, but for the moment (because 4.30 needs
6958 to be released), I have disabled address rewrite logging when running a
6959 filter in a non-root, non-exim process.
6965 1. The buildconfig auxiliary program wasn't quoting the value set for
6966 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6967 not defined. This bug was masked in 4.22 by the effect that was fixed in
6970 2. Some messages that were rejected after a message id was allocated were
6971 shown as "incomplete" by exigrep. It no longer does this for messages that
6972 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6974 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6975 have allow_domain_literals set, the ID did not get logged in the <= line.
6976 Domain literals are now always recognized in Message-ID: header lines.
6978 4. The first argument for a ${extract expansion item is the key name or field
6979 number. Leading and trailing spaces in this item were not being ignored,
6980 causing some misleading effects.
6982 5. When deliver_drop_privilege was set, single queue runner processes started
6983 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6984 same command in the process it spins off) were not dropping privilege.
6986 6. When the daemon running as "exim" started a queue runner, it always
6987 re-executed Exim in the spun-off process. This is a waste of effort when
6988 deliver_drop_privilege is set. The new process now just calls the
6989 queue-runner function directly.
6995 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6998 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6999 ignored. Though the use of -oP was forcing the writing of a pid file, it
7000 was always written to the default place.
7002 3. If the message "no IP address found for host xxxx" is generated during
7003 incoming verification, it is now followed by identification of the incoming
7004 connection (so you can more easily find what provoked it).
7006 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7008 5. Added some features to "harden" Exim a bit more against certain attacks:
7010 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7011 be put in Local/Makefile. This is like the never_users runtime option,
7012 but it cannot be overridden. The default setting is "root".
7014 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7015 prefix string with which any file named in a -C command line option
7018 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7019 is retained for -C and -D only if the caller of Exim is root. Without
7020 it, the exim user may also use -C and -D and retain privilege.
7022 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7023 command line option is disabled.
7025 6. Macro names set by the -D option must start with an upper case letter, just
7026 like macro names defined in the configuration file.
7028 7. Added "dereference=" facility to LDAP.
7030 8. Two instances of the typo "uknown" in the source files are fixed.
7032 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7033 the Configure-Makefile script screwed up while processing it.
7035 10. Incorporated PCRE 4.4.
7037 11. The SMTP synchronization check was not operating right at the start of an
7038 SMTP session. For example, it could not catch a HELO sent before the client
7039 waited for the greeting. There is now a check for outstanding input at the
7040 point when the greeting is written. Because of the duplex, asynchronous
7041 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7042 way, but not yet received, when the check is performed.
7044 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7045 on TCP/IP sockets, because this apparently causes some broken clients to
7048 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7049 unchanged) from the Cygwin maintainer.
7051 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7052 was supported instead of testing for mbx. Effectively a typo.
7054 15. The spa authenticator server code was not checking that the input it
7055 received was valid base64.
7057 16. The debug output line for the "set" modifier in ACLs was not showing the
7058 name of the variable that was being set.
7060 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7062 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7063 lookup on its own. Something else was needed to trigger the lookup. For
7064 example, a match in host_lookup or the need for a host name in a host list.
7065 Now, if $sender_host_name is referenced and the host name has not yet been
7066 looked up, a lookup is performed. If the lookup fails, the variable remains
7067 empty, and $host_lookup_failed is set to "1".
7069 19. Added "eqi" as a case-independent comparison operator.
7071 20. The saslauthd authentication condition could segfault if neither service
7072 nor realm was specified.
7074 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7075 error message that was logged was misleading, and incoming SMTP
7076 connections were dropped. The message is now more accurate, and temporary
7077 errors are given to SMTP connections.
7079 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7080 (except RSET) with a 421 error, until QUIT is received. However, it was
7081 failing to send a response to QUIT.
7083 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7084 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7085 ACL is now run after the helo_try_verify_hosts code.
7087 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7088 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7089 case-independent, so other case variants are also recognized). Apparently
7090 some systems use these upper case variants.
