1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Avoid costly startup code when not strictly needed. This reduces time
10 for some exim process initialisations. It does mean that the logging
11 of TLS configuration problems is only done for the daemon startup.
13 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
15 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
16 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
19 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
20 DKIM signing is being done. Previously a transport_filter would always
21 disable CHUNKING, falling back to traditional DATA.
23 JH/05 Regard command-line receipients as tainted.
25 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
27 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
28 PAM library frees one of the arguments given to it, despite the
29 documentation. Therefore a plain malloc must be used.
31 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
32 on-stack buffers were used, resulting in a taint trap when DSN information
33 copied from a received message was written into the buffer.
35 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
36 the ordering of its ARC headers. This caused a crash.
38 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
39 a new record was being constructed with information from the peer, a trap
42 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
43 installation would get error messages from DMARC verify, when it hit the
44 nonexistent file indicated by the default. Distros wanting DMARC enabled
45 should both provide the file and set the option.
46 Also enforce no DMARC verification for command-line sourced messages.
48 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
49 could, depending on the platform, hang at the STARTTLS response.
51 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
52 message on a connection. Previously if one message had ARC headers and
53 the following one did not, a crash could result when adding an
54 Authentication-Results: header.
56 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
57 Exim main code and Exim-related utities. The introduction of taint
58 tracking also did many adjustments to string handling. Since then, eximon
59 frequently terminated with an assert failure.
61 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
62 check for 452 responses. This slightly helps the inefficieny of doing
63 a large alias-expansion into a recipient-limited target. The max_rcpt
64 transport option still applies (and at the current default, will override
65 the new feature). The check is done for either cause of synch, and forces
66 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
67 connection. The new facility is not tunable at this time.
69 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
70 library live data was being used, so the results became garbage. Make
71 copies while it is still usable.
73 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
74 on delivery deferred (==) and failed (**) lines (if a delivery was
75 attemtped). Previously it was only on completion (=>) lines.
77 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
78 for the expansion of the server_scram_iter and server_scram_salt options.
80 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
81 are now specifically given a NO_DATA response without hitting the system
82 resolver. The library goes on to do the now-standard TXT lookup.
83 Use of dnsdb lookups is not affected.
85 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
86 only retrieve the errormessage once. Previously two calls to dlerror()
87 were used, and the second one (for mainlog/paniclog) retrieved null
90 JH/20 Taint checking: disallow use of tainted data for
91 - the appendfile transport file and directory options
92 - the pipe transport command
93 - file names used by the redirect router (including filter files)
94 Previously this was permitted.
100 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
101 side, to decode a TLS 1.3 packet capture.
103 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
104 Previously the default library behaviour applied, sending two, each in
107 JH/03 Debug output for ACL now gives the config file name and line number for
110 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
112 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
114 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
115 buffer overrun for (non-chunking) other transports.
117 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
118 TLS1.3, means that a server rejecting a client certificate is not visible
119 to the client until the first read of encrypted data (typically the
120 response to EHLO). Add detection for that case and treat it as a failed
121 TLS connection attempt, so that the normal retry-in-clear can work (if
122 suitably configured).
124 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
125 and/or domain. Found and fixed by Jason Betts.
127 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
128 configuration). If a CNAME target was not a wellformed name pattern, a
131 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
132 the OS reports them interleaved with other addresses.
134 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
135 used both for input and for a verify callout, both encrypted, SMTP
136 responses being sent by the server could be lost. This resulted in
137 dropped connections and sometimes bounces generated by a peer sending
140 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
141 string. Previously it was possible to cause undefined behaviour in a
142 library routine (usually a crash). Found by "zerons".
144 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
147 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
148 API was removed, so update to use the newer ones.
150 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
151 any timeout set, is taking a long time. Previously we would hang on to a
152 rotated logfile "forever" if the input was arriving with long gaps
153 (a previous attempt to fix addressed lack, for a long time, of initial
156 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
157 shared (NFS) environment. The length of the tempfile name is now
158 4 + 16 ("hdr.$message_exim_id") which might break on file
159 systems which restrict the file name length to lower values.
162 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
163 shared (NFS) environment.
165 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
166 did for all versions <4.90). Notably -M, -m, --invert, -I may be
169 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
170 on some platforms for bit 31.
172 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
173 to changes apparently associated with TLS1.3 handling some of the APIs
174 previously used were either nonfunctional or inappropriate. Strings
175 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
176 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
177 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
178 This affects log line X= elements, the $tls_{in,out}_cipher variables,
179 and the use of specific cipher names in the encrypted= ACL condition.
181 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
183 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
184 verification result was not updated unless hosts_require_ocsp applied.
186 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
187 queue_list_requires_admin set to false, non-admin users were denied the
190 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
191 directory-of-certs mode. Previously they were advertised despite the
194 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
195 A single TCP connection by a client will now hold a TLS connection open
196 for multiple message deliveries, by default. Previoud the default was to
199 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
200 default. If built with the facility, DANE will be used. The facility
201 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
203 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
204 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
205 must be defined and you must still, unless you define DISABLE_TLS, manage
206 the the include-dir and library-file requirements that go with that
207 choice. Non-TLS builds are still supported.
209 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
212 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
213 default. If the platform supports and has the facility enabled, it will
214 be requested on all coneections.
216 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
217 controlled by the build-time option SUPPORT_PIPE_CONNECT.
219 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
221 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
222 success-DSN messages. Previously the From: header was always the default
223 one for these; the option was ignored.
225 JH/28 Fix the timeout on smtp response to apply to the whole response.
226 Previously it was reset for every read, so a teergrubing peer sending
227 single bytes within the time limit could extend the connection for a
228 long time. Credit to Qualsys Security Advisory Team for the discovery.
230 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
231 delivery address, which leaked information of the results of local
232 forwarding. Change to the original envelope recipient address, per
235 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
236 requested. Previously not bounce was generated and a log entry of
237 error ignored was made.
239 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
241 JH/32 Introduce a general tainting mechanism for values read from the input
242 channel, and values derived from them. Refuse to expand any tainted
243 values, to catch one form of exploit.
245 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
246 was unused and the unexpanded text used for the test. Found and
247 fixed by Ruben Jenster.
249 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
250 an attempt to use a TLS library read routine dereffed a nul pointer,
253 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
254 them in our smtp response.
256 JH/36 Have the general router option retry_use_local_part default to true when
257 any of the restrictive preconditions are set (to anything). Previously it
258 was only for check_local user. The change removes one item of manual
259 configuration which is required for proper retries when a remote router
260 handles a subset of addresses for a domain.
262 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
263 link count into consideration.
265 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
266 caused the extension of big_buffer, the following lines were ignored.
268 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
269 accordance with RFC 2308. Previously there was no expiry, so a longlived
270 receive process (eg. due to ACL delays) versus a short SOA value could
273 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
275 JH/39 Promote DMARC support to mainline.
277 JH/40 Bug 2452: Add a References: header to DSNs.
279 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
280 parameters. The relevant library call is documented as "Deprecated: This
281 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
282 3.6.0, DH parameters are negotiated following RFC7919."
284 HS/06 Change the default of dnssec_request_domains to "*"
286 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
287 carried on and emitted a BDAT command, even when PIPELINING was not
290 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
291 buffer was used for the filename, resulting in a trap when tainted
292 arguments (eg. $domain) were used.
294 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
295 recommended to avoid a possible server-load attack. The feature can be
296 re-enabled via the openssl_options main cofiguration option.
298 JH/45 local_scan API: documented the current smtp_printf() call. This changed
299 for version 4.90 - adding a "more data" boolean to the arguments.
300 Bumped the ABI version number also, this having been missed previously;
301 release versions 4.90 to 4.92.3 inclusive were effectively broken in
302 respect of usage of smtp_printf() by either local_scan code or libraries
303 accessed via the ${dlfunc } expansion item. Both will need coding
304 adjustment for any calls to smtp_printf() to match the new function
305 signature; a FALSE value for the new argument is always safe.
307 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
308 the file-offset (which the Linux syscall does, and exim expects); this
309 resulted in an indefinite loop.
311 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
312 to do ARC verification. The Authentication-Results: header line added
313 by the configuration then had no ARC item.
319 JH/01 Remove code calling the customisable local_scan function, unless a new
320 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
322 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
323 non-signal-safe functions being used.
325 JH/03 Bug 2269: When presented with a received message having a stupidly large
326 number of DKIM-Signature headers, disable DKIM verification to avoid
327 a resource-consumption attack. The limit is set at twenty.
329 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
330 report of oldest_pass in ${authres } in consequence, and separate out
331 some descriptions of reasons for verification fail.
333 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
334 files in the spool were present and unlocked. A queue-runner could spot
335 them, resulting in a duplicate delivery. Fix that by doing the unlock
336 after the unlink. Investigation by Tim Stewart. Take the opportunity to
337 add more error-checking on spoolfile handling while that code is being
340 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
341 No known attacks, no CVE, this is defensive hardening.
343 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
344 a queue-runner could start a delivery while other operations were ongoing.
345 Cutthrough delivery was a common victim, resulting in duplicate delivery.
346 Found and investigated by Tim Stewart. Fix by using the open message data
347 file handle rather than opening another, and not locally closing it (which
348 releases a lock) for that case, while creating the temporary .eml format
349 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
351 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
352 $sender_verify_failure/$recipient_verify_failure to "random".
354 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
357 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
358 Previously this would segfault.
360 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
363 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
364 like zero, since the resolver should be doing this for us, But we need one
365 as a CNAME but no MX presence gets the CNAME returned; we need to check
366 that doesn't point to an MX to declare it "no MX returned" rather than
367 "error, loop". A new main option is added so the older capability of
368 following some limited number of chain links is maintained.
370 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
372 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
373 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
374 now-deprecated earlier definitions used only specified the range up to TLS
375 1.2 (in the older-version library docs).
377 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
379 JH/15 Rework TLS client-side context management. Stop using a global, and
380 explicitly pass a context around. This enables future use of TLS for
381 connections to service-daemons (eg. malware scanning) while a client smtp
382 connection is using TLS; with cutthrough connections this is quite likely.
384 JH/16 Fix ARC verification to do AS checks in reverse order.
386 JH/17 Support a "tls" option on the ${readsocket } expansion item.
388 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
389 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
390 Previously the "utf8" would be re-prepended for every additional message.
392 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
393 Previously thery were accepted, resulting in issues when attempting to
394 forward messages to a non-supporting MTA.
396 PP/02 Let -n work with printing macros too, not just options.
398 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
399 one parent address was copied, and bogus data was used at delivery-logging
400 time. Either a crash (after delivery) or bogus log data could result.
401 Discovery and analysis by Tim Stewart.
403 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
404 Previously if the string ended mid-character, we did not insert the
405 promised '?' replacement.
407 PP/04 Documentation: current string operators work on bytes, not codepoints.
409 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
410 should pack well giving a smaller memory footprint so better caching and
411 therefore performance. Group the declarations where this can't be done so
412 that the byte-sized flag variables are not interspersed among pointer
413 variables, giving a better chance of good packing by the compiler.
415 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
416 non-null, to avoid issues with sites running BATV. Previously reports were
417 sent with an empty envelope sender so looked like bounces.
419 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
420 The ignore_error flag wasn't being returned from the filter subprocess so
421 was not set for later routers. Investigation and fix by Matthias Kurz.
423 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
424 and a msg:complete for the whole, when a message is manually removed using
425 -Mrm. Developement by Matthias Kurz, hacked on by JH.
427 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
428 a "Gnu special" function, asprintf() in the DB utility binary builds; I
429 hope that is portable enough.
431 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
432 requiring a known-CA anchor certificate; make it now rely entirely on the
433 TLSA as an anchor. Checking the name on the leaf cert against the name
434 on the A-record for the host is still done for TA (but not for EE mode).
436 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
437 included in delivery lines for non-proxied connections, when compiled with
438 SUPPORT_SOCKS and running with proxy logging enabled.
440 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
441 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
442 move the existing event to fire before the normal logging of message
443 failure so that custom logging is bracketed by normal logging.
445 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
446 msg:fail:internal event. Developement by Matthias Kurz.
448 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
449 far too small for todays use of crypto signatures stored there. Go all
450 the way to the max DNS message size of 64kB, even though this might be
451 overmuch for IOT constrained device use.
453 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
454 copy a string over itself. The library routine is documented as not
455 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
457 JH/32 For main options check_spool_space and check_inode_space, where the
458 platform supports 64b integers, support more than the previous 2^31 kB
459 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
460 the previous G, M, k.
462 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
463 $authenticated_fail_id variable on authentication failure. Previously
466 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
467 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
468 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
469 GNUTLS_SEC_PARAM_MEDIUM.
471 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
472 side. Previously we would continue as if no SNI had been received.
474 JH/36 Harden the handling of string-lists. When a list consisted of a sole
475 "<" character, which should be a list-separator specification, we walked
476 off past the nul-terimation.
478 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
479 causes) even when the retry time is not yet met. Previously they were
480 not, meaning that when (say) an account was over-quota and temp-rejecting,
481 and multiple senders' messages were queued, only one sender would get
482 notified on each configured delay_warning cycle.
484 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
486 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
487 cipher-suites, an error can be left on the stack even for a succeeding
488 accept; this results in impossible error messages when a later operation
491 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
492 return error codes indicating retry. Under TLS1.3 this becomes required.
494 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
495 it only wrote the new authenticators, resulting in a lack of tracking of
496 peer changes of ESMTP extensions until the next cache flush.
498 JH/41 Fix the loop reading a message header line to check for integer overflow,
499 and more-often against header_maxsize. Previously a crafted message could
500 induce a crash of the recive process; now the message is cleanly rejected.
502 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
503 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
509 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
510 When redis_servers is set to a list of > 1 element, and the Redis servers
511 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
512 case of MOVED into a DEFER case instead, thus moving the query onto the
513 next server in the list. For a cluster of N elements, all N servers must
514 be defined in redis_servers.
516 GF/02 Catch and remove uninitialized value warning in exiqsumm
517 Check for existence of @ARGV before looking at $ARGV[0]
519 JH/01 Replace the store_release() internal interface with store_newblock(),
520 which internalises the check required to safely use the old one, plus
521 the allocate and data copy operations duplicated in both (!) of the
522 extant use locations.
524 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
525 modifier. This matches the restriction on the commandline.
527 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
528 Previously only the last row was returned.
530 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
531 we assumed that tags in the header were well-formed, and parsed the
532 element content after inspecting only the first char of the tag.
533 Assumptions at that stage could crash the receive process on malformed
536 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
537 While running the DKIM ACL we operate on the Permanent memory pool so that
538 variables created with "set" persist to the DATA ACL. Also (at any time)
539 DNS lookups that fail create cache records using the Permanent pool. But
540 expansions release any allocations made on the current pool - so a dnsdb
541 lookup expansion done in the DKIM ACL releases the memory used for the
542 DNS negative-cache, and bad things result. Solution is to switch to the
543 Main pool for expansions.
544 While we're in that code, add checks on the DNS cache during store_reset,
545 active in the testsuite.
546 Problem spotted, and debugging aided, by Wolfgang Breyha.
548 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
549 When none of the hosts presented to a transport match an already-open
550 connection, close it and proceed with the list. Previously we would
551 queue the message. Spotted by Lena with Yahoo, probably involving
554 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
555 Previously a spurious "250 OK id=" response was appended to the proper
558 JH/08 The "support for" informational output now, which built with Content
559 Scanning support, has a line for the malware scanner interfaces compiled
560 in. Interface can be individually included or not at build time.
562 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
563 by the template makefile "src/EDITME". The "STREAM" support for an older
564 ClamAV interface method is removed.
566 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
567 rows affected is given instead).
569 JH/11 The runtime Berkeley DB library version is now additionally output by
570 "exim -d -bV". Previously only the compile-time version was shown.
572 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
573 SMTP connection. Previously, when one had more receipients than the
574 first, an abortive onward connection was made. Move to full support for
575 multiple onward connections in sequence, handling cutthrough connection
576 for all multi-message initiating connections.
578 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
579 routers. Previously, a multi-recipient message would fail to match the
580 onward-connection opened for the first recipient, and cause its closure.
582 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
583 a timeout on read on a GnuTLS initiating connection, resulting in the
584 initiating connection being dropped. This mattered most when the callout
585 was marked defer_ok. Fix to keep the two timeout-detection methods
588 JH/15 Relax results from ACL control request to enable cutthrough, in
589 unsupported situations, from error to silently (except under debug)
590 ignoring. This covers use with PRDR, frozen messages, queue-only and
593 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
595 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
596 metadata, resulting in a crash in free().
598 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
599 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
600 Broken also in d185889f4, with init system revamp.
602 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
603 Previously we abruptly closed the connection after reading a malware-
604 found indication; now we go on to read the "scan ok" response line,
607 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
608 ACL. Previously, a crash would result.
610 JH/19 Speed up macro lookups during configuration file read, by skipping non-
611 macro text after a replacement (previously it was only once per line) and
612 by skipping builtin macros when searching for an uppercase lead character.
614 JH/20 DANE support moved from Experimental to mainline. The Makefile control
615 for the build is renamed.
617 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
618 was allocated for every new TLS startup, meaning one per message. Fix
619 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
621 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
622 reported the original. Fix to report (as far as possible) the ACL
623 result replacing the original.
625 JH/23 Fix memory leak during multi-message connections using STARTTLS under
626 OpenSSL. Certificate information is loaded for every new TLS startup,
627 and the resources needed to be freed.
629 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
631 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
634 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
635 DATA response info to the (existing) per-recipient response info for
636 the "C=" log element. It can have useful tracking info from the
637 destination system. Patch from Simon Arlott.
639 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
640 length value. Previously this would segfault.
642 HS/02 Support Avast multiline protoocol, this allows passing flags to
643 newer versions of the scanner.
645 JH/28 Ensure that variables possibly set during message acceptance are marked
646 dead before release of memory in the daemon loop. This stops complaints
647 about them when the debug_store option is enabled. Discovered specifically
648 for sender_rate_period, but applies to a whole set of variables.
649 Do the same for the queue-runner and queue-list loops, for variables set
650 from spool message files. Do the same for the SMTP per-message loop, for
651 certain variables indirectly set in ACL operations.
653 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
654 as a multi-recipient message from a mailinglist manager). The coding had
655 an arbitrary cutoff number of characters while checking for more input;
656 enforced by writing a NUL into the buffer. This corrupted long / fast
657 input. The problem was exposed more widely when more pipelineing of SMTP
658 responses was introduced, and one Exim system was feeding another.
659 The symptom is log complaints of SMTP syntax error (NUL chars) on the
660 receiving system, and refused recipients seen by the sending system
661 (propating to people being dropped from mailing lists).
662 Discovered and pinpointed by David Carter.
664 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
665 replaced by the ${authresults } expansion.
667 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
669 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
670 allows proper process termination in container environments.
672 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
673 Previously the "final dot" had a newline after it; ensure it is CR,LF.
675 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
676 and "err_perm", deprecated since 4.83 when the RFC-defined words
677 "temperror" and "permerror" were introduced.
679 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
680 transport-filters or DKIM-signing. The restriction was lost in the
681 consolidation of verify-callout and delivery SMTP handling.
682 Extend the restriction to also cover ARC-signing.
684 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
685 in defer=pass mode supply a 450 to the initiator. Previously the message
688 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
689 tls_require_ciphers is used as before.
691 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
692 "pass_unscanned". Only tmpfails from the scanner are written to
693 the paniclog, as they may require admin intervention (permission
694 denied, license issues). Other scanner errors (like decompression
695 bombs) do not cause a paniclog entry.
697 JH/36 Fix reinitialisation of DKIM logging variable between messages.
698 Previously it was possible to log spurious information in receive log
701 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
702 triggered odd behaviour from Outlook Express clients.
704 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
707 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
708 since the IETF WG has not yet settled on that versus the original
709 "bare" representation.
711 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
712 Previously the millisecond value corrupted the output.
713 Fix also for syslog_pid=no and log_selector +pid, for which the pid
714 corrupted the output.
720 JH/01 Rework error string handling in TLS interface so that the caller in
721 more cases is responsible for logging. This permits library-sourced
722 string to be attached to addresses during delivery, and collapses
723 pairs of long lines into single ones.
725 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
726 during configuration. Wildcards are allowed and expanded.
728 JH/02 Rework error string handling in DKIM to pass more info back to callers.
729 This permits better logging.
731 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
732 do not close it down and have the child transport start it up again on
733 the passed-on TCP connection. Instead, proxy the child (and any
734 subsequent ones) for TLS via a unix-domain socket channel. Logging is
735 affected: the continued delivery log lines do not have any DNSSEC, TLS
736 Certificate or OCSP information. TLS cipher information is still logged.
738 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
739 identical IP addresses on different listening ports. Will also affect
742 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
743 add noisy ifdef guards to special-case this sillyness.
744 Patch from Bernd Kuhls.
746 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
747 than 255 are no longer allowed.
749 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
750 Disable the session-cache too, which might reduce our load. Since we
751 currrectly use a new context for every connection, both as server and
752 client, there is no benefit for these.
753 GnuTLS appears to not support tickets server-side by default (we don't
754 call gnutls_session_ticket_enable_server()) but client side is enabled
755 by default on recent versions (3.1.3 +) unless the PFS priority string
758 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
759 <https://reproducible-builds.org/specs/source-date-epoch/>.
761 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
762 the check for any unsuccessful recipients did not notice the limit, and
763 erroneously found still-pending ones.
765 JH/08 Pipeline CHUNKING command and data together, on kernels that support
766 MSG_MORE. Only in-clear (not on TLS connections).
768 JH/09 Avoid using a temporary file during transport using dkim. Unless a
769 transport-filter is involved we can buffer the headers in memory for
770 creating the signature, and read the spool data file once for the
771 signature and again for transmission.
773 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
774 4.77 as the kernel support then wasn't solid, having issues in 64bit
775 mode. Now, it's been long enough. Add support for FreeBSD also.
777 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
778 case where the routing stage had gathered several addresses to send to
779 a host before calling the transport for the first, we previously failed
780 to close down TLS in the old transport process before passing the TCP
781 connection to the new process. The new one sent a STARTTLS command
782 which naturally failed, giving a failed delivery and bloating the retry
783 database. Investigation and fix prototype from Wolfgang Breyha.
785 JH/12 Fix check on SMTP command input synchronisation. Previously there were
786 false-negatives in the check that the sender had not preempted a response
787 or prompt from Exim (running as a server), due to that code's lack of
788 awareness of the SMTP input buffering.
790 PP/04 Add commandline_checks_require_admin option.
791 Exim drops privileges sanely, various checks such as -be aren't a
792 security problem, as long as you trust local users with access to their
793 own account. When invoked by services which pass untrusted data to
794 Exim, this might be an issue. Set this option in main configuration
795 AND make fixes to the calling application, such as using `--` to stop
798 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
799 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
800 responses to those, into a single TLS record each way (this usually means
801 a single packet). As a side issue, smtp_enforce_sync now works on TLS
804 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
805 affects you only if you're dancing at the edge of the param size limits.
806 If you are, and this message makes sense to you, then: raise the
807 configured limit or use OpenSSL 1.1. Nothing we can do for older
810 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
811 cmdline element to get the documented default one. Previously it was
814 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
815 are unsuable for non-message-specific reasons - eg. connection timeout,
816 banner-time rejection.
818 JH/16 Fix logging of delivery remote port, when specified by router, under
821 PP/06 Repair manualroute's ability to take options in any order, even if one
822 is the name of a transport.
825 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
827 JH/17 Change the list-building routines interface to use the expanding-string
828 triplet model, for better allocation and copying behaviour.
830 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
831 Previously it was constructed the first time a possibly-matching string
832 was met in the configuration file input during startup; now it is done
835 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
836 compatible one, to avoid the (poorly documented) possibility of a config
837 file in the working directory redirecting the DB files, possibly correpting
838 some existing file. CVE-2017-10140 assigned for BDB.
840 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
841 cache-hot. Previously, although the result was properly cached, the
842 initial verify call returned a defer.
844 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
845 the main verify for receipient in uncached-mode.
847 JH/22 Retire historical build files to an "unsupported" subdir. These are
848 defined as "ones for which we have no current evidence of testing".
850 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
851 if present. Previously it was ignored.
853 JH/24 Start using specified-initialisers in C structure init coding. This is
854 a C99 feature (it's 2017, so now considered safe).
856 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
857 if was a fixed-sized field and bitmask ops via macros; it is now more
860 PP/07 GitHub PR 56: Apply MariaDB build fix.
861 Patch provided by Jaroslav Škarvada.
863 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
864 during Coverity cleanups [4.87 JH/47]
865 Diagnosis and fix provided by Michael Fischer v. Mollard.
867 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
868 the right size to place the terminating semicolon on its own folded
869 line, the header hash was calculated to an incorrect value thanks to
870 the (relaxed) space the fold became.
872 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
873 and confused the parent.
875 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
876 which could crash as a result. This could lead to undeliverable messages.
878 JH/28 Logging: "next input sent too soon" now shows where input was truncated
881 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
882 matters on fast-turnover and PID-randomising systems, which were getting
883 out-of-order delivery.
885 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
886 a possibly-overlapping copy. The symptom was that "Remote host closed
887 connection in response to HELO" was logged instead of the actual 4xx
890 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
891 Previously only that bufferd was discarded, resulting in SYMTP command
894 JH/32 DKIM: when a message has multiple signatures matching an identity given
895 in dkim_verify_signers, run the dkim acl once for each. Previously only
896 one run was done. Bug 2189.
898 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
899 "panic the current process" to "deliberately defer". The panic log is
900 still written with the problem list name; the mail and reject logs now
901 get a temp-reject line for the message that was being handled, saying
902 something like "domains check lookup or other defer". The SMTP 451
903 message is still "Temporary local problem".
905 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
906 A crafted sequence of BDAT commands could result in in-use memory beeing
907 freed. CVE-2017-16943.
909 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
910 from SMTP input. Previously it was always done; now only done for DATA
911 and not BDAT commands. CVE-2017-16944.
913 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
914 to the message (such as an overlong header line). Previously this was
915 not done and we did not exit BDAT mode. Followon from the previous item
916 though a different problem.
922 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
923 than -2003 did; needs libidn2 in addition to libidn.
925 JH/02 The path option on a pipe transport is now expanded before use.
927 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
928 Patch provided by "Björn", documentation fix added too.
930 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
931 missing a wire-to-host endian conversion.
933 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
934 close after a BDAT command line could be taken as a following command,
935 giving a synch failure. Fix by only checking for synch immediately
936 before acknowledging the chunk.
938 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
939 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
940 macro. Patches provided by Josh Soref.
942 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
943 Previously we did not; the RFC seems ambiguous and VRFY is not listed
944 by IANA as a service extension. However, John Klensin suggests that we
947 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
948 the dkim code may be unix-mode line endings rather than smtp wire-format
949 CRLF, so prepend a CR to any bare LF.
951 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
952 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
954 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
955 the first were themselves being wrongly included in the feed into dkim
956 processing; with most chunk sizes in use this resulted in an incorrect
957 body hash calculated value.
959 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
960 DKIM signature block, for verification. Although advised against by
961 standards it is specifically not ruled illegal.
963 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
965 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
966 missing a body hash (the bh= tag).
968 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
969 It seems that HAProxy sends the Proxy Protocol information in clear and
970 only then does a TLS startup, so do the same.
972 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
973 TCP connections (such as for Spamd) unless the daemon successfully set
974 Fast Open mode on its listening sockets. This fixes breakage seen on
975 too-old kernels or those not configured for Fast Open, at the cost of
976 requiring both directions being enabled for TFO, and TFO never being used
977 by non-daemon-related Exim processes.
979 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
980 endings, at least on the first header line. Try to canonify any that get
981 past that check, despite the cost.
983 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
984 now limited to an arbitrary five deep, while parsing addresses with the
985 strip_excess_angle_brackets option enabled.
987 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
988 instead leave the unprompted TLS handshake in socket buffer for the
989 TLS library to consume.
991 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
993 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
995 JH/16 Drop variables when they go out of scope. Memory management drops a whole
996 region in one operation, for speed, and this leaves assigned pointers
997 dangling. Add checks run only under the testsuite which checks all
998 variables at a store-reset and panics on a dangling pointer; add code
999 explicitly nulling out all the variables discovered. Fixes one known
1000 bug: a transport crash, where a dangling pointer for $sending_ip_address
1001 originally assigned in a verify callout, is re-used.
1003 PP/06 Drop '.' from @INC in various Perl scripts.
1005 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1007 PP/08 Reduce a number of compilation warnings under clang; building with
1008 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1009 should be warning-free.
1011 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1013 HS/01 Fix portability problems introduced by PP/08 for platforms where
1014 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1016 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1017 chunk. This allows us to accept broken chunked messages. We need a more
1018 general solution here.
1020 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1021 already-broken messages in the queue.
1023 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1025 JH/19 Fix reference counting bug in routing-generated-address tracking.
1031 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1032 supports it and a size is available (ie. the sending peer gave us one).
1034 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1035 years of being deprecated). The replacements are the ACLs
1036 acl_smtp_mime and acl_not_smtp_mime.
1038 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1039 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1040 or even an in-clear connection were permitted. Now, if the host lookup
1041 was dnssec and dane was requested then the host is only used if the
1042 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1043 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1044 if one fails this test.
1045 This means that a poorly-configured remote DNS will make it incommunicado;
1046 but it protects against a DNS-interception attack on it.
1048 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1049 non-noisy when a race steals the message being considered.
1051 JH/05 If main configuration option tls_certificate is unset, generate a
1052 self-signed certificate for inbound TLS connections.
1054 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1055 in rewrites and routers.
1057 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1058 and logged a warning sing 4.83; now they are a configuration file error.
1060 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1061 (lacking @domain). Apply the same qualification processing as RCPT.
1063 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1065 JH/10 Support ${sha256:} applied to a string (as well as the previous
1068 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1069 a cutthrough deliver is pending, as we always want to make a connection.
1070 This also avoids re-routing the message when later placing the cutthrough
1071 connection after a verify cache hit.
1072 Do not update it with the verify result either.
1074 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1075 when routing results in more than one destination address.
1077 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1078 signing (which inhibits the cutthrough capability). Previously only
1079 the presence of an option was tested; now an expansion evaluating as
1080 empty is permissible (obviously it should depend only on data available
1081 when the cutthrough connection is made).
1083 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1084 the relevant preceding SMTP command did not note the pipelining mode.
1086 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1087 Previously they were not counted.
1089 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1090 as one having no matching records. Previously we deferred the message
1091 that needed the lookup.
1093 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1094 distinguished as "(=".
1096 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1097 for missing MX records. Previously it only worked for missing A records.
1099 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1101 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1102 after the data-go-ahead and data-ack. Patch from Jason Betts.
1104 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1105 even for a "none" policy. Patch from Tony Meyer.
1107 JH/22 Fix continued use of a connection for further deliveries. If a port was
1108 specified by a router, it must also match for the delivery to be
1111 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1112 When one of the recipients of a message was unsuitable for the connection
1113 (has no matching addresses), we lost track of needing to mark it
1114 deferred. As a result mail would be lost.
1116 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1118 JH/25 Decoding ACL controls is now done using a binary search; the source code
1119 takes up less space and should be simpler to maintain. Merge the ACL
1120 condition decode tables also, with similar effect.
1122 JH/26 Fix problem with one_time used on a redirect router which returned the
1123 parent address unchanged. A retry would see the parent address marked as
1124 delivered, so not attempt the (identical) child. As a result mail would
1127 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1128 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1129 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1132 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1133 Main config options check_{log,spool}_{inodes,space} are now
1134 100 inodes, 10MB unless set otherwise in the configuration.
1136 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1137 Previously it only applied to the main-section connection policy
1140 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1142 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1143 by me. Added RFC7919 DH primes as an alternative.
1145 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1146 are not in the system include path.
1148 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1149 GnuTLS, when a session startup failed (eg because the client disconnected)
1150 Exim did stdio operations after fclose. This was exposed by a recent
1151 change which nulled out the file handle after the fclose.
1153 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1154 signed directly by the cert-signing cert, rather than an intermediate
1155 OCSP-signing cert. This is the model used by LetsEncrypt.
1157 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1159 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1160 an incoming connection.
1162 HS/02 Bug 1802: Do not half-close the connection after sending a request
1165 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1166 fallback to "prime256v1".
1168 JH/34 SECURITY: Use proper copy of DATA command in error message.
1169 Could leak key material. Remotely exploitable. CVE-2016-9963.
1175 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1176 and 3.4.4 - once the server is enabled to respond to an OCSP request
1177 it does even when not requested, resulting in a stapling non-aware
1178 client dropping the TLS connection.
1180 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1181 support variable-length bit vectors. No functional change.
1183 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1184 The I= interface field on outgoing lines is now after the H= remote
1185 host field, same as incoming lines. There is a separate
1186 outgoing_interface log selector which allows you to disable the
1189 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1190 If not running log_selector +smtp_connection the mainlog would be held
1191 open indefinitely after a "too many connections" event, including to a
1192 deleted file after a log rotate. Leave the per net connection logging
1193 leaving it open for efficiency as that will be quickly detected by the
1194 check on the next write.
1196 HS/01 Bug 1671: Fix post transport crash.
1197 Processing the wait-<transport> messages could crash the delivery
1198 process if the message IDs didn't exist for some reason. When
1199 using 'split_spool_directory=yes' the construction of the spool
1200 file name failed already, exposing the same netto behaviour.
1202 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1203 mime_regex ACL conditions.
1205 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1206 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1207 to HELO, local diagnostic string.
1209 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1210 log line to debug. Even when configured with a tls authenticator many
1211 client connections are expected to not authenticate in this way, so
1212 an authenticate fail is not an error.
1214 HS/02 Add the Exim version string to the process info. This way exiwhat
1215 gives some more detail about the running daemon.
1217 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1218 matter for fast-change records such as DNSBLs.
1220 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1221 retry record, even if constant. There may be multiple transports with
1222 different interface settings and the retry behaviour needs to be kept
1225 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1227 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1229 JH/10 Bug 840: fix log_defer_output option of pipe transport
1231 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1232 hosts_randomize. This matters a lot when combined with mua_wrapper.
1234 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1235 ${quote_pgsql:<string>} operator.
1237 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1239 JH/14 Change header file name for freeradius-client. Relevant if compiling
1240 with Radius support; from the Gentoo tree and checked under Fedora.
1242 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1244 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1245 option transport_filter, meaning no filtering.
1247 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1249 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1250 defaults to "*" (all hosts). The variable is now available when not built
1251 with TLS, default unset, mainly to enable keeping the testsuite sane.
1252 If a server certificate is not supplied (via tls_certificate) an error is
1253 logged, and clients will find TLS connections fail on startup. Presumably
1254 they will retry in-clear.
1255 Packagers of Exim are strongly encouraged to create a server certificate
1256 at installation time.
1258 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1259 with the $config_file variable.
1261 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1262 in transport context, after the attempt, and per-recipient. The latter type
1263 is per host attempted. The event data is the error message, and the errno
1264 information encodes the lookup type (A vs. MX) used for the (first) host,
1265 and the trailing two digits of the smtp 4xx response.
1267 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1268 to write to mainlog (or rejectlog, paniclog) in the window between file
1269 creation and permissions/ownership being changed. Particularly affects
1270 installations where exicyclog is run as root, rather than exim user;
1271 result is that the running daemon panics and dies.
1273 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1275 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1276 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1277 "pri" and "weight". Note that the previous implicit priority given by the
1278 list order is no longer honoured.
1280 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1281 for DKIM processing.
1283 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1284 by defining SUPPORT_SOCKS.
1286 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1287 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1288 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1289 variables are renamed to proxy_{local,external}_{address,port}.
1291 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1292 for a build by defining SUPPORT_I18N
1294 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1295 of the query string, and make ${quote_redis:} do that quoting.
1297 JH/29 Move Events support from Experimental to mainline, enabled by default
1298 and removable for a build by defining DISABLE_EVENT.
1300 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1302 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1303 cached by the daemon.
1305 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1306 by defining LOOKUP_REDIS. The libhiredis library is required.
1308 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1309 keys are given for lookup.
1311 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1312 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1313 only supported when built with TLS support. The PolarSSL SHA routines
1314 are still used when the TLS library is too old for convenient support.
1316 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1317 openssl_options), for security. OpenSSL forces this from version 1.1.0
1318 server-side so match that on older versions.
1320 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1321 allocation for $value could be released as the expansion processing
1322 concluded, but leaving the global pointer active for it.
1324 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1325 and to use the domains and local_parts ACL conditions.
1327 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1328 incorrectly not doubled on cutthrough transmission, hence seen as a
1329 body-termination at the receiving system - resulting in truncated mails.
1330 Commonly the sender saw a TCP-level error, and retransmitted the message
1331 via the normal store-and-forward channel. This could result in duplicates
1332 received - but deduplicating mailstores were liable to retain only the
1333 initial truncated version.
1335 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1337 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1339 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1340 we're in there, support oversigning also; bug 1309.
1342 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1344 HS/04 Add support for keep_environment and add_environment options.
1346 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1347 either intentional arithmetic overflow during PRNG, or testing config-
1350 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1351 delivery resulted in actual delivery. Cancel cutthrough before DATA
1354 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1355 rejects a recipient: pass the reject to the originator.
1357 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1358 Many were false-positives and ignorable, but it's worth fixing the
1361 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1362 for the new environment-manipulation done at startup. Move the routines
1363 from being local to tls.c to being global via the os.c file.
1365 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1366 an extract embedded as result-arg for a map, the first arg for extract
1367 is unavailable so we cannot tell if this is a numbered or keyed
1368 extraction. Accept either.
1374 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1377 JH/02 The smtp transport option "multi_domain" is now expanded.
1379 JH/03 The smtp transport now requests PRDR by default, if the server offers
1382 JH/04 Certificate name checking on server certificates, when exim is a client,
1383 is now done by default. The transport option tls_verify_cert_hostnames
1384 can be used to disable this per-host. The build option
1385 EXPERIMENTAL_CERTNAMES is withdrawn.
1387 JH/05 The value of the tls_verify_certificates smtp transport and main options
1388 default to the word "system" to access the system default CA bundle.
1389 For GnuTLS, only version 3.0.20 or later.
1391 JH/06 Verification of the server certificate for a TLS connection is now tried
1392 (but not required) by default. The verification status is now logged by
1393 default, for both outbound TLS and client-certificate supplying inbound
1396 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1399 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1400 Status Notification (bounce) messages are now MIME format per RFC 3464.
1401 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1402 under the control of the dsn_advertise_hosts option, and routers may
1403 have a dsn_lasthop option.
1405 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1406 default, modifiable by a malware= option. The list separator for
1407 the options can now be changed in the usual way. Bug 68.
1409 JH/10 The smtp_receive_timeout main option is now expanded before use.
1411 JH/11 The incoming_interface log option now also enables logging of the
1412 local interface on delivery outgoing connections.
1414 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1415 if the interface and destination host and port all match.
1417 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1420 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1421 Patch from Andrew Lewis.
1423 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1424 now supports optional time-restrictions, weighting, and priority
1425 modifiers per server. Patch originally by <rommer@active.by>.
1427 JH/16 The spamd_address main option now supports a mixed list of local
1428 and remote servers. Remote servers can be IPv6 addresses, and
1429 specify a port-range.
1431 JH/17 Bug 68: The spamd_address main option now supports an optional
1432 timeout value per server.
1434 JH/18 Bug 1581: Router and transport options headers_add/remove can
1435 now have the list separator specified.
1437 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1440 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1443 JH/21 Support for the A6 type of dns record is withdrawn.
1445 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1446 rather than the verbs used.
1448 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1449 from 255 to 1024 chars.
1451 JH/24 Verification callouts now attempt to use TLS by default.
1453 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1454 are generic router options now. The defaults didn't change.
1456 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1457 Original patch from Alexander Shikoff, worked over by JH.
1459 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1460 files if ZCAT_COMMAND is not executable.
1462 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1464 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1466 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1467 Normally benign, it bites when the pair was led to by a CNAME;
1468 modern usage is to not canonicalize the domain to a CNAME target
1469 (and we were inconsistent anyway for A-only vs AAAA+A).
1471 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1473 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1474 when evaluating $sender_host_dnssec.
1476 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1477 $sender_helo_dnssec variable.
1479 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1481 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1483 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1485 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1486 documented as working, but never had. Support all but $spam_report.
1488 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1489 added for tls authenticator.
1491 HS/03 Add perl_taintmode main config option
1497 TL/01 When running the test suite, the README says that variables such as
1498 no_msglog_check are global and can be placed anywhere in a specific
1499 test's script, however it was observed that placement needed to be near
1500 the beginning for it to behave that way. Changed the runtest perl
1501 script to read through the entire script once to detect and set these
1502 variables, reset to the beginning of the script, and then run through
1503 the script parsing/test process like normal.
1505 TL/02 The BSD's have an arc4random API. One of the functions to induce
1506 adding randomness was arc4random_stir(), but it has been removed in
1507 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1508 function when detected.
1510 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1511 cause callback expansion.
1513 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1514 syntax errors in an expansion can be treated as a string instead of
1515 logging or causing an error, due to the internal use of bool_lax
1516 instead of bool when processing it.
1518 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1519 server certificates when making smtp deliveries.
1521 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1523 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1525 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1527 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1528 Merged patch from Sebastian Wiedenroth.
1530 JH/05 Fix results-pipe from transport process. Several recipients, combined
1531 with certificate use, exposed issues where response data items split
1532 over buffer boundaries were not parsed properly. This eventually
1533 resulted in duplicates being sent. This issue only became common enough
1534 to notice due to the introduction of connection certificate information,
1535 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1537 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1538 size buffer was used, resulting in syntax errors when an expansion
1541 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1542 version 3.3.6 or later.
1544 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1545 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1546 both become "event_action", the variables become $event_name, $event_data
1547 and $event_defer_errno. There is a new variable $verify_mode, usable in
1548 routers, transports and related events. The tls:cert event is now also
1549 raised for inbound connections, if the main configuration event_action
1552 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1553 early OCSP support, but no stapling (appears to be less than 1.0.0).
1555 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1556 server certificate names available under the smtp transport option
1557 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1560 JH/10 Time-related extraction expansions from certificates now use the main
1561 option "timezone" setting for output formatting, and are consistent
1562 between OpenSSL and GnuTLS compilations. Bug 1541.
1564 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1565 encoded parameter in the incoming message. Bug 1558.