7092 25. If more than two messages were waiting for the same host, and a transport
7093 filter was specified for the transport, Exim sent two messages over the
7094 same TCP/IP connection, and then failed with "socket operation on non-
7095 socket" when it tried to send the third.
7097 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7099 27. The extern definition of crypt16() in expand.c was not being excluded when
7100 the OS had its own crypt16() function.
7102 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7103 as a preferred synonym for return_size_limit, both as an option and as an
7106 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7108 30. Changed the default configuration ACL to relax the local part checking rule
7109 for addresses that are not in any local domains. For these addresses,
7110 slashes and pipe symbols are allowed within local parts, but the sequence
7111 /../ is explicitly forbidden.
7113 31. SPA server authentication was not clearing the challenge buffer before
7116 32. log_message in a "warn" ACL statement was writing to the reject log as
7117 well as to the main log, which contradicts the documentation and doesn't
7118 seem right (because no rejection is happening). So I have stopped it.
7120 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7121 However, I am unable to do any testing of this.
7123 34. Fixed an infelicity in the appendfile transport. When checking directories
7124 for a mailbox, to see if any needed to be created, it was accidentally
7125 using path names with one or more superfluous leading slashes; tracing
7126 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7128 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7129 contents are added to the log line that is written for every discarded
7130 recipient. (Previously a log_message setting was ignored.)
7132 36. The ${quote: operator now quotes the string if it is empty.
7134 37. The install script runs exim in order to find its version number. If for
7135 some reason other than non-existence or emptiness, which it checks, it
7136 could not run './exim', it was installing it with an empty version number,
7137 i.e. as "exim-". This error state is now caught, and the installation is
7140 38. An argument was missing from the function that creates an error message
7141 when Exim fails to connect to the socket for saslauthd authentication.
7142 This could cause Exim to crash, or give a corrupted message.
7144 39. Added isip, isip4, and isip6 to ${if conditions.
7146 40. The ACL variables $acl_xx are now saved with the message, and can be
7147 accessed later in routers, transports, and filters.
7149 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7150 strings in the file are not string-expanded.
7152 42. If a MAIL command specified a SIZE value that was too large to fit into an
7153 int variable, the check against message_size_limit failed. Such values are
7154 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7155 day this will have to be increased, but I don't think I want to be around
7156 when emails are that large.
7163 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7164 iconv() is not standard in FreeBSD.
7166 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7167 IPv6 enabled. The observed symptom was a segmentation fault on return from
7168 the function os_common_find_running_interfaces() in src/os.c.
7170 3. In the check_special_case() function in daemon.c I had used "errno" as an
7171 argument name, which causes warnings on some systems. This was basically a
7172 typo, since it was named "eno" in the comments!
7174 4. The code that waits for the clock to tick (at a resolution of some fraction
7175 of a second) so as to ensure message-id uniqueness was always waiting for
7176 at least one whole tick, when it could have waited for less. [This is
7177 almost certainly not relevant at current processor speeds, where it is
7178 unlikely to ever wait at all. But we try to future-proof.]
7180 5. The function that sleeps for a time interval that includes fractions of a
7181 second contained a race. It did not block SIGALRM between setting the
7182 timer, and suspending (a couple of lines later). If the interval was short
7183 and the sigsuspend() was delayed until after it had expired, the suspension
7184 never ended. On busy systems this could lead to processes getting stuck for
7187 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7188 process, before it forks any delivery processes. The open lookup caching
7189 mechanism meant that the open file or database connection was passed into
7190 the delivery process. The problem was that delivery processes always tidy
7191 up cached lookup data. This could cause a problem for the next delivery
7192 process started by the queue runner, because the external queue runner
7193 process does not know about the closure. So the next delivery process
7194 still has data in the lookup cache. In the case of a file lookup, there was
7195 no problem because closing a file descriptor in a subprocess doesn't affect
7196 the parent. However, if the lookup was caching a connection to a database,
7197 the connection was closed, and the second delivery process was likely to
7198 see errors such as "PGSQL: query failed: server closed the connection
7199 unexpectedly". The problem has been fixed by closing all cached lookups
7200 in a queue runner before running a delivery process.
7202 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7203 seem to have the "const" qualifier which it has on other OS. I've
7206 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7207 *of the same type* (client or server) with the same public name that an
7208 error should be diagnosed.
7210 9. When Exim looked up a host name for an IP address, but failed to find the
7211 original IP address when looking up the host name (a safety check), it
7212 output the message "<ip address> does not match any IP for NULL", which was
7213 confusing, to say the least. The bug was that the host name should have
7214 appeared instead of "NULL".
7216 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7217 user that is built into the binary, and the -C or -D options is used, root
7218 privilege is dropped before the configuration file is read. In addition,
7219 logging is switched to stderr instead of the normal log files. If the
7220 configuration then re-defines the Exim user, the unprivileged environment
7221 is probably not what is expected, so Exim logs a panic warning message (but
7224 However, if deliver_drop_privilege is set, the unprivileged state may well
7225 be exactly what is intended, so the warning has been cut out in that case,
7226 and Exim is allowed to try to write to its normal log files.
7232 1. smtp_return_error_details was not giving details for temporary sender
7233 or receiver verification errors.
7235 2. Diagnose a configuration error if two authenticators have the same public
7238 3. Exim used not to create the message log file for a message until the first
7239 delivery attempt. This could be confusing when incoming messages were held
7240 for policy or load reasons. The message log file is now created at the time
7241 the message is received, and an initial "Received" line is written to it.
7243 4. The automatically generated man page for command line options had a minor
7244 bug that caused no ill effects; however, a more serious problem was that
7245 the procedure for building the man page automatically didn't always
7246 operate. Consequently, release 4.20 contains an out-of-date version. This
7247 shouldn't happen again.
7249 5. When building Exim with embedded Perl support, the script that builds the
7250 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7251 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7253 6. The freeze_tell option was not being used for messages that were frozen on
7254 arrival, either by an ACL or by local_scan().
7256 7. Added the smtp_incomplete_transaction log selector.
7258 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7259 was accepting AUTH without a new EHLO.
7261 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7262 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7265 10. Logging of TCP/IP connections (when configured) now happens in the main
7266 daemon process instead of the child process, so that the TCP/IP connection
7267 count is more accurate (but it can never be perfect).
7269 11. The use of "drop" in a nested ACL was not being handled correctly in the
7270 outer ACL. Now, if condition failure induced by the nested "drop" causes
7271 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7272 or "require"), the connection is dropped.
7274 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7275 that yield "discard" can now be used with an "accept" or a "discard" verb,
7276 but an error is generated for any others (because I can't see a useful way
7277 to define what should happen).
7279 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7280 are now processed in the same way as lines in the Exim configuration file.
7281 In particular, continuation lines are supported.
7283 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7285 15. Added -ti meaning -t -i.
7287 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7288 domains, and warn by logging if others are found.
7290 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7291 returned from GIFCONF when Exim is trying to find the list of interfaces on
7292 a host. The code in os.c has been modified to copy each ifreq to an aligned
7293 structure in all cases.
7295 Also, in some cases, the returned ifreq's were being copied to a 'struct
7296 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7297 means the last couple of bytes of an IPv6 address could be chopped if the
7298 ifreq contained only a normal sockaddr (14 bytes storage).
7300 18. Named domain lists were not supported in the hosts_treat_as_local option.
7301 An entry such as +xxxx was not recognized, and was treated as a literal
7304 19. Ensure that header lines added by a DATA ACL are included in the reject log
7305 if the ACL subsequently rejects the message.
7307 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7308 MD5 (which is deprecated).
7310 21. When testing a filter file using -bf, Exim was writing a message when it
7311 took the sender from a "From " line in the message, but it was not doing so
7312 when it took $return_path from a Return-Path: header line. It now does.
7314 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7315 with a valid header line field name (a series of printing characters
7316 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7318 23. Changed "disc" in the source to "disk" to conform to the documentation and
7319 the book and for uniformity.