1567 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1568 include certificate info, eximon was claiming there were spoolfile
1571 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1573 JH/14 Log delivery-related information more consistently, using the sequence
1574 "H=<name> [<ip>]" wherever possible.
1576 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1577 are problematic for Debian distribution, omit them from the release
1580 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1582 JH/16 Fix string representation of time values on 64bit time_t architectures.
1585 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1586 output list separator was used.
1591 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1592 checkers that were complaining about end of non-void function with no
1595 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1596 This was a regression introduced in 4.83 by another bugfix.
1598 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1600 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1601 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1607 TF/01 Correctly close the server side of TLS when forking for delivery.
1609 When a message was received over SMTP with TLS, Exim failed to clear up
1610 the incoming connection properly after forking off the child process to
1611 deliver the message. In some situations the subsequent outgoing
1612 delivery connection happened to have the same fd number as the incoming
1613 connection previously had. Exim would try to use TLS and fail, logging
1614 a "Bad file descriptor" error.
1616 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1617 utilities have not been installed.
1619 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1620 temporary space as the ACL may create new global variables.
1622 TL/01 LDAP support uses per connection or global context settings, depending
1623 upon the detected version of the libraries at build time.
1625 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1626 to extract and use the src ip:port in logging and expansions as if it
1627 were a direct connection from the outside internet. PPv2 support was
1628 updated based on HAProxy spec change in May 2014.
1630 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1632 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1633 Properly escape header and check for NULL return.
1635 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1638 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1640 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1641 characters in header names, implemented as a verify condition.
1642 Contributed by Michael Fischer v. Mollard.
1644 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1645 results permerror and temperror. Previous values are deprecated but
1646 still accepted. In a future release, err_perm and err_temp will be
1647 completely removed, which will be a backward incompatibility if the
1648 ACL tests for either of these two old results. Patch contributed by
1649 user bes-internal on the mailing list.
1651 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1653 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1654 selectors, in both main and reject logs.
1656 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1659 JH/07 Add malware type "sock" for talking to simple daemon.
1661 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1663 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1664 routers/transports under cutthrough routing.
1666 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1667 numbers. Touch up "bool" conditional to keep the same definition.
1669 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1671 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1673 JH/12 Expand items in router/transport headers_add or headers_remove lists
1674 individually rather than the list as a whole. Bug 1452.
1676 Required for reasonable handling of multiple headers_ options when
1677 they may be empty; requires that headers_remove items with embedded
1678 colons must have them doubled (or the list-separator changed).
1680 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1681 view the policy declared in the DMARC record. Currently, $dmarc_status
1682 is a combined value of both the record presence and the result of the
1685 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1687 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1688 dnslookup router and the smtp transport (applying to the forward
1691 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1692 of ldap servers used for a specific lookup. Patch provided by Heiko
1695 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1696 New variable $lookup_dnssec_authenticated for observability.
1698 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1699 Patch submitted by Lars Timman.
1701 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1703 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1704 Requires trusted mode and valid format message id, aborts otherwise.
1705 Patch contributed by Heiko Schlichting.
1707 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1708 certextract with support for various fields. Bug 1358.
1710 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1711 is requested by default, modifiable by smtp transport option
1714 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1715 operate on certificate variables to give certificate fingerprints
1716 Also new ${sha256:cert_variable}.
1718 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1720 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1721 Christian Aistleitner.
1723 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1725 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1726 file. Patch from Wolfgang Breyha.
1728 JH/25 Expand the coverage of the delivery $host and $host_address to
1729 client authenticators run in verify callout. Bug 1476.
1731 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1732 align with daemon_smtp_ports. Bug 72.
1734 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1735 support and error reporting did not work properly.
1737 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1738 and is readable. Patch from Andrew Colin Kissa.
1740 TL/14 Enhance documentation of ${run expansion and how it parses the
1741 commandline after expansion, particularly in the case when an
1742 unquoted variable expansion results in an empty value.
1744 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1746 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1747 by renaming away from it. Addresses GH issue 15, reported by
1750 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1752 TL/15 SECURITY: prevent double expansion in math comparison functions
1753 (can expand unsanitized data). Not remotely exploitable.
1760 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1762 PP/02 Make -n do something, by making it not do something.
1763 When combined with -bP, the name of an option is not output.
1765 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1768 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1769 $sender_host_name and config options to manage this, and basic check
1772 PP/05 DSCP support for outbound connections and control modifier for inbound.
1774 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1775 (Only plugin which currently uses this is kerberos4, which nobody should
1776 be using, but we should make it available and other future plugins might
1777 conceivably use it, even though it would break NAT; stuff *should* be
1778 using channel bindings instead).
1780 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1781 name; added for Sendmail compatibility; requires admin caller.
1782 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1783 just ignore it); requires trusted caller.
1784 Also parse but ignore: -Ac -Am -X<logfile>
1787 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1789 TL/02 Add +smtp_confirmation as a default logging option.
1791 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1792 Patch by Magnus Holmgren from 2007-02-20.
1794 TL/04 Bugzilla 1281 - Spec typo.
1795 Bugzilla 1283 - Spec typo.
1796 Bugzilla 1290 - Spec grammar fixes.
1798 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1800 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1802 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1803 mailing list by by Dmitry Isaikin.
1805 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1807 JH/02 Support "G" suffix to numbers in ${if comparisons.
1809 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1811 NM/01 Bugzilla 1197 - Spec typo
1812 Bugzilla 1196 - Spec examples corrections
1814 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1816 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1817 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1820 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1821 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1823 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1824 "acl {{name}{arg}...}", and optional args on acl condition
1827 JH/05 Permit multiple router/transport headers_add/remove lines.
1829 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1831 JH/07 Avoid using a waiting database for a single-message-only transport.
1832 Performance patch from Paul Fisher. Bugzilla 1262.
1834 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1837 JH/09 Add $headers_added variable, with content from use of ACL modifier
1838 add_header (but not yet added to the message). Bugzilla 199.
1840 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1841 Pulled from Bugzilla 817 by Wolfgang Breyha.
1843 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1845 (nb: this is the same fix as in Exim 4.80.1)
1847 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1850 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1853 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1855 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1856 advertise SMTP AUTH mechanism to us, instead of a generic
1857 protocol violation error. Also, make Exim more robust to bad
1858 data from the Dovecot auth socket.
1860 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1862 When a queue runner is handling a message, Exim first routes the
1863 recipient addresses, during which it prunes them based on the retry
1864 hints database. After that it attempts to deliver the message to
1865 any remaining recipients. It then updates the hints database using
1868 So if a recipient address works intermittently, it can get repeatedly
1869 deferred at routing time. The retry hints record remains fresh so the
1870 address never reaches the final cutoff time.
1872 This is a fairly common occurrence when a user is bumping up against
1873 their storage quota. Exim had some logic in its local delivery code
1874 to deal with this. However it did not apply to per-recipient defers
1875 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1877 This change adds a proper retry rule check during routing so that the
1878 final cutoff time is checked against the message's age. We only do
1879 this check if there is an address retry record and there is not a
1880 domain retry record; this implies that previous attempts to handle
1881 the address had the retry_use_local_parts option turned on. We use
1882 this as an approximation for the destination being like a local
1883 delivery, as in LMTP.
1885 I suspect this new check makes the old local delivery cutoff check
1886 redundant, but I have not verified this so I left the code in place.
1888 TF/02 Correct gecos expansion when From: is a prefix of the username.
1890 Test 0254 submits a message to Exim with the header
1894 When I ran the test suite under the user fanf2, Exim expanded
1895 the header to contain my full name, whereas it should have added
1896 a Resent-Sender: header. It erroneously treats any prefix of the
1897 username as equal to the username.
1899 This change corrects that bug.
1901 GF/01 DCC debug and logging tidyup
1902 Error conditions log to paniclog rather than rejectlog.
1903 Debug lines prefixed by "DCC: " to remove any ambiguity.
1905 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1907 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1908 Bug spotted by Jeremy Harris; was flawed since initial commit.
1909 Would have resulted in OCSP responses post-SNI triggering an Exim
1910 NULL dereference and crash.
1912 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1914 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1915 Bug detection, analysis and fix by Samuel Thibault.
1916 Bugzilla 1331, Debian bug #698092.
1918 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1920 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1921 Server implementation by Todd Lyons, client by JH.
1922 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1923 config variable "prdr_enable" controls whether the server
1924 advertises the facility. If the client requests PRDR a new
1925 acl_data_smtp_prdr ACL is called once for each recipient, after
1926 the body content is received and before the acl_smtp_data ACL.
1927 The client is controlled by both of: a hosts_try_prdr option
1928 on the smtp transport, and the server advertisement.
1929 Default client logging of deliveries and rejections involving
1930 PRDR are flagged with the string "PRDR".
1932 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1933 fclose(). Diagnosis by Todd Lyons.
1935 PP/17 Update configure.default to handle IPv6 localhost better.
1936 Patch by Alain Williams (plus minor tweaks).
1939 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1940 This is now consistent with GnuTLS, and is now documented: the
1941 previous undocumented portable approach to treating the option as
1942 unset was to force an expansion failure. That still works, and
1943 an empty string is now equivalent.
1945 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1946 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1947 not performing validation itself.
1949 PP/20 Added force_command boolean option to pipe transport.
1950 Patch from Nick Koston, of cPanel Inc.
1952 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1955 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1957 PP/21 Fix eximon continuous updating with timestamped log-files.
1958 Broken in a format-string cleanup in 4.80, missed when I repaired the
1959 other false fix of the same issue.
1960 Report and fix from Heiko Schlichting.
1963 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1964 Report from Prashanth Katuri.
1966 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1967 It's SecureTransport, so affects any MacOS clients which use the
1968 system-integrated TLS libraries, including email clients.
1970 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1971 using a MIME ACL for non-SMTP local injection.
1972 Report and assistance in diagnosis by Warren Baker.
1974 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1976 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1978 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1979 last id that failed so it may be referenced in subsequent ACL's.
1981 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1984 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1985 ldap library initialization, allowing self-signed CA's to be
1986 used. Also properly sets require_cert option later in code by
1987 using NULL (global ldap config) instead of ldap handle (per
1988 session). Bug diagnosis and testing by alxgomz.
1990 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1991 the src/util/ subdirectory.
1993 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1994 renamed to Transport Post Delivery Action by Jeremy Harris, as
1997 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1998 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1999 redis_servers = needs to be configured which will be used by the redis
2000 lookup. Patch from Warren Baker, of The Packet Hub.
2002 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2004 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2005 hostname or reverse DNS when processing a host list. Used suggestions
2006 from multiple comments on this bug.
2008 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2010 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2011 Missed a few lines, added it to make the runtest require no keyboard
2014 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2015 contains upper case chars. Make router use caseful_local_part.
2017 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2018 support when GnuTLS has been built with p11-kit.
2024 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2026 This, or similar/improved, will also be change PP/11 of 4.82.
2032 PP/01 Handle short writes when writing local log-files.
2033 In practice, only affects FreeBSD (8 onwards).
2034 Bugzilla 1053, with thanks to Dmitry Isaikin.
2036 NM/01 Bugzilla 949 - Documentation tweak
2038 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2041 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2043 PP/02 Implemented gsasl authenticator.
2045 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2047 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2048 `pkg-config foo` for cflags/libs.
2050 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2051 with rest of GSASL and with heimdal_gssapi.
2053 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2054 `pkg-config foo` for cflags/libs for the TLS implementation.
2056 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2057 properties get this fed in as external SSF. A number of robustness
2058 and debugging improvements to the cyrus_sasl authenticator.
2060 PP/08 cyrus_sasl server now expands the server_realm option.
2062 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2063 Patch by Jeremy Harris.
2065 PP/10 Added dbmjz lookup type.
2067 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2069 PP/12 MAIL args handles TAB as well as SP, for better interop with
2070 non-compliant senders.
2071 Analysis and variant patch by Todd Lyons.
2073 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2074 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2075 Patch from Dirk Mueller <dmueller@suse.com>
2077 PP/13 tls_peerdn now print-escaped for spool files.
2078 Observed some $tls_peerdn in wild which contained \n, which resulted
2079 in spool file corruption.
2081 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2082 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2083 or write after TLS renegotiation, which otherwise led to messages
2086 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2087 as a tracking header (ie: a signed header comes before the signature).
2088 Patch from Wolfgang Breyha.
2090 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2091 comma-sep list; embedded commas doubled.
2093 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2095 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2097 Report and patch from Dmitry Banschikov.
2099 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2100 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2101 failures. We appear to now support TLS1.1+ with Exim.
2103 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2104 lets Exim select keys and certificates based upon TLS SNI from client.
2105 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2106 before an outbound SMTP session. New log_selector, +tls_sni.
2108 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2109 NULL dereference. Report and patch from Alun Jones.
2111 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2112 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2113 Not seeing resolver debug output on NetBSD, but suspect this is a
2114 resolver implementation change.
2116 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2117 Left warnings. Added "eximon gdb" invocation mode.
2119 PP/21 Defaulting "accept_8bitmime" to true, not false.
2121 PP/22 Added -bw for inetd wait mode support.
2123 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2124 locate the relevant includes and libraries. Made this the default.
2126 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2127 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2129 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2130 This may cause build issues on older platforms.
2132 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2133 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2134 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2135 Added SNI support via GnuTLS too.
2136 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2138 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2140 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2141 Applied second patch from Janne, implementing suggestion to default
2142 multiple-strings-in-record handling to match SPF spec.
2144 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2146 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2147 read-only, out of scope).
2148 Patch from Wolfgang Breyha, report from Stuart Northfield.
2150 PP/29 Fix three issues highlighted by clang analyser static analysis.
2151 Only crash-plausible issue would require the Cambridge-specific
2152 iplookup router and a misconfiguration.
2153 Report from Marcin Mirosław.
2155 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2157 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2158 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2159 As part of this, removing so much warning spew let me fix some minor
2160 real issues in debug logging.
2162 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2163 assignment on my part. Fixed.
2165 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2166 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2167 Janne Snabb (who went above and beyond: thank you).
2169 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2170 string otherwise requires a connection and a bunch more work and it's
2171 relatively easy to get wrong. Should also expose TLS library linkage
2174 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2175 64-bit ${eval} (JH/03).
2177 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2178 GNU libc to support some of the 64-bit stuff, should not lead to
2179 conflicts. Defined before os.h is pulled in, so if a given platform
2180 needs to override this, it can.
2182 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2183 protection layer was required, which is not implemented.
2184 Bugzilla 1254, patch from Wolfgang Breyha.
2186 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2187 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2188 tls_dhparam take prime identifiers. Also unbreak combination of
2189 OpenSSL+DH_params+TLSSNI.
2191 PP/39 Disable SSLv2 by default in OpenSSL support.
2197 PP/01 Solaris build fix for Oracle's LDAP libraries.
2198 Bugzilla 1109, patch from Stephen Usher.
2200 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2202 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2205 TF/02 Fix a couple more cases where we did not log the error message
2206 when unlink() failed. See also change 4.74-TF/03.
2208 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2209 lock up or crash if it happened to be inside a call to libc when it
2210 got a SIGUSR1 from exiwhat.
2212 The SIGUSR1 handler appends the current process status to the process
2213 log which is later printed by exiwhat. It used to use the general
2214 purpose logging code to do this, but several functions it calls are
2215 not safe for signals.
2217 The new output code in the SIGUSR1 handler is specific to the process
2218 log, and simple enough that it's easy to inspect for signal safety.
2219 Removing some special cases also simplifies the general logging code.
2220 Removing the spurious timestamps from the process log simplifies
2223 TF/04 Improved ratelimit ACL condition.
2225 The /noupdate option has been deprecated in favour of /readonly which
2226 has clearer semantics. The /leaky, /strict, and /readonly update modes
2227 are mutually exclusive. The update mode is no longer included in the
2228 database key; it just determines when the database is updated. (This
2229 means that when you upgrade Exim will forget old rate measurements.)
2231 Exim now checks that the per_* options are used with an update mode that
2232 makes sense for the current ACL. For example, when Exim is processing a
2233 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2234 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2235 must specify per_mail/readonly. If you omit the update mode it defaults to
2236 /leaky where that makes sense (as before) or /readonly where required.
2238 The /noupdate option is now undocumented but still supported for
2239 backwards compatibility. It is equivalent to /readonly except that in
2240 ACLs where /readonly is required you may specify /leaky/noupdate or
2241 /strict/noupdate which are treated the same as /readonly.
2243 A useful new feature is the /count= option. This is a generalization
2244 of the per_byte option, so that you can measure the throughput of other
2245 aggregate values. For example, the per_byte option is now equivalent
2246 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2248 The per_rcpt option has been generalized using the /count= mechanism
2249 (though it's more complicated than the per_byte equivalence). When it is
2250 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2251 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2252 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2253 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2254 Note that using per_rcpt with a non-readonly update mode in more than
2255 one ACL will cause the recipients to be double-counted. (The per_mail
2256 and per_byte options don't have this problem.)
2258 The handling of very low rates has changed slightly. If the computed rate
2259 is less than the event's count (usually one) then this event is the first
2260 after a long gap. In this case the rate is set to the same as this event's
2261 count, so that the first message of a spam run is counted properly.
2263 The major new feature is a mechanism for counting the rate of unique
2264 events. The new per_addr option counts the number of different
2265 recipients that someone has sent messages to in the last time period. It
2266 behaves like per_rcpt if all the recipient addresses are different, but
2267 duplicate recipient addresses do not increase the measured rate. Like
2268 the /count= option this is a general mechanism, so the per_addr option
2269 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2270 example, measure the rate that a client uses different sender addresses
2271 with the options per_mail/unique=$sender_address. There are further
2272 details in the main documentation.
2274 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2276 TF/06 Removed a few PCRE remnants.
2278 TF/07 Automatically extract Exim's version number from tags in the git
2279 repository when doing development or release builds.
2281 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2282 Bugzilla 879. Patch from Paul Fisher.
2284 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2285 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2288 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2290 PP/05 Variable $av_failed, true if the AV scanner deferred.
2291 Bugzilla 1078. Patch from John Horne.
2293 PP/06 Stop make process more reliably on build failure.
2294 Bugzilla 1087. Patch from Heiko Schlittermann.
2296 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2297 Bugzilla 1089. Patch from Heiko Schlittermann.
2299 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2300 Bugzilla 1131. Patch from Holger Weiß.
2302 PP/09 Handle IPv6 addresses with SPF.
2303 Bugzilla 860. Patch from Wolfgang Breyha.
2305 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2307 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2310 PP/11 match_* no longer expand right-hand-side by default.
2311 New compile-time build option, EXPAND_LISTMATCH_RHS.
2312 New expansion conditions, "inlist", "inlisti".
2314 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2316 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2318 PP/14 fix log_write() format string regression from TF/03.
2319 Bugzilla 1152. Patch from Dmitry Isaikin.
2325 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2327 PP/02 Harmonised TLS library version reporting; only show if debugging.
2328 Layout now matches that introduced for other libraries in 4.74 PP/03.
2330 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2332 PP/04 New "dns_use_edns0" global option.
2334 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2337 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2338 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2340 TK/01 Updated PolarSSL code to 0.14.2.
2341 Bugzilla 1097. Patch from Andreas Metzler.
2343 PP/07 Catch divide-by-zero in ${eval:...}.
2344 Fixes bugzilla 1102.
2346 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2349 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2350 format-string attack -- SECURITY: remote arbitrary code execution.
2352 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2353 time unintentionally subject to list matching rules, letting the header
2354 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2355 arbitrary string expansion). This allowed for information disclosure.
2357 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2358 INT_MIN/-1 -- value coerced to INT_MAX.
2364 NM/01 Workaround for PCRE version dependency in version reporting
2367 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2368 This fixes portability to compilers other than gcc, notably
2369 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2371 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2372 makefiles for portability to HP-UX and POSIX correctness.
2374 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2375 Also via indented variable definition in the Makefile.
2376 (Debugging by Oliver Heesakkers).
2378 PP/02 Restore caching of spamd results with expanded spamd_address.
2379 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2381 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2382 Improves build reliability. Fix from: Frank Elsner
2384 NM/02 Fix wide character breakage in the rfc2047 coding
2385 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2387 NM/03 Allow underscore in dnslist lookups
2388 Fixes bug 1026. Patch from Graeme Fowler
2390 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2391 Code patches from Adam Ciarcinski of NetBSD.
2393 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2396 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2397 is logged, to avoid truncation. Patch from John Horne.
2399 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2400 Patch from Jakob Hirsch.
2402 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2403 SQL string expansion failure details.
2404 Patch from Andrey Oktyabrski.
2406 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2407 Patch from Simon Arlott.
2409 PP/09 New lookups functionality failed to compile on old gcc which rejects
2410 extern declarations in function scope.
2411 Patch from Oliver Fleischmann
2413 PP/10 Use sig_atomic_t for flags set from signal handlers.
2414 Check getgroups() return and improve debugging.
2415 Fixed developed for diagnosis in bug 927 (which turned out to be
2418 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2419 Patch from Mark Zealey.
2421 PP/12 Bugzilla 1056: Improved spamd server selection.
2422 Patch from Mark Zealey.
2424 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2425 Based on patch from Heiko Schlittermann.
2427 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2428 Patch from Uwe Doering, sign-off by Michael Haardt.
2430 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2431 variable declaration deep within a block. Bug and patch from
2434 PP/15 lookups-Makefile IRIX compatibility coercion.
2436 PP/16 Make DISABLE_DKIM build knob functional.
2438 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2439 Patch by Simon Arlott
2441 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2442 variable argument macros. Our copy now differs from upstream.