7321 24. Ignore Sendmail's -Ooption=value command line item.
7323 25. When execve() failed while trying to run a command in a pipe transport,
7324 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7325 could be confused with a return value of 69 from the command itself. This
7326 has been changed to 127, the value the shell returns if it is asked to run
7327 a non-existent command. The wording for the related log line suggests a
7328 non-existent command as the problem.
7330 26. If received_header_text expands to an empty string, do not add a Received:
7331 header line to the message. (Well, it adds a token one on the spool, but
7332 marks it "old" so that it doesn't get used or transmitted.)
7334 27. Installed eximstats 1.28 (addition of -nt option).
7336 28. There was no check for failure on the call to getsockname() in the daemon
7337 code. This can fail if there is a shortage of resources on the system, with
7338 ENOMEM, for example. A temporary error is now given on failure.
7340 29. Contrary to the C standard, it seems that in some environments, the
7341 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7342 program. Exim now does this explicitly; it affects the formatting of
7343 timestamps using strftime().
7345 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7346 complaints. I've now initialized all the variables, to avoid this.
7348 32. Header lines added by a system filter were not being "seen" during
7349 transport-time rewrites.
7351 33. The info_callback() function passed to OpenSSL is set up with type void
7352 (*)(SSL *, int, int), as described somewhere. However, when calling the
7353 function (actually a macro) that sets it up, the type void(*)() is
7354 expected. I've put in a cast to prevent warnings from picky compilers.
7356 34. If a DNS black list lookup found a CNAME record, but there were no A
7357 records associated with the domain it pointed at, Exim crashed.
7359 35. If a DNS black list lookup returned more than one A record, Exim ignored
7360 all but the first. It now scans all returned addresses if a particular IP
7361 value is being sought. In this situation, the contents of the
7362 $dnslist_value variable are a list of all the addresses, separated by a
7365 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7366 to accept a host name and all its aliases if the forward lookup for any of
7367 them yielded the IP address of the incoming connection. Now it accepts only
7368 those names whose forward lookup yields the correct IP address. Any other
7369 names are discarded. This closes a loophole whereby a rogue DNS
7370 administrator could create reverse DNS records to break through a
7371 wildcarded host restriction in an ACL.
7373 37. If a user filter or a system filter that ran in a subprocess used any of
7374 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7375 the wrong values were passed to the pipe command ($thisaddress had the
7376 value of $0, $0 had the value of $1, etc). This bug was introduced by
7377 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7379 38. Improved the line breaking for long SMTP error messages from ACLs.
7380 Previously, if there was no break point between 40 and 75 characters, Exim
7381 left the rest of the message alone. Two changes have been made: (a) I've
7382 reduced the minimum length to 35 characters; (b) if it can't find a break
7383 point between 35 and 75 characters, it looks ahead and uses the first one
7384 that it finds. This may give the occasional overlong line, but at least the
7385 remaining text gets split now.
7387 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7388 file descriptors might be low, and that setting 1000 would always raise it.
7389 It turns out that in some environments, the limit is already over 1000 and
7390 that lowering it causes trouble. So now Exim takes care not to decrease it.
7392 40. When delivering a message, the value of $return_path is set to $sender_
7393 address at the start of routing (routers may change the value). By an
7394 oversight, this default was not being set up when an address was tested by
7395 -bt or -bv, which affected the outcome if any router or filter referred to
7398 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7399 log only when "message" or "log_message" are set. However, if one of the
7400 conditions was an address verification, or a call to a nested ACL, the
7401 messages generated by the underlying test were being passed through. This
7402 no longer happens. The underlying message is available in $acl_verify_
7403 message for both "message" and "log_message" expansions, so it can be
7404 passed through if needed.
7406 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7407 new expansion $bh_ to give the encoded byte string without charset
7408 translation. Translation happens only if iconv() is available; HAVE_ICONV
7409 indicates this at build time. HEADERS_CHARSET gives the charset to
7410 translate to; headers_charset can change it in the configuration, and
7411 "headers charset" can change it in an individual filter file.
7413 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7414 that creates RFC 2047 encoded "words" labels them as that charset instead
7415 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7416 expansion operator; (ii) when Exim creates a From: line for a local
7417 message; (iii) when a header line is rewritten to include a "phrase" part.
7419 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7420 buggy, causing it to skip the first lines of messages whose message ID
7421 ended in 'D'. This would not have bitten before Exim release 4.14, because
7422 message IDs were unlikely to end in 'D' before then. The effect was to have
7423 incorrect size information for certain domains.
7425 45. #include "config.h" was missing at the start of the crypt16.c module. This
7426 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7429 46. If there was a timeout during a "random" callout check, Exim treated it as
7430 a failure of the random address, and carried on sending RSET and the real
7431 address. If the delay was just some slowness somewhere, the response to the
7432 original RCPT would be taken as a response to RSET and so on, causing
7433 mayhem of various kinds.
7435 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7436 when I implemented it. It didn't allow for the fact that some option values
7437 may legitimately be negative (e.g. size_addition), and it didn't even do
7438 the right test for positive values.
7440 48. Domain names in DNS records are case-independent. Exim always looks them up
7441 in lower case. Some resolvers return domain names in exactly the case they
7442 appear in the zone file, that is, they may contain uppercase letters. Not
7443 all resolvers do this - some return always lower case. Exim was treating a
7444 change of case by a resolver as a change of domain, similar to a widening
7445 of a domain abbreviation. This triggered its re-routing code and so it was
7446 trying to route what was effectively the same domain again. This normally
7447 caused routing to fail (because the router wouldn't handle the domain
7448 twice). Now Exim checks for this case specially, and just changes the
7449 casing of the domain that it ultimately uses when it transmits the message
7452 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7455 50. If a filter generated a file delivery with a non-absolute name (possible if
7456 no home directory exists for the router), the forbid_file option was not
7459 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7460 the existing equality matching.
7462 52. Exim was using ints instead of ino_t variables in some places where it was
7463 dealing with inode numbers.
7465 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7466 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7467 called TMPDIR, and if it finds it is different, it changes its value.
7469 54. The smtp_printf() function is now made available to local_scan() so
7470 additional output lines can be written before returning. There is also an
7471 smtp_fflush() function to enable the detection of a dropped connection.
7472 The variables smtp_input and smtp_batched_input are exported to
7475 55. Changed the default runtime configuration: the message "Unknown user"
7476 has been removed from the ACL, and instead placed on the localuser router,
7477 using the cannot_route_message feature. This means that any verification
7478 failures that generate their own messages won't get overridden. Similarly,
7479 the "Unrouteable address" message that was in the ACL for unverifiable
7480 relay addresses has also been removed.
7482 56. Added hosts_avoid_esmtp to the smtp transport.
7484 57. The exicyclog script was not checking for the esoteric option
7485 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7486 will work only if exicyclog is run under the appropriate euid.
7488 58. Following a discussion on the list, the rules by which Exim recognises line
7489 endings on incoming messages have been changed. The -dropcr and drop_cr
7490 options are now no-ops, retained only for backwards compatibility. The
7491 following line terminators are recognized: LF CRLF CR. However, special
7492 processing applies to CR:
7494 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7495 nor a local message in the state where . is a terminator.
7497 (ii) If a bare CR is encountered in a header line, an extra space is added
7498 after the line terminator so as not to end the header. The reasoning
7499 behind this is that bare CRs in header lines are most likely either
7500 to be mistakes, or people trying to play silly games.
7502 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7503 was being incorrectly given as 18 bytes larger than it should have been.
7504 This is a VOB (very old bug).
7506 60. This may never have affected anything current, but just in case it has:
7507 When the local host is found other than at the start of a list of hosts,
7508 the local host, those with the same MX, and any that follow, are discarded.