2448 TF/01 Failure to get a lock on a hints database can have serious
2449 consequences so log it to the panic log.
2451 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2452 controlled using the smtp_confirmation log selector.
2454 TF/03 Include the error message when we fail to unlink a spool file.
2456 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2457 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2458 for maintaining out-of-tree patches for some time.
2460 PP/01 Bugzilla 139: Documentation and portability issues.
2461 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2462 Handle per-OS dynamic-module compilation flags.
2464 PP/02 Let /dev/null have normal permissions.
2465 The 4.73 fixes were a little too stringent and complained about the
2466 permissions on /dev/null. Exempt it from some checks.
2467 Reported by Andreas M. Kirchwitz.
2469 PP/03 Report version information for many libraries, including
2470 Exim version information for dynamically loaded libraries. Created
2471 version.h, now support a version extension string for distributors
2472 who patch heavily. Dynamic module ABI change.
2474 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2475 privilege escalation vulnerability whereby the Exim run-time user
2476 can cause root to append content of the attacker's choosing to
2479 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2482 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2483 If dropping privileges for untrusted macros, we disabled normal logging
2484 on the basis that it would fail; for the Exim run-time user, this is not
2485 the case, and it resulted in successful deliveries going unlogged.
2486 Fixed. Reported by Andreas Metzler.
2492 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2493 only prepend for the Resent-* case. Fixes regression introduced in
2494 Exim 4.70 by NM/22 for Bugzilla 607.
2496 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2497 increasing numbers of administrators be bitten by this.
2499 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2501 PP/03 Bugzilla 994: added openssl_options main configuration option.
2503 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2505 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2507 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2509 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2510 without a peer certificate, leading to a segfault because of an
2511 assumption that peers always have certificates. Be a little more
2512 paranoid. Problem reported by Martin Tscholak.
2514 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2515 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2516 NB: ClamAV planning to remove STREAM in "middle of 2010".
2517 CL also introduces -bmalware, various -d+acl logging additions and
2518 more caution in buffer sizes.
2520 PP/09 Implemented reverse_ip expansion operator.
2522 PP/10 Bugzilla 937: provide a "debug" ACL control.
2524 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2526 PP/12 Bugzilla 973: Implement --version.
2528 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2530 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2532 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2534 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2535 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2536 ignore trailing whitespace.
2538 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2540 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2543 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2544 Notification from Dr Andrew Aitchison.
2546 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2547 ExtendedDetectionInfo response format.
2548 Notification from John Horne.
2550 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2553 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2554 XSL and documented dependency on system catalogs, with examples of how
2557 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2560 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2561 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2562 configuration file which is writeable by the Exim user or group.
2564 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2565 of configuration files to cover files specified with the -C option if
2566 they are going to be used with root privileges, not just the default
2569 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2570 option (effectively making it always true).
2572 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2573 files to be used while preserving root privileges.
2575 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2576 that rogue child processes cannot use them.
2578 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2579 run-time user, instead of root.
2581 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2582 Exim run-time user without dropping privileges.
2584 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2585 result string, instead of calling string_vformat() twice with the same
2588 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2589 for other users. Others should always drop root privileges if they use
2590 -C on the command line, even for a whitelisted configure file.
2592 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2594 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2600 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2601 $data_path, and $header_path variables; fixed documentation bugs and
2604 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2605 exipick to access non-standard spools, including the "frozen" queue
2608 NM/01 Bugzilla 965: Support mysql stored procedures.
2609 Patch from Alain Williams
2611 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2613 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2614 Patch from Andreas Metzler
2616 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2617 Patch from Kirill Miazine
2619 NM/05 Bugzilla 671: Added umask to procmail example.
2621 JJ/03 installed exipick 20100323.0, fixing doc bug
2623 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2624 directory. Notification and patch from Dan Rosenberg.
2626 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2628 TK/02 Improve log output when DKIM signing operation fails.
2630 MH/01 Treat the transport option dkim_domain as a colon separated
2631 list, not as a single string, and sign the message with each element,
2632 omitting multiple occurences of the same signer.
2634 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2635 Bugzilla 985, 986. Patch by Simon Arlott
2637 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2638 Patch by Simon Arlott
2640 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2641 MBX locking. Notification from Dan Rosenberg.
2647 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2649 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2651 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2653 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2655 NM/04 Bugzilla 915: Flag broken perl installation during build.
2661 TK/01 Added patch by Johannes Berg that expands the main option
2662 "spamd_address" if it starts with a dollar sign.
2664 TK/02 Write list of recipients to X-Envelope-Sender header when building
2665 the mbox-format spool file for content scanning (suggested by Jakob
2668 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2669 (http://www.dcc-servers.net/) support via dccifd. Activated by
2670 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2672 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2673 by Mark Daniel Reidel <mr@df.eu>.
2675 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2676 When building exim an external PCRE library is now needed -
2677 PCRE is a system library on the majority of modern systems.
2678 See entry on PCRE_LIBS in EDITME file.
2680 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2681 conversation. Added nologin parameter to request.
2682 Patch contributed by Kirill Miazine.
2684 TF/01 Do not log submission mode rewrites if they do not change the address.
2686 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2688 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2689 log files in place. Contributed by Roberto Lima.
2691 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2693 TF/03 Bugzilla 615: When checking the local_parts router precondition
2694 after a local_part_suffix or local_part_prefix option, Exim now
2695 does not use the address's named list lookup cache, since this
2696 contains cached lookups for the whole local part.
2698 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2699 Robert Millan. Documentation is in experimental-spec.txt.
2701 TF/04 Bugzilla 668: Fix parallel build (make -j).
2703 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2705 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2706 Patch provided by Jan Srzednicki.
2708 TF/05 Leading white space used to be stripped from $spam_report which
2709 wrecked the formatting. Now it is preserved.
2711 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2712 that they are available at delivery time.
2714 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2716 TF/08 TLS error reporting now respects the incoming_interface and
2717 incoming_port log selectors.
2719 TF/09 Produce a more useful error message if an SMTP transport's hosts
2720 setting expands to an empty string.
2722 NM/06 Bugzilla 744: EXPN did not work under TLS.
2723 Patch provided by Phil Pennock.
2725 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2726 Patch provided by Richard Godbee.
2728 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2729 acl_smtp_notquit, added index entry.
2731 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2732 Patch provided by Eugene Bujak.
2734 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2735 accept(). Patch provided by Maxim Dounin.
2737 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2738 Patch provided by Phil Pennock.
2740 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2742 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2743 Patch provided by Brad "anomie" Jorsch.
2745 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2746 Patch provided by Dean Brooks.
2748 TK/05 Add native DKIM support (does not depend on external libraries).
2750 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2751 Patch provided by Graeme Fowler.
2753 NM/16 Bugzilla 851: Documentation example syntax fix.
2755 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2757 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2760 NM/19 Bugzilla 745: TLS version reporting.
2761 Patch provided by Phil Pennock.
2763 NM/20 Bugzilla 167: bool: condition support.
2764 Patch provided by Phil Pennock.
2766 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2767 clients. Patch provided by Phil Pennock.
2769 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2770 Patch provided by Brad "anomie" Jorsch.
2772 NM/23 Bugzilla 687: Fix misparses in eximstats.
2773 Patch provided by Heiko Schlittermann.
2775 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2776 Patch provided by Heiko Schlittermann.
2778 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2779 plus update to original patch.
2781 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2783 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2784 Patch provided by David Brownlee.
2786 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2788 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2790 NM/30 Bugzilla 888: TLS documentation bugfixes.
2792 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2794 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2795 Unlike the original bugzilla I have changed all shell scripts in src tree.
2797 NM/33 Bugzilla 898: Transport filter timeout fix.
2798 Patch by Todd Rinaldo.
2800 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2801 Patch by Serge Demonchaux.
2803 NM/35 Bugzilla 39: Base64 decode bug fixes.
2804 Patch by Jakob Hirsch.
2806 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2808 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2810 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2812 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2818 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2819 ALT-N's libdkim that I have put here:
2820 http://duncanthrax.net/exim-experimental/
2822 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2823 were called 'true' and it seems that C99 defines that as a reserved
2824 keyword to be used with 'bool' variable types. That means you could
2825 not include C99-style headers which use bools without triggering
2826 build errors in sieve.c.
2828 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2829 as mailq or other aliases. Changed the --help handling significantly
2830 to do whats expected. exim_usage() emits usage/help information.
2832 SC/01 Added the -bylocaldomain option to eximstats.
2834 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2836 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2838 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2844 PH/01 Another patch from the Sieve maintainer.
2846 PH/02 When an IPv6 address is converted to a string for single-key lookup
2847 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2848 dots are used instead of colons so that keys in lsearch files need not
2849 contain colons. This was done some time before quoting was made available
2850 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2851 (notated using the quote facility) so as to distinguish them from IPv4
2852 keys. This meant that lookups for IP addresses in host lists did not work
2853 for iplsearch lookups.
2855 This has been fixed by arranging for IPv6 addresses to be expressed with
2856 colons if the lookup type is iplsearch. This is not incompatible, because
2857 previously such lookups could never work.
2859 The situation is now rather anomalous, since one *can* have colons in
2860 ordinary lsearch keys. However, making the change in all cases is
2861 incompatible and would probably break a number of configurations.
2863 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2866 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2867 conversion specification without a maximum field width, thereby enabling
2868 a rogue spamd server to cause a buffer overflow. While nobody in their
2869 right mind would setup Exim to query an untrusted spamd server, an
2870 attacker that gains access to a server running spamd could potentially
2871 exploit this vulnerability to run arbitrary code as the Exim user.
2873 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2874 $primary_hostname instead of what libspf2 thinks the hosts name is.
2876 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2877 a directory entry by the name of the lookup key. Previously, if a
2878 symlink pointed to a non-existing file or a file in a directory that
2879 Exim lacked permissions to read, a lookup for a key matching that
2880 symlink would fail. Now it is enough that a matching directory entry
2881 exists, symlink or not. (Bugzilla 503.)
2883 PH/03 The body_linecount and body_zerocount variables are now exported in the
2886 PH/04 Added the $dnslist_matched variable.
2888 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2889 This means they are set thereafter only if the connection becomes
2892 PH/06 Added the client_condition to authenticators so that some can be skipped
2893 by clients under certain conditions.
2895 PH/07 The error message for a badly-placed control=no_multiline_responses left
2896 "_responses" off the end of the name.
2898 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2900 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2901 (without spaces) instead of just copying the configuration text.
2903 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2905 PH/11 Added $max_received_linelength.
2907 PH/12 Added +ignore_defer and +include_defer to host lists.
2909 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2910 way in which PCRE > 7.0 is built.
2912 PH/14 Implemented queue_only_load_latch.
2914 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2915 MAIL command. The effect was to mangle the value on 64-bit systems.
2917 PH/16 Another patch from the Sieve maintainer.
2919 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2921 PH/18 If a system quota error occurred while trying to create the file for
2922 a maildir delivery, the message "Mailbox is full" was not appended to the
2923 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2924 only to a quota excession during the actual writing of the file.
2926 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2927 characters?) which causes problems in log lines. The DN values are now
2928 passed through string_printing() before being added to log lines.
2930 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2931 and InterBase are left for another time.)
2933 PH/21 Added message_body_newlines option.
2935 PH/22 Guard against possible overflow in moan_check_errorcopy().
2937 PH/23 POSIX allows open() to be a macro; guard against that.
2939 PH/24 If the recipient of an error message contained an @ in the local part
2940 (suitably quoted, of course), incorrect values were put in $domain and
2941 $local_part during the evaluation of errors_copy.
2947 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2948 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2951 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2952 issue a MAIL command.
2954 PH/02 In an ACL statement such as
2956 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2958 if a client was not listed at all, or was listed with a value other than
2959 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2960 the condition was not true (as it should be), so access was not denied.
2961 The bug was that the ! inversion was incorrectly passed on to the second
2962 item. This has been fixed.
2964 PH/03 Added additional dnslists conditions == and =& which are different from
2965 = and & when the dns lookup returns more than one IP address.
2967 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2968 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2970 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2971 FSYNC, which compiles an option called disable_fsync that allows for
2972 bypassing fsync(). The documentation is heavily laced with warnings.
2974 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2976 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2977 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2978 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2979 including adding "make clean"; (3) Added -fPIC when compiling the test
2980 dynamically loaded module, to get rid of a warning.
2982 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2983 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2984 = 0s. The bug is otherwise harmless.
2986 PH/07 There was a bug in the dovecot authenticator such that the value of
2987 $auth1 could be overwritten, and so not correctly preserved, after a
2988 successful authentication. This usually meant that the value preserved by
2989 the server_setid option was incorrect.
2991 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2993 PH/09 Installed PCRE release 7.0.
2995 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2996 run for batched SMTP input. It is now run at the start of every message
2997 in the batch. While fixing this I discovered that the process information
2998 (output by running exiwhat) was not always getting set for -bs and -bS
2999 input. This is fixed, and it now also says "batched" for BSMTP.
3001 PH/11 Added control=no_pipelining.
3003 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3004 patch, slightly modified), and move the expansion of helo_data till after
3005 the connection is made in the smtp transport (so it can use these
3008 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3010 PH/14 Added log_selector = +pid.
3012 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3014 PH/16 Add ${if forany and ${if forall.
3016 PH/17 Added dsn_from option to vary the From: line in DSNs.
3018 PH/18 Flush SMTP output before performing a callout, unless control =
3019 no_callout_flush is set.
3021 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3022 was true (the default) a successful delivery failed to delete the retry
3023 item, thus causing premature timeout of the address. The bug is now
3026 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3028 PH/21 Long custom messages for fakedefer and fakereject are now split up
3029 into multiline responses in the same way that messages for "deny" and
3030 other ACL rejections are.
3032 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3033 with slight modification.
3035 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3036 draft, changing the syntax and factoring some duplicate code".
3038 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3039 for deliveries of the second and subsequent messages over the same SMTP
3042 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3043 ${reduce, with only minor "tidies".
3045 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3047 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3048 expansion side effects.
3050 PH/27 When a message times out after an over-quota error from an Exim-imposed
3051 quota, the bounce message says "mailbox is full". This message was not
3052 being given when it was a system quota that was exceeded. It now should
3055 MH/03 Made $recipients available in local_scan(). local_scan() already has
3056 better access to the recipient list through recipients_list[], but
3057 $recipients can be useful in postmaster-provided expansion strings.
3059 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3060 in the case of a MAIL command with additional options following the
3061 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3062 were accidentally chopped off.
3064 PH/29 SMTP synchronization checks are implemented when a command is read -
3065 there is a check that no more input is waiting when there shouldn't be
3066 any. However, for some commands, a delay in an ACL can mean that it is
3067 some time before the response is written. In this time, more input might
3068 arrive, invalidly. So now there are extra checks after an ACL has run for
3069 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3070 pipelining has not been advertised.
3072 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3074 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3075 returned in the SMTP response when smtp_return_error_details was false.
3076 This has been fixed.
3078 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3079 instead of the C I/O that was originally supplied, because problems were
3080 reported on Solaris.
3082 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3083 Exim which did not show up earlier: it was assuming that a call to
3084 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3085 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3086 was a macro that became an assignment, so it seemed to work. This has
3087 changed to a proper function call with a void return, hence the compile
3088 error. Exim's code has been fixed.
3090 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3093 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3095 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3097 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3098 because it turns out to be too big or there is a timeout) there is no
3099 "Completed" line in the log. When some messages of this type were
3100 selected by exigrep, they were listed as "not completed". Others were
3101 picked up by some special patterns. I have improved the selection
3102 criteria to be more general.
3104 PH/38 The host_find_failed option in the manualroute router can now be set
3105 to "ignore", to completely ignore a host whose IP address cannot be
3106 found. If all hosts are ignored, the behaviour is controlled by the new
3107 host_all_ignored option.
3109 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3110 homing or because of multiple MX records with /mx) generated more than
3111 one IP address, and the following item turned out to be the local host,
3112 all the secondary addresses of the first item were incorrectly removed
3113 from the list, along with the local host and any following hosts (which
3114 is what is supposed to happen).
3116 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3117 whoever called Exim into the -H file. In the case of the daemon it was
3118 behaving confusingly. When first started, it used values for whoever
3119 started the daemon, but after a SIGHUP it used the Exim user (because it
3120 calls itself on a restart). I have changed the code so that it now always
3123 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3124 message are rejected with the same error (e.g. no authentication or bad
3125 sender address), and a DATA command is nevertheless sent (as can happen
3126 with PIPELINING or a stupid MUA), the error message that was given to the
3127 RCPT commands is included in the rejection of the DATA command. This is
3128 intended to be helpful for MUAs that show only the final error to their
3131 PH/42 Another patch from the Sieve maintainer.
3133 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3134 Eximstats - Fixed some broken HTML links and added missing column headers
3136 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3137 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3139 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3141 PH/43 Yet another patch from the Sieve maintainer.
3143 PH/44 I found a way to check for a TCP/IP connection going away before sending
3144 the response to the final '.' that terminates a message, but only in the
3145 case where the client has not sent further data following the '.'
3146 (unfortunately, this is allowed). However, in many cases there won't be
3147 any further data because there won't be any more messages to send. A call
3148 to select() can be used: if it shows that the input is "ready", there is
3149 either input waiting, or the socket has been closed. An attempt to read
3150 the next input character can distinguish the two cases. Previously, Exim
3151 would have sent an OK response which the client would never have see.
3152 This could lead to message repetition. This fix should cure that, at
3153 least in a lot of common cases.
3155 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3156 advertised in response to EHLO.
3162 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3163 fixed by 4.65/MH/01 (is this a record?) are fixed:
3165 (i) An empty string was always treated as zero by the numeric comparison
3166 operators. This behaviour has been restored.
3168 (ii) It is documented that the numeric comparison operators always treat
3169 their arguments as decimal numbers. This was broken in that numbers
3170 starting with 0 were being interpreted as octal.
3172 While fixing these problems I realized that there was another issue that
3173 hadn't been noticed. Values of message_size_limit (both the global option
3174 and the transport option) were treated as octal if they started with 0.
3175 The documentation was vague. These values are now always treated as
3176 decimal, and I will make that clear in the documentation.
3182 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3183 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3186 MH/01 Don't check that the operands of numeric comparison operators are
3187 integers when their expansion is in "skipping" mode (fixes bug
3188 introduced by 4.64-PH/07).
3190 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3191 child addresses, Exim now panics and dies. Previously, because the count
3192 is held in a short int, deliveries were likely to be lost. As such a
3193 large number of recipients for a single message is ridiculous
3194 (performance will be very, very poor), I have chosen to impose a limit
3195 rather than extend the field.
3201 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3202 leftover -K file (the existence of which was triggered by #402).
3203 While we were at it, introduced process PID as part of the -K
3204 filename. This should rule out race conditions when creating
3207 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3208 processing considerably. Previous code took too long for large mails,
3209 triggering a timeout which in turn triggers #401.
3211 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3212 in the DK code in transports.c. sendfile() is not really portable,
3213 hence the _LINUX specificness.
3215 TF/01 In the add_headers option to the mail command in an Exim filter,
3216 there was a bug that Exim would claim a syntax error in any
3217 header after the first one which had an odd number of characters
3220 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3221 callout verification, Exim cached a "reject" for the entire domain. This
3222 is correct for most verifications, but it is not correct for a recipient
3223 verification with use_sender or use_postmaster set, because in that case
3224 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3225 case of MAIL FROM:<> rejection from other early rejections (e.g.
3226 rejection of HELO). When verifying a recipient using a non-null MAIL
3227 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3228 Whatever the result of the callout, the value of the domain cache is
3229 left unchanged (for any other kind of callout, getting as far as trying
3230 RCPT means that the domain itself is ok).
3232 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3235 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3236 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3237 session, the daemon ignores ECONNECT errors and logs others; it now
3238 ignores EPIPE as well.
3240 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3241 (quoted-printable decoding).
3243 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3244 later the small subsequent patch to fix an introduced bug.
3246 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3248 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3250 PH/08 An error is now given if message_size_limit is specified negative.
3252 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3253 to be given (somewhat) arbitrary names.
3255 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3258 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3259 miscellaneous code fixes
3261 PH/10 Added the log_reject_target ACL modifier to specify where to log
3264 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3265 hostname. This is wrong, because it relates to the incoming message (and
3266 probably the interface on which it is arriving) and not to the outgoing
3267 callout (which could be using a different interface). This has been
3268 changed to use the value of the helo_data option from the smtp transport
3269 instead - this is what is used when a message is actually being sent. If
3270 there is no remote transport (possible with a router that sets up host
3271 addresses), $smtp_active_hostname is used.
3273 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3274 tweaks were necessary in order to get it to work (see also 21 below):
3275 (a) The code assumed that strncpy() returns a negative number on buffer
3276 overflow, which isn't the case. Replaced with Exim's string_format()
3278 (b) There were several signed/unsigned issues. I just did the minimum
3279 hacking in of casts. There is scope for a larger refactoring.
3280 (c) The code used strcasecmp() which is not a standard C function.
3281 Replaced with Exim's strcmpic() function.
3282 (d) The code set only $1; it now sets $auth1 as well.
3283 (e) A simple test gave the error "authentication client didn't specify
3284 service in request". It would seem that Dovecot has changed its
3285 interface. Fortunately there's a specification; I followed it and
3286 changed what the client sends and it appears to be working now.
3288 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3291 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3292 address A is aliased to B and C, where B exists and C does not. Without
3293 -v the output is "A verified" because verification stops after a
3294 successful redirection if more than one address is generated. However,
3295 with -v the child addresses are also verified. Exim was outputting "A
3296 failed to verify" and then showing the successful verification for C,
3297 with its parentage. It now outputs "B failed to verify", showing B's
3298 parentage before showing the successful verification of C.
3300 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3301 look up a TXT record in a specific list after matching in a combined
3304 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3305 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3306 they consult the DNS. I had assumed they would set it the way they
3307 wanted; and indeed my experiments on Linux seem to show that in some
3308 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3309 To be on the safe side, however, I have now made the interface to
3310 host_find_byname() similar to host_find_bydns(), with an argument
3311 containing the DNS resolver options. The host_find_byname() function now
3312 sets these options at its start, just as host_find_bydns() does. The smtp
3313 transport options dns_qualify_single and dns_search_parents are passed to
3314 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3315 of host_find_byname() use the default settings of RES_DEFNAMES
3316 (qualify_single) but not RES_DNSRCH (search_parents).
3318 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3319 spool_read_header() do less string testing, by means of a preliminary
3320 switch on the second character of optional "-foo" lines. (This is
3321 overdue, caused by the large number of possibilities that now exist.
3322 Originally there were few.) While I was there, I also converted the
3323 str(n)cmp tests so they don't re-test the leading "-" and the first
3324 character, in the hope this might squeeze out yet more improvement.
3326 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3327 flag allowing group syntax was set by the header_syntax check but not
3328 turned off, possible causing trouble later; (2) The flag was not being
3329 set at all for the header_verify test, causing "group"-style headers to
3330 be rejected. I have now set it in this case, and also caused header_
3331 verify to ignore an empty address taken from a group. While doing this, I
3332 came across some other cases where the code for allowing group syntax
3333 while scanning a header line wasn't quite right (mostly, not resetting
3334 the flag correctly in the right place). These bugs could have caused
3335 trouble for malformed header lines. I hope it is now all correct.
3337 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3338 with the "reply" argument non-NULL. The code, however (which originally
3339 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3340 but it didn't always do it. This confused somebody who was copying the
3341 code for some other use. I have removed all the tests.
3343 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3344 feature that was used to support insecure browsers during the U.S. crypto
3345 embargo. It requires special client support, and Exim is probably the
3346 only MTA that supported it -- and would never use it because real RSA is
3347 always available. This code has been removed, because it had the bad
3348 effect of slowing Exim down by computing (never used) parameters for the
3349 RSA_EXPORT functionality.
3351 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3352 authenticator to fail if there's a tab character in the incoming data
3353 (there should never be unless someone is messing about, as it's supposed
3354 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3355 if the connection is using TLS or if the remote IP is the same as the
3356 local IP, and the "valid-client-cert option" if a client certificate has
3359 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3360 authenticators. This can be used for authorization after authentication
3361 succeeds. (In the case of plaintext, it servers for both authentication
3364 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3365 if any retry times were supplied.
3367 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3368 connection, where there is no client IP address to check. In this
3369 situation, the verify now always succeeds.
3371 PH/25 Applied John Jetmore's -Mset patch.
3373 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3375 PH/27 In a string expansion for a processed (not raw) header when multiple
3376 headers of the same name were present, leading whitespace was being
3377 removed from all of them, but trailing whitespace was being removed only
3378 from the last one. Now trailing whitespace is removed from each header
3379 before concatenation. Completely empty headers in a concatenation (as
3380 before) are ignored.
3382 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3383 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3385 PH/29 [Removed. This was a change that I later backed out, and forgot to
3386 correct the ChangeLog entry (that I had efficiently created) before
3387 committing the later change.]
3389 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3390 address errors (4xx response to RCPT) over the same connection as other
3391 messages routed to the same hosts. Such deliveries are always "forced",
3392 so retry times are not inspected. This resulted in far too many retries
3393 for the affected addresses. The effect occurred only when there were more
3394 hosts than the hosts_max_try setting in the smtp transport when it had
3395 the 4xx errors. Those hosts that it had tried were not added to the list
3396 of hosts for which the message was waiting, so if all were tried, there
3397 was no problem. Two fixes have been applied:
3399 (i) If there are any address or message errors in an SMTP delivery, none
3400 of the hosts (tried or untried) are now added to the list of hosts
3401 for which the message is waiting, so the message should not be a
3402 candidate for sending over the same connection that was used for a
3403 successful delivery of some other message. This seems entirely
3404 reasonable: after all the message is NOT "waiting for some host".
3405 This is so "obvious" that I'm not sure why it wasn't done
3406 previously. Hope I haven't missed anything, but it can't do any
3407 harm, as the worst effect is to miss an optimization.
3409 (ii) If, despite (i), such a delivery is accidentally attempted, the
3410 routing retry time is respected, so at least it doesn't keep
3411 hammering the server.
3413 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3414 in ${readsocket because some servers need this prod.
3416 PH/32 Added some extra debug output when updating a wait-xxx database.
3418 PH/33 The hint "could be header name not terminated by colon", which has been
3419 given for certain expansion errors for a long time, was not being given
3420 for the ${if def:h_colon_omitted{... case.
3422 PH/34 The spec says: "With one important exception, whenever a domain list is
3423 being scanned, $domain contains the subject domain." There was at least
3424 one case where this was not true.
3426 PH/35 The error "getsockname() failed: connection reset by peer" was being
3427 written to the panic log as well as the main log, but it isn't really
3428 panic-worthy as it just means the connection died rather early on. I have
3429 removed the panic log writing for the ECONNRESET error when getsockname()
3432 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3433 runs only) independently of the message's sender address. This meant
3434 that, if the 4xx error was in fact related to the sender, a different
3435 message to the same recipient with a different sender could confuse
3436 things. In particular, this can happen when sending to a greylisting
3437 server, but other circumstances could also provoke similar problems.
3438 I have changed the default so that the retry time for these errors is now
3439 based a combination of the sender and recipient addresses. This change
3440 can be overridden by setting address_retry_include_sender=false in the
3443 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3444 remote server are returned as part of bounce messages. This was not
3445 happening for LMTP over a pipe (the lmtp transport), but now it is the
3446 same for both kinds of LMTP.
3448 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3449 in header lines that were in fact CNAMEs. This is no longer the case.
3451 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3452 and queue runs started by the daemon processed all messages. This has
3453 been fixed so that -R and -S can now usefully be given with -q<time>.
3455 PH/40 Import PCRE release 6.7 (fixes some bugs).
3457 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3459 PH/42 Give an error if -q is specified more than once.
3461 PH/43 Renamed the variables $interface_address and $interface_port as
3462 $received_ip_address and $received_port, to make it clear that these
3463 values apply to message reception, and not to the outgoing interface when
3464 a message is delivered. (The old names remain recognized, of course.)
3466 PH/44 There was no timeout on the connect() call when using a Unix domain
3467 socket in the ${readsocket expansion. There now is.
3469 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3470 be meaningful with "accept".
3472 SC/01 Eximstats V1.43
3473 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3475 SC/02 Eximstats V1.44
3476 Use a glob alias rather than an array ref in the generated
3477 parser. This improves both readability and performance.
3479 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3480 Collect SpamAssassin and rejection statistics.
3481 Don't display local sender or destination tables unless
3482 there is data to show.
3483 Added average volumes into the top table text output.
3485 SC/04 Eximstats V1.46
3486 Collect data on the number of addresses (recipients)
3487 as well as the number of messages.
3489 SC/05 Eximstats V1.47
3490 Added 'Message too big' to the list of mail rejection
3491 reasons (thanks to Marco Gaiarin).
3493 SC/06 Eximstats V1.48
3494 Mainlog lines which have GMT offsets and are too short to
3495 have a flag are now skipped.
3497 SC/07 Eximstats V1.49 (Alain Williams)
3498 Added the -emptyok flag.
3500 SC/08 Eximstats V1.50
3501 Fixes for obtaining the IP address from reject messages.
3503 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3504 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3505 whitespace changes from 4.64-PH/27
3507 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3510 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3511 are found, allow negative numbers in numeric criteria)
3513 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3515 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3516 to match changes made in 4.64-PH/43
3518 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3520 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3521 "domain" is actually an IP address, and force a failure. This locks out
3522 those revolvers/nameservers that support "A-for-A" lookups, in
3523 contravention of the specifications.
3525 PH/48 When a host name was looked up from an IP address, and the subsequent
3526 forward lookup of the name timed out, the host name was left in
3527 $sender_host_name, contrary to the specification.
3529 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3530 restricted to single-key lookups, Exim was not diagnosing an error if
3531 * or *@ was used with a query-style lookup.
3533 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3535 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3536 long ago, but noone interested enough thought of it. Let's just say that
3537 the "1.1" means that there are some new functions that weren't there at
3538 some point in the past.
3540 PH/51 Error processing for expansion failure of helo_data from an smtp
3541 transport during callout processing was broken.
3543 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3544 tested/used via the -bh/-bhc/-bs options.
3546 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3547 bug, fixed in subsequent PCRE releases).
3549 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3550 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3552 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3558 SC/01 Use a glob alias rather than an array ref in eximstats generated
3559 parser. This improves both readability and performance.
3561 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3562 Don't display local sender or destination tables in eximstats unless
3563 there is data to show.
3564 Added average volumes into the eximstats top table text output.
3566 SC/03 Collect data on the number of addresses (recipients) as well
3567 as the number of messages in eximstats.
3569 TF/01 Correct an error in the documentation for the redirect router. Exim
3570 does (usually) call initgroups() when daemonizing.
3572 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3573 with consistent privilege compared to when running as a daemon.
3575 TF/03 Note in the spec that $authenticated_id is not set for local
3576 submissions from trusted users.
3578 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3579 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3581 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3582 by adding some example configuration directives to the default
3583 configuration file. A little bit of work is required to uncomment the
3584 directives and define how usernames and passwords are checked, but
3585 there is now a framework to start from.
3587 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3588 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3589 without this. I don't know how relevant this is to other LDAP libraries.
3591 PH/02 Add the verb name to the "unknown ACL verb" error.
3593 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3595 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3597 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3598 directory not expanded when it should be if an expanded home directory
3599 was set for the address (which is overridden by the transport).
3601 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3604 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3605 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3606 because it is too late at that time, and has no effect.
3608 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3609 security issue with \' (bugzilla #107). I could not use the
3610 PQescapeStringConn() function, because it needs a PGconn value as one of
3613 PH/08 When testing addresses using -bt, indicate those final addresses that
3614 are duplicates that would not cause an additional delivery. At least one
3615 person was confused, thinking that -bt output corresponded to deliveries.
3616 (Suppressing duplicates isn't a good idea as you lose the information
3617 about possibly different redirections that led to the duplicates.)
3619 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3620 systems where poll() doesn't work, in particular OS X.
3622 PH/10 Added more information to debugging output for retry time not reached.
3624 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3625 operations in malware.c.
3627 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3630 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3631 syslog_duplication set false, log lines that would normally be written
3632 both the the main log and to the reject log were not written to syslog at
3635 PH/14 In the default configuration, change the use of "message" in ACL warn
3636 statements to "add_header".
3638 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3639 not followed by a command (e.g. "seen endif").
3641 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3642 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3645 PH/17 Added extra conditions to the default value of delay_warning_condition
3649 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3650 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3651 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3654 The Auto-Submitted: and various List- headers are standardised, whereas I
3655 don't think Precedence: ever was.
3657 PH/18 Refactored debugging code in route_finduser() to show more information,
3658 in particular, the error code if getpwnam() issues one.
3660 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3661 This is apparently needed in addition to the PH/07 change above to avoid
3662 any possible encoding problems.
3664 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3665 but not after initializing Perl.
3667 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3668 output them only if debugging. By default they are written stderr,
3669 apparently, which is not desirable.
3671 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3674 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3677 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3679 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3680 authenticated or an ident call has been made. Suppress the default
3681 values for $authenticated_id and $authenticated_sender (but permit -oMai
3682 and -oMas) when testing with -bh.
3684 PH/24 Re-jigged the order of the tests in the default configuration so that the
3685 tests for valid domains and recipients precede the DNS black list and CSA
3686 tests, on the grounds that those ones are more expensive.
3688 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3689 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3690 This bug exists in every version of Exim that I still have, right back to
3693 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3694 However, an attempt to turn on case-sensitivity in a regex key by
3695 including (?-i) didn't work because the subject string was already
3696 lowercased, and the effects were non-intuitive. It turns out that a
3697 one-line patch can be used to allow (?-i) to work as expected.
3703 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3704 other effects) broke the use of negated acl sub-conditions.
3706 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3709 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3710 "Deny" causes Exim to reject the incoming connection with a 554 error.
3711 Unfortunately, if there is a major crisis, such as a disk failure,
3712 tcp-wrappers gives "deny", whereas what one would like would be some
3713 kind of temporary error. A kludge has been added to help with this.
3714 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3715 554 error is used if errno is still zero or contains ENOENT (which occurs
3716 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3719 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3721 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3722 errors. Otherwise a message that provokes a temporary error (when other
3723 messages do not) can cause a whole host to time out.
3725 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3726 addresses were routed directly to files or pipes from a redirect router.
3727 File deliveries just didn't batch; pipe deliveries might have suffered
3730 PH/06 A failure to get a lock for a hints database would erroneously always say
3731 "Failed to get write lock", even when it was really a read lock.
3733 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3734 of 0600. This has been changed to use the value of the lockfile_mode
3735 option (which defaults to 0600).
3737 PH/08 Applied small patch from the Sieve maintainer.
3739 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3740 folder from quota calculations, a direct delivery into this folder messed
3741 up the contents of the maildirsize file. This was because the regex was
3742 used only to exclude .Trash (or whatever) when the size of the mailbox
3743 was calculated. There was no check that a delivery was happening into an
3744 excluded directory. This bug has been fixed by ignoring all quota
3745 processing for deliveries into excluded directories.
3747 PH/10 Added the maildirfolder_create_regex option to appendfile.
3753 PH/01 The code for finding all the local interface addresses on a FreeBSD
3754 system running IPv6 was broken. This may well have applied to all BSD
3755 systems, as well as to others that have similar system calls. The broken
3756 code found IPv4 interfaces correctly, but gave incorrect values for the
3757 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3758 that it would not match correctly against @[] and not recognize the IPv6
3761 PH/02 The ipliteral router was not recognizing addresses of the form user@
3762 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3764 PH/03 Added disable_ipv6.
3766 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3767 decoded form, because it is most often used to construct To: headers
3768 lines in autoreplies, and the decoded form may well be syntactically
3769 invalid. However, $reply_address has leading white space removed, and all
3770 newlines turned into spaces so that the autoreply transport does not
3773 PH/05 If group was specified without a user on a router, and no group or user
3774 was specified on a transport, the group from the router was ignored.
3776 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3777 for visible compile-time settings that can be used to change these
3778 numbers, for those that want even more. Backwards compatibility with old
3779 spool files has been maintained. However, going back to a previous Exim
3780 release will lost any variables that are in spool files.
3782 PH/07 Two small changes when running in the test harness: increase delay when
3783 passing a TCP/IP connection to a new process, in case the original
3784 process has to generate a bounce, and remove special handling of
3785 127.0.0.2 (sic), which is no longer necessary.
3787 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3788 be the same on different OS.
3790 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3793 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3794 whether --show-vars was specified or not
3796 JJ/02 exipick: Added support for new ACL variable spool format introduced
3799 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3800 syntactically invalid From: or Reply-to: line, and a filter used this to
3801 generate an autoreply, and therefore failed to obtain an address for the
3802 autoreply, Exim could try to deliver to a non-existent relative file
3803 name, causing unrelated and misleading errors. What now happens is that
3804 it logs this as a hard delivery error, but does not attempt to create a
3807 PH/11 The exinext utility has a -C option for testing purposes, but although
3808 the given file was scanned by exinext itself; it wasn't being passed on
3809 when Exim was called.
3811 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3812 an end-of-file indication when reading a command response.
3814 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3815 compiled. In many other places in Exim, IPv6 addresses are always
3816 recognized, so I have changed this. It also means that IPv4 domain
3817 literals of the form [IPV4:n.n.n.n] are now always recognized.
3819 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3820 used if the router is not running as root, for example, when verifying at
3821 ACL time, or when using -bh. The debugging output from this situation was
3822 non-existent - all you got was a failure to exec. I have made two
3825 (a) Failures to set uid/gid, the current directory, or a process leader
3826 in a subprocess such as that created by queryprogram now generate
3827 suitable debugging output when -d is set.
3829 (b) The queryprogram router detects when it is not running as root,
3830 outputs suitable debugging information if -d is set, and then runs
3831 the subprocess without attempting to change uid/gid.
3833 PH/15 Minor change to Makefile for building test_host (undocumented testing
3836 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3837 additional section of a DNS packet that returns MX or SRV records.
3838 Instead, it always explicitly searches for A/AAAA records. This avoids
3839 major problems that occur when a DNS server includes only records of one
3840 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3841 fixed another bug: if SRV records were looked up and the corresponding
3842 address records were *not* found in the additional section, the port
3843 values from the SRV records were lost.
3845 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3846 using the correct key (the original address) when searching the retry
3847 rules in order to find which one to use for generating the retry hint.
3849 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3850 adding the default one. Similarly, if it contains a Reply-To: header, the
3851 errors_reply_to option, if set, is not used.
3853 PH/19 When calculating a retry time, Exim used to measure the "time since
3854 failure" by looking at the "first failed" field in the retry record. Now
3855 it does not use this if it is later than than the arrival time of the
3856 message. Instead it uses the arrival time. This makes for better
3857 behaviour in cases where some deliveries succeed, thus re-setting the
3858 "first failed" field. An example is a quota failure for a huge message
3859 when small messages continue to be delivered. Without this change, the
3860 "time since failure" will always be short, possible causing more frequent
3861 delivery attempts for the huge message than are intended.
3862 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3864 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3865 $1, $2, $3) because the numerical variables can be reset during some
3866 expansion items (e.g. "match"), thereby losing the authentication data.
3868 PH/21 Make -bV show the size of off_t variables so that the test suite can
3869 decide whether to run tests for quotas > 2G.
3871 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3872 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3873 filecount value is greater than 2G or if a quota value is greater than 2G
3874 on a system where the size of off_t is not greater than 4, a panic error
3877 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3878 never match. The debug and -bh output now contains an explicit error
3879 message indicating a malformed IPv4 address or mask.
3881 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3882 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3883 PH/23 above applies.
3885 PH/25 Do not write to syslog when running in the test harness. The only
3886 occasion when this arises is a failure to open the main or panic logs
3887 (for which there is an explicit test).
3889 PH/26 Added the /no_tell option to "control=freeze".
3891 PH/27 If a host name lookup failed very early in a connection, for example, if
3892 the IP address matched host_lookup and the reverse lookup yielded a name
3893 that did not have a forward lookup, an error message of the form "no IP
3894 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3895 could be logged. Now it outputs the IP address instead of "NULL".
3897 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3898 allows the sender and the authenticated sender to be set when
3899 submitting a message from within Exim. Since child_open_exim() is
3900 documented for local_scan(), the new function should be too.
3902 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3903 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3904 results in an empty string is now treated as unset.
3906 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3908 PH/31 Added sender_verify_fail logging option.
3910 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3911 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3912 tidied the source and removed it altogether.
3914 PH/33 When a queue run was abandoned because the load average was too high, a
3915 log line was always written; now it is written only if the queue_run log
3916 selector is set. In addition, the log line for abandonment now contains
3917 information about the queue run such as the pid. This is always present
3918 in "start" and "stop" lines but was omitted from the "abandon" line.
3920 PH/34 Omit spaces between a header name and the colon in the error message that
3921 is given when verify = headers_syntax fails (if there are lots of them,
3922 the message gets confusing).
3924 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3925 names, as there are now some PTR records that contain slashes. This check
3926 is only to protect against broken name servers that fall over on strange
3927 characters, so the fact that it applies to all lookups doesn't matter.
3929 PH/36 Now that the new test suite is complete, we can remove some of the
3930 special code in Exim that was needed for the old test suite. For example,
3931 sorting DNS records because real resolvers return them in an arbitrary
3932 order. The new test suite's fake resolver always returns records in the
3935 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3936 bounces) except when queue_only is set, to avoid logging races between
3937 the different processes.
3939 PH/38 Panic-die if .include specifies a non-absolute path.
3941 PH/39 A tweak to the "H" retry rule from its user.
3943 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3944 a label. They prevented compilation on older perls.
3946 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3947 a warning to be raised on newish perls.
3949 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3950 on queue. Changes to match documented behaviour of showing count of
3951 messages matching specified criteria.
3953 PH/40 Changed the default ident timeout from 30s to 5s.
3955 PH/41 Added support for the use of login_cap features, on those BSD systems
3956 that have them, for controlling the resources used by pipe deliveries.
3958 PH/42 The content-scanning code uses fopen() to create files in which to put
3959 message data. Previously it was not paying any attention to the mode of
3960 the files. Exim runs with umask(0) because the rest of the code creates
3961 files with open(), and sets the required mode explicitly. Thus, these
3962 files were ending up world-writeable. This was not a big issue, because,
3963 being within the spool directory, they were not world-accessible. I have
3964 created a function called modefopen, which takes an additional mode
3965 argument. It sets umask(777), creates the file, chmods it to the required
3966 mode, then resets the umask. All the relevant calls to fopen() in the
3967 content scanning code have been changed to use this function.
3969 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3970 to 24 hours. This avoids potential overflow problems when processing G
3971 and H retry rules. I suspect nobody ever tinkers with this value.
3973 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3975 PH/45 When the plaintext authenticator is running as a client, the server's
3976 challenges are checked to ensure they are valid base64 strings. By
3977 default, the authentication attempt is cancelled if an invalid string is
3978 received. Setting client_ignore_invalid_base64 true ignores these errors.
3979 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3980 they are received. Thus, the responses can be made to depend on the
3981 challenges. If an invalid string is ignored, an empty string is placed in
3984 PH/46 Messages that are created by the autoreply transport now contains a
3985 References: header, in accordance with RFCs 2822 and 3834.
3987 PH/47 Added authenticated_sender_force to the smtp transport.
3989 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3991 PH/49 Installed latest patch from the Sieve maintainer.
3993 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3994 was also set, the appendfile transport was unnecessarily scanning a
3995 directory of message files (e.g. for maildir delivery) to find the count
3996 of files (along with the size), even though it did not need this
3997 information. It now does the scan only if it needs to find either the
3998 size of the count of files.
4000 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4002 PH/52 Two bugs concerned with error handling when the smtp transport is
4005 (i) Exim was not creating retry information for temporary errors given
4006 for individual recipients after the DATA command when the smtp transport
4007 was used in LMTP mode. This meant that they could be retried too
4008 frequently, and not timed out correctly.
4010 (ii) Exim was setting the flag that allows error details to be returned
4011 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4012 recipients that were returned after the DATA command.
4014 PH/53 This is related to PH/52, but is more general: for any failing address,
4015 when detailed error information was permitted to be returned to the
4016 sender, but the error was temporary, then after the final timeout, only
4017 "retry timeout exceeded" was returned. Now it returns the full error as
4018 well as "retry timeout exceeded".
4020 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4021 do this, and (what is worse) MTAs that accept it.
4023 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4024 will now be deprecated.
4026 PH/56 New os.c-cygwin from the Cygwin maintainer.
4028 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4029 formats (previously only available in exim formats via -bpr, -bpru,
4030 and -bpra. Now also available in native and exiqgrep formats)
4032 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4033 with very large, slow to parse queues
4035 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4037 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4039 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4040 responses to authentication challenges, though it was showing the
4041 challenges; (ii) I've removed the CR characters from the debug output for
4044 PH/58 Allow for the insertion of a newline as well as a space when a string
4045 is turned into more than one encoded-word during RFC 2047 encoding. The
4046 Sieve code now uses this.
4048 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4049 data_4xx, lost_connection, tls_required.
4051 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4052 message was being sent as an SMTP response.
4054 PH/61 Add -l and -k options to exicyclog.
4056 PH/62 When verifying, if an address was redirected to one new address, so that
4057 verification continued, and the new address failed or deferred after
4058 having set something in $address_data, the value of $address_data was not
4059 passed back to the ACL. This was different to the case when no
4060 redirection occurred. The value is now passed back in both cases.
4062 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4063 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4064 use login_cap.h, so on its own it isn't the distinguishing feature. The
4065 new name refers directly to the setclassresources() function.
4067 PH/65 Added configuration files for NetBSD3.
4069 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4071 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4072 is preferred over IPv4.
4074 PH/68 The bounce_return_message and bounce_return_body options were not being
4075 honoured for bounces generated during the reception of non-SMTP messages.
4076 In particular, this applied to messages rejected by the ACL. This bug has
4077 been fixed. However, if bounce_return_message is true and bounce_return_
4078 body is false, the headers that are returned for a non-SMTP message
4079 include only those that have been read before the error was detected.
4080 (In the case of an ACL rejection, they have all been read.)
4082 PH/69 The HTML version of the specification is now built in a directory called
4083 spec_html instead of spec.html, because the latter looks like a path with
4084 a MIME-type, and this confuses some software.
4086 PH/70 Catch two compiler warnings in sieve.c.
4088 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4089 function verify_get_ident() calls ip_connect() to connect a socket, but
4090 if the "connect()" function timed out, ip_connect() used to close the
4091 socket. However, verify_get_ident() also closes the socket later, and in
4092 between Exim writes to the log, which may get opened at this point. When
4093 the socket was closed in ip_connect(), the log could get the same file
4094 descriptor number as the socket. This naturally causes chaos. The fix is
4095 not to close the socket in ip_connect(); the socket should be closed by
4096 the function that creates it. There was only one place in the code where
4097 this was missing, in the iplookup router, which I don't think anybody now
4098 uses, but I've fixed it anyway.
4100 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4101 well as to direct DNS lookups. Otherwise the handling of names in host
4102 lists is inconsistent and therefore confusing.
4108 PH/01 Two changes to the default runtime configuration:
4110 (1) Move the checks for relay_from_hosts and authenticated clients from
4111 after to before the (commented out) DNS black list checks.
4113 (2) Add control=submission to the relay_from_hosts and authenticated
4114 clients checks, on the grounds that messages accepted by these
4115 statements are most likely to be submissions.
4117 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4119 (1) Generate an error if the third argument for the ${prvs expansion is
4122 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4125 (3) Reset the variables that are obtained from the first argument of
4126 ${prvscheck and used in the second argument before leaving the code,
4127 because their memory is reclaimed, so using them afterwards may do
4130 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4131 one (it's much easier than Tom thought :-).
4133 (5) Because of (4), we can now allow for the use of $prvscheck_result
4134 inside the third argument.
4136 PH/03 For some reason, the default setting of PATH when running a command from
4137 a pipe transport was just "/usr/bin". I have changed it to
4140 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4141 anything to be listed in the output from -bV.
4143 PH/05 When a filter generated an autoreply, the entire To: header line was
4144 quoted in the delivery log line, like this:
4146 => >A.N.Other <ano@some.domain> <original@ddress> ...
4148 This has been changed so that it extracts the operative address. There
4149 may be more than one such address. If so, they are comma-separated, like
4152 => >ano@some.domain,ona@other.domain <original@ddress> ...
4154 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4155 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4156 not being looked up in the rDNS to get a host name, Exim was showing the
4157 IP address twice in Received: lines, even though the IP addresses were
4158 identical. For example:
4160 Received: from [1.2.3.4] (helo=[1.2.3.4])
4162 However, if the real host name was known, it was omitting the HELO data
4163 if it matched the actual IP address. This has been tidied up so that it
4164 doesn't show the same IP address twice.
4166 PH/07 When both +timestamp and +memory debugging was on, the value given by
4167 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4168 called by the debug printing, thereby overwriting the timestamp buffer.
4169 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4171 PH/08 When the original message was included in an autoreply transport, it
4172 always said "this is a copy of the message, including all the headers",
4173 even if body_only or headers_only was set. It now gives an appropriate
4176 PH/09 Applied a patch from the Sieve maintainer which:
4178 o fixes some comments
4179 o adds the (disabled) notify extension core
4180 o adds some debug output for the result of if/elsif tests
4181 o points to the current vacation draft in the documentation
4182 and documents the missing references header update
4186 o fixes a bug in processing the envelope test (when testing
4187 multiple envelope elements, the last element determined the
4190 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4191 Electronic Mail") by including:
4193 Auto-submitted: auto-generated
4195 in the messages that it generates (bounce messages and others, such as
4196 warnings). In the case of bounce messages for non-SMTP messages, there was
4197 also a typo: it was using "Auto_submitted" (underscore instead of
4198 hyphen). Since every message generated by Exim is necessarily in response
4199 to another message, thes have all been changed to:
4201 Auto-Submitted: auto-replied
4203 in accordance with these statements in the RFC:
4205 The auto-replied keyword:
4207 - SHOULD be used on messages sent in direct response to another
4208 message by an automatic process,
4210 - MUST NOT be used on manually-generated messages,
4212 - MAY be used on Delivery Status Notifications (DSNs) and Message
4213 Disposition Notifications (MDNs),
4215 - MUST NOT be used on messages generated by automatic or periodic
4216 processes, except for messages which are automatic responses to
4219 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4220 to the default Received: header definition.
4222 PH/12 Added log selector acl_warn_skipped (default on).
4224 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4225 variables because (a) they are in the wrong storage pool and (b) even if
4226 they were copied, it wouldn't work properly because of the caching.
4228 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4229 checking when decoding. Apparently there are clients that generate
4230 overlong encoded strings. Why am I not surprised?
4232 PH/15 If the first argument of "${if match_address" was not empty, but did not
4233 contain an "@" character, Exim crashed. Now it writes a panic log message
4234 and treats the condition as false.
4236 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4238 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4239 extension "envelope-auth". The code is finished and in agreement with
4240 other implementations, but there is no documentation so far and in fact,
4241 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4242 not changing the active code.
4244 Print executed "if" and "elsif" statements when debugging is used. This
4245 helps a great deal to understand what a filter does.
4247 Document more things not specified clearly in RFC3028. I had all this
4248 sorted out, when out of a sudden new issues came to my mind. Oops."
4250 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4253 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4254 canonical form (as documented). However, after a host name lookup from
4255 the IP address, check_host() was doing a simple string comparison with
4256 addresses acquired from the DNS when checking that the found name did
4257 have the original IP as one of its addresses. Since any found IPv6
4258 addresses are likely to be in abbreviated form, the comparison could
4259 fail. Luckily, there already exists a function for doing the comparison
4260 by converting both addresses to binary, so now that is used instead of
4261 the text comparison.
4263 PH/20 There was another similar case to PH/19, when a complete host name was
4264 given in a host list; looking up its IP address could give an abbreviated
4265 form, whereas the current host's name might or might not be abbreviated.
4266 The same fix has been applied.
4272 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4273 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4276 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4277 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4279 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4281 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4282 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4283 Also, for "Auto-Submitted", treat anything other than "no" as
4284 non-personal, in accordance with RFC 3834. (Previously it treated
4285 anything starting "auto-" as non-personal.)
4287 TF/01 The control=submission/name=... option had a problem with syntax
4288 errors if the name included a slash character. The /name= option
4289 now slurps the rest of the string, so it can include any characters
4290 but it must come last in the list of options (after /sender_retain
4293 PH/05 Some modifications to the interface to the fake nameserver for the new
4301 TK/01 Added the "success_on_redirect" address verification option. See
4302 NewStuff for rationale and an example.
4304 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4306 PH/02 Patch to exigrep to allow it to work on syslog lines.
4308 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4309 fread() to skip over the body file's header line, because in Cygwin the
4310 header line is locked and is inaccessible.
4312 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4313 co-exist for some time) to make it clear that it is the Exim ID that is
4314 referenced, not the Message-ID: header line.
4316 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4317 string_format() function, because snprintf() does not exist on all
4320 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4321 ${stat: expansion item.
4323 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4324 protocol synchronization error", to keep the pedants happy.
4326 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4327 well as for IRIX systems, when gcc is being used. See the host.c source
4330 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4332 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4335 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4336 transport and to the smtp transport in LMTP mode.
4338 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4340 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4341 run a filter in a subprocess. This could lead to confusion in subsequent
4342 lookups in the parent process. There should also be a search_tidyup() at
4343 the end of the subprocess.
4345 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4346 only if the host matched helo_try_verify_hosts, which caused the
4347 verification to occur when the EHLO/HELO command was issued. The ACL just
4348 tested the remembered result. Now, if a previous verification attempt has
4349 not happened, "verify = helo" does it there and then.
4351 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4353 TK/03 Fix log output including CR from clamd.
4355 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4356 exist provoked a memory error which could cause a segfault.
4358 PH/15 Installed PCRE 6.2
4360 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4362 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4363 of the problem. Specifically, suggested +O2 rather than +O1 for the
4366 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4368 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4369 record, the port was not correctly logged when the outgoing_port log
4370 selector was set (it logged the transort's default port).
4372 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4373 fallback_hosts, and "hosts" in the smtp transport.
4375 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4376 host errors such as "Connection refused".
4378 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4379 authentication with radiusclient 0.4.9:
4381 - Error returned from rc_read_config was caught wrongly
4382 - Username/password not passed on to radius server due to wrong length.
4384 The presumption is that some radiusclient API changes for 4.51/PH/17
4385 were not taken care of correctly. The code is still untested by me (my
4386 Linux distribution still has 0.3.2 of radiusclient), but it was
4387 contributed by a Radius user.
4389 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4390 expanding the "port" option of the smtp transport.
4392 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4393 while reading a MIME header. Thanks to Tom Hughes for a patch.
4395 PH/24 Include config.h inside local_scan.h so that configuration settings are
4398 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4399 that in an ACL for RCPT (for example), you can examine exactly what was
4402 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4403 commands, but it was not correctly comparing the address with the actual
4404 client host address. Thus, it would show the EHLO address in Received:
4405 header lines when this was not necessary.
4407 PH/27 Added the % operator to ${eval:}.
4409 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4410 it should be ignoring failures (because with -C, for example, it has lost
4411 privilege). It wasn't ignoring creation failures other than "already
4414 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4417 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4418 because an input line was too long, either on its own, or by virtue of
4419 too many continuations, the temporary file was not being removed, and the
4420 return code was incorrect.
4422 PH/31 Missing "BOOL" in function definition in filtertest.c.
4424 PH/32 Applied Sieve patches from the maintainer.
4426 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4428 PH/33 Added "verify = not_blind".
4430 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4431 Local/Makefile (with some defaults set). These are used in built scripts
4432 such as exicyclog, but they have never been used in the exim_install
4433 script (though there are many overriding facilities there). I have
4434 arranged that the exim_install script now takes note of these two
4437 PH/35 Installed configuration files for Dragonfly.
4439 PH/36 When a locally submitted message by a trusted user did not contain a
4440 From: header, and the sender address was obtained from -f or from an SMTP
4441 MAIL command, and the trusted user did not use -F to supply a sender
4442 name, $originator_name was incorrectly used when constructing a From:
4443 header. Furthermore, $originator_name was used for submission mode
4444 messages from external hosts without From: headers in a similar way,
4445 which is clearly wrong.
4447 PH/37 Added control=suppress_local_fixups.
4449 PH/38 When log_selector = +received_sender was set, and the addition of the
4450 sender made the log line's construction buffer exactly full, or one byte
4451 less than full, an overflow happened when the terminating "\n" was
4454 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4455 when the result of a list match is failure because a DNS lookup failed.
4457 PH/40 RM_COMMAND is now used in the building process.
4459 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4460 the "build-* directories that it finds.
4462 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4463 address in a domain literal was a prefix of an interface address.
4465 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4466 when verifying a sender address, unless rewrite_headers is false.
4468 PH/44 Wrote a long comment about why errors_to addresses are verified as
4469 recipients, not senders.
4471 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4472 the ratelimit ACL was added.
4474 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4476 PH/46 Added extra information about PostgreSQL errors to the error string.
4478 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4479 avoiding the need to install special zones in a real server. This is
4480 backwards compatible; if it can't find the fake resolver, it drops back.
4481 Thus, both old and new test suites can be run.
4483 TF/02 Added util/ratelimit.pl
4485 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4486 clock is set back in time.
4488 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4489 Candler <B.Candler@pobox.com>.
4491 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4492 OK for addresses that are the result of redirections.
4494 PH/48 A number of further additions for the benefit of the new test suite,
4495 including a fake gethostbyname() that interfaces to the fake DNS resolver
4498 TF/06 The fix for widen_domains has also been applied to qualify_single and
4499 search_parents which are the other dnslookup options that can cause
4502 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4505 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4507 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4508 service" if it encountered an MX record with an empty target hostname.
4509 The message is now "an MX or SRV record indicated no SMTP service".
4511 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4512 if the DNS of the sending site is misconfigured. This is quite a
4513 common situation. This change restores the behaviour of treating a
4514 helo verification defer as a failure.
4516 PH/51 If self=fail was set on a router, the bounce message did not include the
4517 actual error message.
4523 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4525 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4526 command itself ended in error, the underlying message about the transport
4527 filter timeout was being overwritten with the pipe command error. Now the
4528 underlying error message should be appended to the second error message.
4530 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4532 PH/02 Reduce the amount of output that "make" produces by default. Full output
4533 can still be requested.
4535 PH/03 The warning log line about a condition test deferring for a "warn" verb
4536 was being output only once per connection, rather than after each
4537 occurrence (because it was using the same function as for successful
4538 "warn" verbs). This seems wrong, so I have changed it.
4540 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4541 it should not have, which might have caused a crash in the right
4542 circumstances, but probably never did.
4544 PH/04 Installed a modified version of Tony Finch's patch to make submission
4545 mode fix the return path as well as the Sender: header line, and to
4546 add a /name= option so that you can make the user's friendly name appear
4549 TF/03 Added the control = fakedefer ACL modifier.
4551 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4552 Mark Lowes for thorough testing.
4554 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4556 TK/03 Merged latest SRS patch from Miles Wilton.
4558 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4559 with the definition in sysexits.h (which is #included earlier).
4560 Fortunately, Exim does not actually use EX_OK. The code used to try to
4561 preserve the sysexits.h value, by assuming that macro definitions were
4562 scanned for macro replacements. I have been disabused of this notion,
4563 so now the code just undefines EX_OK before #including unistd.h.
4565 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4566 in the smtp transport. When a block could not be written in a single
4567 write() function, the timeout was being re-applied to each part-write.
4568 This seems wrong - if the receiver was accepting one byte at a time it
4569 would take for ever. The timeout is now adjusted when this happens. It
4570 doesn't have to be particularly precise.
4572 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4573 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4575 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4576 without a domain if the check to <postmaster@domain> fails.
4578 SC/01 Eximstats: added -xls and the ability to specify output files
4579 (patch written by Frank Heydlauf).
4581 SC/02 Eximstats: use FileHandles for outputting results.
4583 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4585 SC/04 Eximstats: fixed display of large numbers with -nvr option
4587 SC/05 Eximstats: fixed merging of reports with empty tables.
4589 SC/06 Eximstats: added the -include_original_destination flag
4591 SC/07 Eximstats: removed tabs and trailing whitespace.
4593 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4595 TK/06 MBOX spool code: Add real "From " MBOX separator line
4596 so the .eml file is really in mbox format (even though
4597 most programs do not really care). Patch from Alex Miller.
4599 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4600 The latter is generated from $received_to and is only set if the
4601 message has one envelope recipient. SA can use these headers,
4602 obviously out-of-the-box. Patch from Alex Miller.
4604 PH/08 The ${def test on a variable was returning false if the variable's
4605 value was "0", contrary to what the specification has always said!
4606 The result should be true unless the variable is empty.
4608 PH/09 The syntax error of a character other than { following "${if
4609 def:variable_name" (after optional whitespace) was not being diagnosed.
4610 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4611 accidental colon was present, for example, could give incorrect results.
4613 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4614 result is used (not including appendfile, where other changes are about
4617 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4618 This involved changing a lot of size variables from int to off_t. It
4619 should work with maildirs and everything.
4621 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4622 spamd dying while we are connected to it.
4624 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4627 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4628 function for BDB 4.3.
4630 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4632 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4633 the sources for PCRE have been re-arranged and more files are now
4636 PH/15 The code I had for printing potentially long long variables in PH/11
4637 above was not the best (it lost precision). The length of off_t variables
4638 is now inspected at build time, and an appropriate printing format (%ld
4639 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4640 to be "long long int" or "long int". This is needed for the internal
4641 formatting function string_vformat().
4643 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4644 the configuration file to be ":syslog", then the script "guesses" where
4645 the logs files are, rather than using the compiled in default. In our
4646 case the guess is not the same as the compiled default, so the script
4647 suddenly stopped working when I started to use syslog. The patch checks
4648 to see if log_file_path is "". If so, it attempts to read it from exim
4649 with no configuration file to get the compiled in version, before it
4650 falls back to the previous guessing code."
4652 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4653 implementing BATV in an Exim configuration. See NewStuff for the gory
4656 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4657 Makefile that are specific to HP-UX.
4659 PH/18 If the "use_postmaster" option was set for a recipient callout together
4660 with the "random" option, the postmaster address was used as the MAIL
4661 FROM address for the random test, but not for the subsequent recipient
4662 test. It is now used for both.
4664 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4665 patch removes a few documentation additions to RFC 3028, because the
4666 latest draft now contains them. It adds the new en;ascii-case comparator
4667 and a new error check for 8bit text in MIME parts. Comparator and
4668 require names are now matched exactly. I enabled the subaddress
4669 extension, but it is not well tested yet (read: it works for me)."
4671 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4672 rework some of the code of TK/09 above to avoid the hardwired use of
4673 "%lld" and "long long". Replaced the call to snprintf() with a call to
4676 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4677 records point to non-existent hosts", "retry timeout exceeded", and
4678 "retry time not reached for any host after a long failure period".
4680 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4681 experimental DomainKeys support:
4683 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4684 (2) On an error such as an illegally used "control", the wrong name for
4685 the control was given.
4687 These problems did NOT occur unless DomainKeys support was compiled.
4689 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4691 PH/24 Added ${if match_ip condition.
4693 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4694 databases so that it will be absolutely obvious if a crash occurs in the
4695 DB library. This is a regular occurrence (often caused by mis-matched
4698 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4699 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4700 picked up on a user's system that detects such things. There doesn't seem
4701 to be a gcc warning option for this - only an attribute that has to be
4702 put on the function's prototype. It seems that in Fedora Core 4 they have
4703 set this on a number of new functions. No doubt there will be more in due
4706 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4707 specify a transport. However, if an address that was verified by such a
4708 router was the subject of a callout, Exim crashed because it tried to
4709 read the rcpt_include_affixes from the non-existent transport. Now it
4710 just assumes that the setting of that option is false. This bug was
4711 introduced by 4.51/PH/31.
4713 PH/28 Changed -d+all to exclude +memory, because that information is very
4714 rarely of interest, but it makes the output a lot bigger. People tend to
4715 do -d+all out of habit.
4717 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4718 code in os-type was giving problems when libc.so lives in lib64, like on
4721 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4722 aren't the modern standard, and it seems that some systems' include files
4723 don't always have them. Exim was already checking for some of the newer
4724 ones like T_AAAA, and defining it itself. I've added checks for all the
4725 record types that Exim uses.
4727 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4728 not automatically generating a new one, as it is supposed to. This
4729 prevented TLS from working. If the file did exist, but contained invalid
4730 data, a new version was generated, as expected. It was only the case of a
4731 non-existent file that was broken.
4733 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4734 with a change in libdomainkeys > 0.64.
4736 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4737 from DNS. If the selector record carries the flag, it now has
4738 precedence over the domain-wide flag.
4740 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4742 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4743 the use of a port name that isn't defined in /etc/services) occurred, the
4744 message was deferred as in a normal delivery, and thus remained on the
4745 spool, instead of being failed because of the mua_wrapper setting. This
4746 is now fixed, and I tidied up some of the mua_wrapper messages at the
4749 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4750 the messages in a hash of arrays rather than using individual hashes.
4751 This is a bit cleaner and results in dramatic memory savings, albeit
4752 at a slight CPU cost.
4754 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4755 as requested by Marc Sherman.
4757 SC/10 Eximstats: added histograms for user specified patterns as requested
4760 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4762 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4763 fopen() in the content-scanning modules that did not already have it.
4769 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4770 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4772 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4774 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4776 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4777 to test Sieve filters that use "vacation".
4779 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4780 that changes the way the GnuTLS parameters are stored in the cache file.
4781 The new format can be generated externally. For backward compatibility,
4782 if the data in the cache doesn't make sense, Exim assumes it has read an
4783 old-format file, and it generates new data and writes a new file. This
4784 means that you can't go back to an older release without removing the
4787 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4788 work if there are any delivery delays because "one_time" forces the
4789 parent to be marked "delivered", so its unseen clone is never tried
4790 again. For this reason, Exim now forbids the simultaneous setting of
4793 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4794 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4795 entry if you want to know the details. The fix, however, neglected to
4796 consider the case when local delivery batching is involved. The test for
4797 "previously delivered" was not happening when checking to see if an
4798 address could be batched with a previous (undelivered) one; under
4799 certain circumstances this could lead to multiple deliveries to the same
4802 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4803 in its include files, and this causes problems building Exim.
4805 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4806 header_syntax) but Exim was just ignoring anything given after a slash.
4807 In particular, this caused confusion with an attempt to use "verify =
4808 reverse_host_lookup/defer_ok". An error is now given when options are
4809 supplied for verify items that do not have them. (Maybe reverse_host_
4810 lookup should have a defer_ok option, but that's a different point.)
4812 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4813 defined by RFC 821) to 2048, because there were problems with some AUTH
4814 commands, and RFC 1869 says the size should be increased for extended
4815 SMTP commands that take arguments.
4817 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4820 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4821 "unknown" error; now it says that the functionality isn't in the binary.
4823 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4824 an address' error message when a string expansion fails (syntax or
4825 whatever). Otherwise the password may appear in the log. Following change
4826 PH/42 below, there is no longer a chance of it appearing in a bounce
4829 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4831 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4832 of its addresses was ever tried. (Bugzilla bug #2.)
4834 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4835 the result incorrectly in the debug output. (It correctly added a newline
4836 to what was transported.)
4838 TF/01 Added $received_time.
4840 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4841 commented out examples of how to interface to a virus scanner and to
4842 SpamAssassin. Also added commented examples of av_scanner and
4843 spamd_address settings.
4845 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4846 and controls are allowed in which ACLs. There were a couple of minor
4847 errors. Some of the entries in the conditions table (which is a table of
4848 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4849 negation of where the condition IS allowed.
4851 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4853 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4854 header file does not have a version number, so I've had to invent a new
4855 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4856 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4857 radiusclient), but it was contributed by a Radius user.
4859 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4860 files or directories, for OpenSSL.
4862 PH/19 When an Exim process that is running as root has to create an Exim log
4863 file, it does so in a subprocess that runs as exim:exim so as to get the
4864 ownership right at creation (otherwise, other Exim processes might see
4865 the file with the wrong ownership). There was no test for failure of this
4866 fork() call, which would lead to the process getting stuck as it waited
4867 for a non-existent subprocess. Forks do occasionally fail when resources
4868 run out. I reviewed all the other calls to fork(); they all seem to check
4871 PH/20 When checking for unexpected SMTP input at connect time (before writing
4872 the banner), Exim was not dealing correctly with a non-positive return
4873 from the read() function. If the client had disconnected by this time,
4874 the result was a log entry for a synchronization error with an empty
4875 string after "input=" when read() returned zero. If read() returned -1
4876 (an event I could not check), uninitialized data bytes were printed.
4877 There were reports of junk text (parts of files, etc) appearing after
4880 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4882 PH/22 Added support for macro redefinition, and (re)definition in between
4883 driver and ACL definitions.
4885 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4886 forgetting to use the resulting value; it was using the unexpanded value.
4888 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4889 hadn't been configured. The fix is from Juergen Kreileder, who
4890 understands it better than I do:
4892 "Here's what I see happening with three configured cyrus_sasl
4893 authenticators configured (plain, login, cram-md5):
4895 On startup auth_cyrus_sasl_init() gets called for each of these.
4896 This means three calls to sasl_listmech() without a specified mech_list.
4897 => SASL tests which mechs of all available mechs actually work
4898 => three warnings about OTP not working
4899 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4901 With the patch, sasl_listmech() also gets called three times. But now
4902 SASL's mech_list option is set to the server_mech specified in the the
4903 authenticator. Or in other words, the answer from sasl_listmech()
4904 gets limited to just the mech you're testing for (which is different
4906 => the return list contains just 'plain' or 'login', 'cram-md5' or
4907 nothing depending on the value of ob->server_mech.
4909 I've just tested the patch: Authentication still works fine,
4910 unavailable mechs specified in the exim configuration are still
4911 caught, and the auth.log warnings about OTP are gone."
4913 PH/25 When debugging is enabled, the contents of the command line are added
4914 to the debugging output, even when log_selector=+arguments is not
4917 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4918 answer is "GNU", and only if the return is "GNU/something" is the answer
4921 PH/27 $acl_verify_message is now set immediately after the failure of a
4922 verification in an ACL, and so is available in subsequent modifiers. In
4923 particular, the message can be preserved by coding like this:
4925 warn !verify = sender
4926 set acl_m0 = $acl_verify_message
4928 Previously, $acl_verify_message was set only while expanding "message"
4929 and "log_message" when a very denied access.
4931 PH/28 Modified OS/os.c-Linux with
4933 -#ifndef OS_LOAD_AVERAGE
4934 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4936 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4937 nomenclature these days.)
4939 PH/29 Installed patch from the Sieve maintainer that adds the options
4940 sieve_useraddress and sieve_subaddress to the redirect router.
4942 PH/30 In these circumstances:
4943 . Two addresses routed to the same list of hosts;
4944 . First host does not offer TLS;
4945 . First host accepts first address;
4946 . First host gives temporary error to second address;
4947 . Second host offers TLS and a TLS session is established;
4948 . Second host accepts second address.
4949 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4950 and peerdn, if requested) that were in fact used only for the second
4953 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4954 attention to any local part prefix or suffix that was matched by the
4955 router that accepted the address. It now behaves in the same way as it
4956 does for delivery: the affixes are removed from the local part unless
4957 rcpt_include_affixes is set on the transport.
4959 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4960 timeout during the DATA phase of an incoming message.
4962 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4963 applied a patch sanctioned by the Sieve maintainer.
4965 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4966 the uid or gid is negative. A case of a negative gid caused this to be
4967 noticed. The fix allows for either to be negative.
4969 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4970 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4973 PH/36 The change PH/12 above was broken. Fixed it.
4975 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4976 the grounds that routing the same address twice would always produce the
4977 same answer. This might have been true once, but it is certainly no
4978 longer true now. Routing a child address may depend on the previous
4979 routing that produced that child. Some complicated redirection strategies
4980 went wrong when messages had multiple recipients, and made Exim's
4981 behaviour dependent on the order in which the addresses were given.
4983 I have moved the duplicate checking until after the routing is complete.
4984 Exim scans the addresses that are assigned to local and remote
4985 transports, and removes any duplicates. This means that more work will be
4986 done, as duplicates will always all be routed, but duplicates are
4987 presumably rare, so I don't expect this is of any significance.
4989 For deliveries to pipes, files, and autoreplies, the duplicate checking
4990 still happens during the routing process, since they are not going to be
4993 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4994 It corrects a timeout issue with spamd. This is Ian's comment: "The
4995 background is that sometimes spamd either never reads data from a
4996 connection it has accepted, or it never writes response data. The exiscan
4997 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4998 blindly assumes that writes won't block so it may never time out."
5000 PH/39 Allow G after quota size as well as K and M.
5002 PH/40 The value set for $authenticated_id in an authenticator may not contain
5003 binary zeroes or newlines because the value is written to log lines and
5004 to spool files. There was no check on this. Now the value is run through
5005 the string_printing() function so that such characters are converted to
5006 printable escape sequences.
5008 PH/41 $message_linecount is a new variable that contains the total number of
5009 lines in the message. Compare $body_linecount, which is the count for the
5012 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5013 bounce and delay warning messages, except in certain special cases, which
5016 (a) An SMTP error message from a remote host;
5017 (b) A message specified in a :fail: redirection;
5018 (c) A message specified in a "fail" command in a system filter;
5019 (d) A message specified in a FAIL return from the queryprogram router;
5020 (e) A message specified by the cannot_route_message router option.
5022 In these cases only, Exim does include the error details in bounce and
5023 warning messages. There are also a few cases where bland messages such
5024 as "unrouteable address" or "local delivery error" are given.
5026 PH/43 $value is now also set for the "else" part of a ${run expansion.
5028 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5029 being worked on, but at least Exim now implements the latest version to
5032 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5033 process to complete was treated as a delivery failure, a timeout while
5034 writing the message to the pipe was logged, but erroneously treated as a
5035 successful delivery. Such timeouts include transport filter timeouts. For
5036 consistency with the overall process timeout, these timeouts are now
5037 treated as errors, giving rise to delivery failures by default. However,
5038 there is now a new Boolean option for the pipe transport called
5039 timeout_defer, which, if set TRUE, converts the failures into defers for
5040 both kinds of timeout. A transport filter timeout is now identified in
5043 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5044 systems where "make" and "gmake" are different, calling "gmake" at top
5045 level broke things. I've arranged for the value of $(MAKE) to be passed
5046 from the Makefile to this script so that it can call the same version of
5050 A note about Exim versions 4.44 and 4.50
5051 ----------------------------------------
5053 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5054 changes of various kinds. As a consequence, a big documentation update was
5055 needed. This delayed the release for rather longer than seemed good, especially
5056 in the light of a couple of (minor) security issues. Therefore, the changes
5057 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5058 release. So 4.44 and 4.50 are in effect two different branches that both start
5061 I have left the 4.50 change log unchanged; it contains all the changes since
5062 4.43. The change log for 4.44 is below; many of its items are identical to
5063 those for 4.50. This seems to be the most sensible way to preserve the
5064 historical information.
5070 1. Minor wording change to the doc/README.SIEVE file.
5072 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5073 computation of the current number of files was incorrect.
5075 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5076 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5079 4. Give more explanation in the error message when the command for a transport
5080 filter fails to execute.
5082 5. There are several places where Exim runs a non-Exim command in a
5083 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5084 was being done only for the command run by the queryprogram router. It is
5085 now done for all such subprocesses. The other cases are: ${run, transport
5086 filters, and the commands run by the lmtp and pipe transports.
5088 6. Added CONFIGURE_GROUP build-time option.
5090 7. Some older OS have a limit of 256 on the maximum number of file
5091 descriptors. Exim was using setrlimit() to set 1000 as a large value
5092 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5093 systems. I've change it so that if it can't get 1000, it tries for 256.
5095 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5096 was an oversight, and furthermore, ever since the addition of extra
5097 controls (e.g. 4.43/32), the checks on when to allow different forms of
5098 "control" were broken. There should now be diagnostics for all cases when a
5099 control that does not make sense is encountered.
5101 9. Added the /retain_sender option to "control=submission".
5103 10. $recipients is now available in the predata ACL (oversight).
5105 11. Tidy the search cache before the fork to do a delivery from a message
5106 received from the command line. Otherwise the child will trigger a lookup
5107 failure and thereby defer the delivery if it tries to use (for example) a
5108 cached ldap connection that the parent has called unbind on.
5110 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5111 of $address_data from the recipient verification was clobbered by the
5112 sender verification.
5114 13. The value of address_data from a sender verification is now available in
5115 $sender_address_data in subsequent conditions in the ACL statement.
5117 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5119 15. Added a new option "connect=<time>" to callout options, to set a different
5122 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5123 was its contents. (It was OK if the option was not defined at all.)
5125 17. A "Completed" log line is now written for messages that are removed from
5126 the spool by the -Mrm option.
5128 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5129 information about exactly what failed.
5131 19. Added -dd to debug only the daemon process.
5133 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5134 handles timeouts, both on the server side and network timeouts. Renamed the
5135 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5137 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5138 It is now set to "smtps".
5140 22. $host_address is now set to the target address during the checking of
5141 ignore_target_hosts.
5143 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5144 being passed; this would have caused $sender_host_name to have been used if
5145 matching the list had actually called for a host name (not very likely,
5146 since this list is usually IP addresses). A host name is now passed as
5149 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5150 code that specifies a non-restarting handler (typically sigaction() in
5151 modern systems) in an attempt to fix a rare and obscure crash bug.
5153 25. Narrowed the window for a race in the daemon that could cause it to ignore
5154 SIGCHLD signals. This is not a major problem, because they are used only to
5155 wake it up if nothing else does.
5157 26. A malformed maildirsize file could cause Exim to calculate negative values
5158 for the mailbox size or file count. Odd effects could occur as a result.
5159 The maildirsize information is now recalculated if the size or filecount
5162 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5163 support for a long time. Removed HAVE_SYS_VFS_H.
5165 28. Installed the latest version of exipick from John Jetmore.
5167 29. In an address list, if the pattern was not a regular expression, an empty
5168 subject address (from a bounce message) matched only if the pattern was an
5169 empty string. Non-empty patterns were not even tested. This was the wrong
5170 because it is perfectly reasonable to use an empty address as part of a
5171 database query. An empty address is now tested by patterns that are
5172 lookups. However, all the other forms of pattern expect the subject to
5173 contain a local part and a domain, and therefore, for them, an empty
5174 address still always fails if the pattern is not itself empty.
5176 30. Exim went into a mad DNS loop when attempting to do a callout where the
5177 host was specified on an smtp transport, and looking it up yielded more
5178 than one IP address.
5180 31. Re-factored the code for checking spool and log partition space into a
5181 function that finds that data and another that does the check. The former
5182 is then used to implement four new variables: $spool_space, $log_space,
5183 $spool_inodes, and $log_inodes.
5185 32. The RFC2047 encoding function was originally intended for short strings
5186 such as real names; it was not keeping to the 75-character limit for
5187 encoded words that the RFC imposes. It now respects the limit, and
5188 generates multiple encoded words if necessary. To be on the safe side, I
5189 have increased the buffer size for the ${rfc2047: expansion operator from
5192 33. It is now permitted to omit both strings after an "if" condition; if the
5193 condition is true, the result is "true". As before, when the second string
5194 is omitted, a false condition yields an empty string. This makes it less
5195 cumbersome to write custom ACL and router conditions.
5197 34. Failure to deliver a bounce message always caused it to be frozen, even if
5198 there was an errors_to setting on the router. The errors_to setting is now
5201 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5202 canonical form (fully expanded) before being placed in
5203 $sender_host_address.
5205 36. The table in the code that translates DNS record types into text (T_A to
5206 "A" for instance) was missing entries for NS and CNAME. It is just possible
5207 that this could have caused confusion if both these types were looked up
5208 for the same domain, because the text type is used as part of Exim's
5209 per-process caching. But the chance of anyone hitting this buglet seems
5212 37. The dnsdb lookup has been extended in a number of ways.
5214 (1) There is a new type, "zns", which walks up the domain tree until it
5215 finds some nameserver records. It should be used with care.
5217 (2) There is a new type, "mxh", which is like "mx" except that it returns
5218 just the host names, not the priorities.
5220 (3) It is now possible to give a list of domains (or IP addresses) to be
5221 looked up. The behaviour when one of the lookups defers can be
5222 controlled by a keyword.
5224 (4) It is now possible to specify the separator character for use when
5225 multiple records are returned.
5227 38. The dnslists ACL condition has been extended: it is now possible to supply
5228 a list of IP addresses and/or domains to be looked up in a particular DNS
5231 39. Added log_selector=+queue_time_overall.
5233 40. When running the queue in the test harness, wait just a tad after forking a
5234 delivery process, to get repeatability of debugging output.
5236 41. Include certificate and key file names in error message when GnuTLS fails
5237 to set them up, because the GnuTLS error message doesn't include the name
5238 of the failing file when there is a problem reading it.
5240 42. Allow both -bf and -bF in the same test run.
5242 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5244 44. The "Exiscan patch" is now merged into the mainline Exim source.
5246 45. Sometimes the final signoff response after QUIT could fail to get
5247 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5248 < 0 before doing a fflush(). This bug looks as though it goes back to the
5249 introduction of TLS in release 3.20, but "sometimes" must have been rare
5250 because the tests only now provoked it.
5252 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5253 (this can affect the format of dates).
5255 47. exim_tidydb, when checking for the continued existence of a message for
5256 which it has found a message-specific retry record, was not finding
5257 messages that were in split spool directories. Consequently, it was
5258 deleting retry records that should have stayed in existence.
5260 48. Steve fixed some bugs in eximstats.
5262 49. The SPA authentication driver was not abandoning authentication and moving
5263 on to the next authenticator when an expansion was forced to fail,
5264 contradicting the general specification for all authenticators. Instead it
5265 was generating a temporary error. It now behaves as specified.
5267 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5268 (the order specifies the preference for clients). The order is now AES256,
5269 AES128, 3DES, ARCFOUR128.
5271 51. Small patch to Sieve code - explicitly set From: when generating an
5274 52. Exim crashed if a remote delivery caused a very long error message to be
5275 recorded - for instance if somebody sent an entire SpamAssassin report back
5276 as a large number of 550 error lines. This bug was coincidentally fixed by
5277 increasing the size of one of Exim's internal buffers (big_buffer) that
5278 happened as part of the Exiscan merge. However, to be on the safe side, I
5279 have made the code more robust (and fixed the comments that describe what
5282 53. Now that there can be additional text after "Completed" in log lines (if
5283 the queue_time_overall log selector is set), a one-byte patch to exigrep
5284 was needed to allow it to recognize "Completed" as not the last thing in
5287 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5288 patch that reportedly fixes this has been added. I am not expert enough to
5289 create a test for it. This is what the patch creator wrote:
5291 "I found a little strange behaviour of ldap code when working with
5292 Windows 2003 AD Domain, where users was placed in more than one
5293 Organization Units. When I tried to give exim partial DN, the exit code
5294 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5295 But simultaneously result of request was absolutely normal ldap result,
5296 so I produce this patch..."
5298 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5299 REFERENCE, so I have modified the code to exclude the patch when that macro
5302 55. Some experimental protocols are using DNS PTR records for new purposes. The
5303 keys for these records are domain names, not reversed IP addresses. The
5304 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5305 leaves it alone. Component reversal etc. now happens only for IP addresses.
5308 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5310 57. Double the size of the debug message buffer (to 2048) so that more of very
5311 long debug lines gets shown.
5313 58. The exicyclog utility now does better if the number of log files to keep
5314 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5316 59. Two changes related to the smtp_active_hostname option:
5318 (1) $smtp_active_hostname is now available as a variable.
5319 (2) The default for smtp_banner uses $smtp_active_hostname instead
5320 of $primary_hostname.
5322 60. The host_aton() function is supposed to be passed a string that is known
5323 to be a valid IP address. However, in the case of IPv6 addresses, it was
5324 not checking this. This is a hostage to fortune. Exim now panics and dies
5325 if the condition is not met. A case was found where this could be provoked
5326 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5327 components; fortuitously, this particular loophole had already been fixed
5328 by change 4.50/55 above.
5330 If there are any other similar loopholes, the new check in host_aton()
5331 itself should stop them being exploited. The report I received stated that
5332 data on the command line could provoke the exploit when Exim was running as
5333 exim, but did not say which command line option was involved. All I could
5334 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5335 running as the user.
5338 61. There was a buffer overflow vulnerability in the SPA authentication code
5339 (which came originally from the Samba project). I have added a test to the
5340 spa_base64_to_bits() function which I hope fixes it.
5343 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5344 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5346 63. The daemon start-up calls getloadavg() while still root for those OS that
5347 need the first call to be done as root, but it missed one case: when
5348 deliver_queue_load_max is set with deliver_drop_privilege. This is
5349 necessary for the benefit of the queue runner, because there is no re-exec
5350 when deliver_drop_privilege is set.
5352 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5353 This has been fixed.
5355 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5356 was in use, was not putting the data itself into the right store pool;
5357 consequently, it could be overwritten for a subsequent message in the same
5358 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5361 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5363 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5364 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5365 the function were treating the return as a boolean value, which happened to
5366 work because 0=false and not-0=true, but is not correct code.
5368 68. The host_aton() function was not handling scoped IPv6 addresses (those
5369 with, for example, "%eth0" on the end) correctly.
5371 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5372 negated items (that is, ~something) in unsigned ints. Some compilers
5373 apparently mutter when there is no cast.
5375 70. If an address verification called from an ACL failed, and did not produce a
5376 user-specific message (i.e. there was only a "system" message), nothing was
5377 put in $acl_verify_message. In this situation, it now puts the system
5380 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5381 session; change 4.31/43 added the unwanted input to the log line - except
5382 that it did not do this in the start of session case. It now does.
5384 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5385 This is wrong and can cause the other end to generate a synchronization
5386 error if it is another Exim or anything else that does the synchronization
5387 check. A QUIT command is no longer sent after a timeout.
5389 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5390 during host lookups.
5392 74. The defer_ok option of callout verification was not working if it was used
5393 when verifying addresses in header lines, that is, for this case:
5395 verify = header_sender/callout=defer_ok
5397 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5398 those file descriptors could be used for SMTP connections. If anything
5399 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5400 could be sent to the SMTP client, causing chaos. The daemon now opens
5401 stdin, stdout, and stderr to /dev/null when it puts itself into the
5404 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5405 log by default. The user can override this with suitable Perl magic.
5407 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5408 the log message when discard triggers, was not working for the DATA ACL or
5409 for the non-SMTP ACL.
5411 78. Error message wording change in sieve.c.
5413 79. If smtp_accept_max_per_host was set, the number of connections could be
5414 restricted to fewer than expected, because the daemon was trying to set up
5415 a new connection before checking whether the processes handling previous
5416 connections had finished. The check for completed processes is now done
5417 earlier. On busy systems, this bug wouldn't be noticed because something
5418 else would have woken the daemon, and it would have reaped the completed
5421 80. If a message was submitted locally by a user whose login name contained one
5422 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5423 It caused a spool format error. I have fixed the spool reading code. A
5424 related problem was that the "from" clause in the Received: line became
5425 illegal because of the space(s). It is now covered by ${quote_local_part.
5427 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5430 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5432 83. Patch from Sieve maintainer to fix unterminated string problem in
5433 "vacation" handling.
5435 84. Some minor changes to the Linux configuration files to help with other
5436 OS variants using glibc.
5438 85. One more patch for Sieve to update vacation handling to latest spec.
5441 ----------------------------------------------------
5442 See the note above about the 4.44 and 4.50 releases.
5443 ----------------------------------------------------
5449 1. Change 4.43/35 introduced a bug that caused file counts to be
5450 incorrectly computed when quota_filecount was set in an appendfile
5453 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5454 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5457 3. Give more explanation in the error message when the command for a transport
5458 filter fails to execute.
5460 4. There are several places where Exim runs a non-Exim command in a
5461 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5462 was being done only for the command run by the queryprogram router. It is
5463 now done for all such subprocesses. The other cases are: ${run, transport
5464 filters, and the commands run by the lmtp and pipe transports.
5466 5. Some older OS have a limit of 256 on the maximum number of file
5467 descriptors. Exim was using setrlimit() to set 1000 as a large value
5468 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5469 systems. I've change it so that if it can't get 1000, it tries for 256.
5471 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5472 was an oversight, and furthermore, ever since the addition of extra
5473 controls (e.g. 4.43/32), the checks on when to allow different forms of
5474 "control" were broken. There should now be diagnostics for all cases when a
5475 control that does not make sense is encountered.
5477 7. $recipients is now available in the predata ACL (oversight).
5479 8. Tidy the search cache before the fork to do a delivery from a message
5480 received from the command line. Otherwise the child will trigger a lookup
5481 failure and thereby defer the delivery if it tries to use (for example) a
5482 cached ldap connection that the parent has called unbind on.
5484 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5485 of $address_data from the recipient verification was clobbered by the
5486 sender verification.
5488 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5489 was its contents. (It was OK if the option was not defined at all.)
5491 11. A "Completed" log line is now written for messages that are removed from
5492 the spool by the -Mrm option.
5494 12. $host_address is now set to the target address during the checking of
5495 ignore_target_hosts.
5497 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5498 being passed; this would have caused $sender_host_name to have been used if
5499 matching the list had actually called for a host name (not very likely,
5500 since this list is usually IP addresses). A host name is now passed as
5503 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5504 code that specifies a non-restarting handler (typically sigaction() in
5505 modern systems) in an attempt to fix a rare and obscure crash bug.
5507 15. Narrowed the window for a race in the daemon that could cause it to ignore
5508 SIGCHLD signals. This is not a major problem, because they are used only to
5509 wake it up if nothing else does.
5511 16. A malformed maildirsize file could cause Exim to calculate negative values
5512 for the mailbox size or file count. Odd effects could occur as a result.
5513 The maildirsize information is now recalculated if the size or filecount
5516 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5517 support for a long time. Removed HAVE_SYS_VFS_H.
5519 18. Updated exipick to current release from John Jetmore.
5521 19. Allow an empty sender to be matched against a lookup in an address list.
5522 Previously the only cases considered were a regular expression, or an
5525 20. Exim went into a mad DNS lookup loop when doing a callout where the
5526 host was specified on the transport, if the DNS lookup yielded more than
5529 21. The RFC2047 encoding function was originally intended for short strings
5530 such as real names; it was not keeping to the 75-character limit for
5531 encoded words that the RFC imposes. It now respects the limit, and
5532 generates multiple encoded words if necessary. To be on the safe side, I
5533 have increased the buffer size for the ${rfc2047: expansion operator from
5536 22. Failure to deliver a bounce message always caused it to be frozen, even if
5537 there was an errors_to setting on the router. The errors_to setting is now
5540 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5541 canonical form (fully expanded) before being placed in
5542 $sender_host_address.
5544 24. Updated eximstats to version 1.33
5546 25. Include certificate and key file names in error message when GnuTLS fails
5547 to set them up, because the GnuTLS error message doesn't include the name
5548 of the failing file when there is a problem reading it.
5550 26. Expand error message when OpenSSL has problems setting up cert/key files.
5553 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5554 (this can affect the format of dates).
5556 28. exim_tidydb, when checking for the continued existence of a message for
5557 which it has found a message-specific retry record, was not finding
5558 messages that were in split spool directories. Consequently, it was
5559 deleting retry records that should have stayed in existence.
5561 29. eximstats updated to version 1.35
5562 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5563 1.35 - bugfix such that pie charts by volume are generated correctly
5565 30. The SPA authentication driver was not abandoning authentication and moving
5566 on to the next authenticator when an expansion was forced to fail,
5567 contradicting the general specification for all authenticators. Instead it
5568 was generating a temporary error. It now behaves as specified.
5570 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5571 (the order specifies the preference for clients). The order is now AES256,
5572 AES128, 3DES, ARCFOUR128.
5574 31. Small patch to Sieve code - explicitly set From: when generating an
5577 32. Exim crashed if a remote delivery caused a very long error message to be
5578 recorded - for instance if somebody sent an entire SpamAssassin report back
5579 as a large number of 550 error lines. This bug was coincidentally fixed by
5580 increasing the size of one of Exim's internal buffers (big_buffer) that
5581 happened as part of the Exiscan merge. However, to be on the safe side, I
5582 have made the code more robust (and fixed the comments that describe what
5585 33. Some experimental protocols are using DNS PTR records for new purposes. The
5586 keys for these records are domain names, not reversed IP addresses. The
5587 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5588 leaves it alone. Component reversal etc. now happens only for IP addresses.
5591 34. The host_aton() function is supposed to be passed a string that is known
5592 to be a valid IP address. However, in the case of IPv6 addresses, it was
5593 not checking this. This is a hostage to fortune. Exim now panics and dies
5594 if the condition is not met. A case was found where this could be provoked
5595 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5596 components; fortuitously, this particular loophole had already been fixed
5597 by change 4.50/55 or 4.44/33 above.
5599 If there are any other similar loopholes, the new check in host_aton()
5600 itself should stop them being exploited. The report I received stated that
5601 data on the command line could provoke the exploit when Exim was running as
5602 exim, but did not say which command line option was involved. All I could
5603 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5604 running as the user.
5607 35. There was a buffer overflow vulnerability in the SPA authentication code
5608 (which came originally from the Samba project). I have added a test to the
5609 spa_base64_to_bits() function which I hope fixes it.
5612 36. The daemon start-up calls getloadavg() while still root for those OS that
5613 need the first call to be done as root, but it missed one case: when
5614 deliver_queue_load_max is set with deliver_drop_privilege. This is
5615 necessary for the benefit of the queue runner, because there is no re-exec
5616 when deliver_drop_privilege is set.
5618 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5619 was in use, was not putting the data itself into the right store pool;
5620 consequently, it could be overwritten for a subsequent message in the same
5621 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5624 38. Sometimes the final signoff response after QUIT could fail to get
5625 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5626 < 0 before doing a fflush(). This bug looks as though it goes back to the
5627 introduction of TLS in release 3.20, but "sometimes" must have been rare
5628 because the tests only now provoked it.
5634 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5635 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5636 Now it returns an integer. A number of calls were still expecting a T/F
5637 return. Fortuitously, in all cases, the tests worked in OK situations,
5638 which is the norm. However, things would have gone wrong on any write
5639 failures on the smtp file descriptor. This function is used when sending
5640 messages over SMTP and also when doing verify callouts.
5642 2. When Exim is called to do synchronous delivery of a locally submitted
5643 message (the -odf or -odi options), it no longer closes stderr before doing
5646 3. Implemented the mua_wrapper option.
5648 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5650 5. Implemented the functions header_remove(), header_testname(),
5651 header_add_at_position(), and receive_remove_recipient(), and exported them
5654 6. If an ACL "warn" statement specified the addition of headers, Exim already
5655 inserted X-ACL-Warn: at the start if there was no header name. However, it
5656 was not making this test for the second and subsequent header lines if
5657 there were newlines in the string. This meant that an invalid header could
5658 be inserted if Exim was badly configured.
5660 7. Allow an ACL "warn" statement to add header lines at the start or after all
5661 the Received: headers, as well as at the end.
5663 8. Added the rcpt_4xx retry error code.
5665 9. Added postmaster_mailfrom=xxx to callout verification option.
5667 10. Added mailfrom=xxxx to the callout verification option, for verify=
5670 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5671 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5673 12. Inserted some casts to stop certain compilers warning when using pointer
5674 differences as field lengths or precisions in printf-type calls (mostly
5675 affecting debugging statements).
5677 13. Added optional readline() support for -be (dynamically loaded).
5679 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5680 same clock tick as a message's arrival, so that its received time was the
5681 same as the "first fail" time on the retry record, and that message
5682 remained on the queue past the ultimate address timeout, every queue runner
5683 would try a delivery (because it was past the ultimate address timeout) but
5684 after another failure, the ultimate address timeout, which should have then
5685 bounced the address, did not kick in. This was a "< instead of <=" error;
5686 in most cases the first failure would have been in the next clock tick
5687 after the received time, and all would be well.
5689 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5690 being recognized when the domain list was tested by the match_domain
5691 condition in an expansion string.
5693 16. Added the ${str2b64: operator.
5695 17. Exim was always calling setrlimit() to set a large limit for the number of
5696 processes, without checking whether the existing limit was already
5697 adequate. (It did check for the limit on file descriptors.) Furthermore,
5698 errors from getrlimit() and setrlimit() were being ignored. Now they are
5699 logged to the main and panic logs, but Exim does carry on, to try to do its
5700 job under whatever limits there are.
5702 18. Imported PCRE 5.0.
5704 19. Trivial typo in log message " temporarily refused connection" (the leading
5707 20. If the log selector return_path_on_delivery was set and an address was
5708 redirected to /dev/null, the delivery process crashed because it assumed
5709 that a return path would always be set for a "successful" delivery. In this
5710 case, the whole delivery is bypassed as an optimization, and therefore no
5713 21. Internal re-arrangement: the function for sending a challenge and reading
5714 a response while authentication was assuming a zero-terminated challenge
5715 string. It's now changed to take a pointer and a length, to allow for
5716 binary data in such strings.
5718 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5720 23. Exim was not respecting finduser_retries when seeking the login of the
5721 uid under which it was called; it was always trying 10 times. (The default
5722 setting of finduser_retries is zero.) Also, it was sleeping after the final
5723 failure, which is pointless.
5725 24. Implemented tls_on_connect_ports.
5727 25. Implemented acl_smtp_predata.
5729 26. If the domain in control=submission is set empty, Exim assumes that the
5730 authenticated id is a complete email address when it generates From: or
5731 Sender: header lines.
5733 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5734 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5735 chown and chgrp in /bin and hostname in /usr/bin.
5737 28. Exim was keeping the "process log" file open after each use, just as it
5738 does for the main log. This opens the possibility of it remaining open for
5739 long periods when the USR1 signal hits a daemon. Occasional processlog
5740 errors were reported, that could have been caused by this. Anyway, it seems
5741 much more sensible not to leave this file open at all, so that is what now
5744 29. The long-running daemon process does not normally write to the log once it
5745 has entered its main loop, and it closes the log before doing so. This is
5746 so that log files can straightforwardly be renamed and moved. However,
5747 there are a couple of unusual error situations where the daemon does write
5748 log entries, and I had neglected to close the log afterwards.
5750 30. The text of an SMTP error response that was received during a remote
5751 delivery was being truncated at 512 bytes. This is too short for some of
5752 the long messages that one sometimes sees. I've increased the limit to
5755 31. It is now possible to make retry rules that apply only when a message has a
5756 specific sender, in particular, an empty sender.
5758 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5759 it possible to be selective about when SMTP synchronization is enforced.
5761 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5763 32. Implemented hosts_connection_nolog.
5765 33. Added an ACL for QUIT.
5767 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5770 35. Added mailbox_size and mailbox_filecount to appendfile.
5772 36. Added control = no_multiline_responses to ACLs.
5774 37. There was a bug in the logic of the code that waits for the clock to tick
5775 in the case where the clock went backwards by a substantial amount such
5776 that the microsecond fraction of "now" was more than the microsecond
5777 fraction of "then" (but the whole seconds number was less).
5779 38. Added support for the libradius Radius client library this is found on
5780 FreeBSD (previously only the radiusclient library was supported).
5786 1. When certain lookups returned multiple values in the form name=value, the
5787 quoting of the values was not always being done properly. Specifically:
5788 (a) If the value started with a double quote, but contained no whitespace,
5790 (b) If the value contained whitespace other than a space character (i.e.
5791 tabs or newlines or carriage returns) it was not quoted.
5792 This fix has been applied to the mysql and pgsql lookups by writing a
5793 separate quoting function and calling it from the lookup code. The fix
5794 should probably also be applied to nisplus, ibase and oracle lookups, but
5795 since I cannot test any of those, I have not disturbed their existing code.
5797 2. A hit in the callout cache for a specific address caused a log line with no
5798 reason for rejecting RCPT. Now it says "Previous (cached) callout
5799 verification failure".
5801 3. There was an off-by-one bug in the queryprogram router. An over-long
5802 return line was truncated at 256 instead of 255 characters, thereby
5803 overflowing its buffer with the terminating zero. As well as fixing this, I
5804 have increased the buffer size to 1024 (and made a note to document this).
5806 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5807 when Exim is waiting for an SMTP response from a remote server, Exim
5808 restarts its select() call on the socket, thereby resetting its timeout.
5809 This is not a problem when such interrupts are rare. Somebody set up a cron
5810 job to run exiwhat every 2 minutes, which is less than the normal select()
5811 timeout (5 or 10 minutes). This meant that the select() timeout never
5812 kicked in because it was always reset. I have fixed this by comparing the
5813 time when an interrupt arrives with the time at the start of the first call
5814 to select(). If more time than the timeout has elapsed, the interrupt is
5815 treated as a timeout.
5817 5. Some internal re-factoring in preparation for the addition of Sieve
5818 extensions (by MH). In particular, the "personal" test is moved to a
5819 separate function, and given an option for scanning Cc: and Bcc: (which is
5820 not set for Exim filters).
5822 6. When Exim created an email address using the login of the caller as the
5823 local part (e.g. when creating a From: or Sender: header line), it was not
5824 quoting the local part when it contained special characters such as @.
5826 7. Installed new OpenBSD configuration files.
5828 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5829 try to make them clearer.
5831 9. Callout options, other than the timeout value, were being ignored when
5832 verifying sender addresses in header lines. For example, when using
5834 verify = header_sender/callout=no_cache
5836 the cache was (incorrectly) being used.
5838 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5839 only the Cygwin environment.
5841 11. When return_path_on_delivery was set as a log selector, if different remote
5842 addresses in the same message used different return paths and parallel
5843 remote delivery occurred, the wrong values would sometimes be logged.
5844 (Whenever a remote delivery process finished, the return path value from
5845 the most recently started remote delivery process was logged.)
5847 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5848 lines when AUTH and/or TLS are in use. This is the "received protocol"
5849 field. Exim used to use "asmtp" for authenticated SMTP, without any
5850 indication (in the protocol name) for TLS use. Now it follows the RFC and
5851 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5852 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5853 names appear in log lines as well as in Received: header lines.
5855 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5856 extensions, and comparison tests, and to fix some bugs.
5858 14. Changes to the "personal" filter test:
5860 (1) The test was buggy in that it was just doing the equivalent of
5861 "contains" tests on header lines. For example, if a user's address was
5862 anne@some.where, the "personal" test would incorrectly be true for
5864 To: susanne@some.where
5866 This test is now done by extracting each address from the header in turn,
5867 and checking the entire address. Other tests that are part of "personal"
5868 are now done using regular expressions (for example, to check local parts
5869 of addresses in From: header lines).
5871 (2) The list of non-personal local parts in From: addresses has been
5872 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5873 taken from the Sieve specification recommendations.
5875 (3) If the message contains any header line starting with "List-" it is
5876 treated as non-personal.
5878 (4) The test for "circular" in the Subject: header line has been removed
5879 because it now seems ill-conceived.
5881 15. Minor typos in src/EDITME comments corrected.
5883 16. Installed latest exipick from John Jetmore.
5885 17. If headers_add on a router specified a text string that was too long for
5886 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5887 of string_sprintf() is now avoided.
5889 18. $message_body_size was not set (it was always zero) when running the DATA
5890 ACL and the local_scan() function.
5892 19. For the "mail" command in an Exim filter, no default was being set for
5893 the once_repeat time, causing a random time value to be used if "once" was
5894 specified. (If the value happened to be <= 0, no repeat happened.) The
5895 default is now 0s, meaning "never repeat". The "vacation" command was OK
5896 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5897 (I found it when inspecting the code).
5899 20. There is now an overall timeout for performing a callout verification. It
5900 defaults to 4 times the callout timeout, which applies to individual SMTP
5901 commands during the callout. The overall timeout applies when there is more
5902 than one host that can be tried. The timeout is checked before trying the
5903 next host. This prevents very long delays if there are a large number of
5904 hosts and all are timing out (e.g. when the network connections are timing
5905 out). The value of the overall timeout can be changed by specifying an
5906 additional sub-option for "callout", called "maxwait". For example:
5908 verify = sender/callout=5s,maxwait=20s
5910 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5911 to the end before writing, but this should make it even safer).
5913 22. Exim was forgetting that it had advertised PIPELINING for the second and
5914 subsequent messages on an SMTP connection. It was also not resetting its
5915 memory on STARTTLS and an internal HELO.
5917 23. When Exim logs an SMTP synchronization error within a session, it now
5918 records whether PIPELINING has been advertised or not.
5920 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5921 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5922 rather than long int.
5924 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5926 26. Added the never_mail option to autoreply.
5932 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5933 crash if the getsockname() call failed; this can happen if a connection is
5934 closed very soon after it is established. The problem was simply in the
5935 order in which certain operations were done, causing Exim to try to write
5936 to the SMTP stream before it had set up the file descriptor. The bug has
5937 been fixed by making things happen in the correct order.
5943 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5944 before the connection was closed, thus losing the rejection response.
5946 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5947 some early Solaris releases, but causes trouble in current releases where
5948 socklen_t is defined.
5950 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5953 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5956 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5957 that releasing the top of it at the end releases what was used for sub-
5958 expansions (unless the block got too big). However, discard this block if
5959 the first thing is a variable or header, so that we can use its block when
5960 it is dynamic (useful for very large $message_headers, for example).
5962 6. Lookups now cache *every* query, not just the most recent. A new, separate
5963 store pool is used for this. It can be recovered when all lookup caches are
5964 flushed. Lookups now release memory at the end of their result strings.
5965 This has involved some general refactoring of the lookup sources.
5967 7. Some code has been added to the store_xxx() functions to reduce the amount
5968 of flapping