7509 When the list in question was part of a longer list of hosts, the following
7510 hosts (not currently being processed) were also being discarded. This no
7511 longer happens. I'm not sure if this situation could ever has previously
7514 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7517 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7518 header. This is something that is recommended in a new Internet Draft, and
7519 is something that is documented as being done by Sendmail. There are two
7520 possible values. For messages generated by the autoreply transport, Exim
7523 Auto-Submitted: auto-replied
7525 whereas for all other generated messages (e.g. bounces) it adds
7527 Auto-Submitted: auto-generated
7529 63. The "personal" condition in filters now includes a test for the
7530 Auto-Submitted: header. If it contains the string "auto-" the message it
7531 not considered personal.
7533 64. Added rcpt_include_affixes as a generic transport option.
7535 65. Added queue_only_override (default true).
7537 66. Added the syslog_duplication option.
7539 67. If what should have been the first header line of a message consisted of
7540 a space followed by a colon, Exim was mis-interpreting it as a header line.
7541 It isn't of course - it is syntactically invalid and should therefore be
7542 treated as the start of the message body. The misbehaviour could have
7543 caused a number of strange effects, including loss of data in subsequent
7544 header lines, and spool format errors.
7546 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7547 client host had authenticated. This control can now be exercised by an ACL
7548 for more flexibility.
7550 69. By default, callouts do not happen when testing with -bh. There is now a
7551 variant, -bhc, which does actually run the callout code, including
7552 consulting and updating the callout cache.
7554 70. Added support for saslauthd authentication, courtesy of Alexander
7557 71. If statvfs() failed on the spool or log directories while checking their
7558 size for availability, Exim confusingly gave the error "space shortage".
7559 Furthermore, in debugging mode it crashed with a floating point exception.
7560 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7561 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7562 serious problem, Exim now writes to the main and panic logs when this
7563 happens, with details of the failure. It then refuses to accept the
7564 incoming message, giving the message "spool directory problem" or "log
7565 directory problem" with a 421 code for SMTP messages.
7567 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7568 0, 1, and 2 exist, because some OS complain for execs without them (see
7569 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7570 descriptors. However, the code omitted to check that the open succeeded,
7571 causing mysterious errors if for some reason the permissions on /dev/null
7572 got screwed. Now Exim writes a message to the main and panic logs, and
7573 bombs out if it can't open /dev/null.
7575 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7576 interact so that it is all more flexible. It is supposed to remain
7577 backwards compatible. Also added extra_local_interfaces.
7579 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7580 to bomb out with an assertion failure - to the client this appears as a
7581 connection drop. This problem occurs in the part of the code that was taken
7582 from the Samba project. Fortunately, the assertion is in a very simple
7583 function, so I have fixed this by reproducing the function inline in the
7584 one place where it is called, and arranging for authentication to fail
7585 instead of killing the process with assert().
7587 75. The SPA client code was not working when the server requested OEM rather
7588 than Unicode encoding.
7590 76. Added code to make require_files with a specific uid setting more usable in
7591 the case where statting the file as root fails - usually a non-root-mounted
7592 NFS file system. When this happens and the failure is EACCES, Exim now
7593 forks a subprocess and does the per-uid checking as the relevant uid.
7595 77. Added process_log_path.
7597 78. If log_file_path was not explicitly set, a setting of check_log_space or
7598 check_log_inodes was ignored.
7600 79. If a space check for the spool or log partitions fails, the incident is now
7601 logged. Of course, in the latter case the data may get lost...
7603 80. Added the %p formatting code to string_format() so that it can be used to
7604 print addresses in debug_print(). Adjusted all the address printing in the
7605 debugging in store.c to use %p rather than %d.
7607 81. There was a concern that a line of code in smtp_in.c could overflow a
7608 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7609 initially expressed, the concern was not well-founded, because trailing
7610 spaces are removed early. However, if the trailing spaces were followed by
7611 a NULL, they did not get removed, so the overflow was possible. Two fixes
7614 (a) I re-wrote the offending code in a cleaner fashion.
7615 (b) If an incoming SMTP command contains a NULL character, it is rejected
7618 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7619 runs initgroups(), so that if the Exim user is in any additional groups,
7620 they will be used during message reception.
7626 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob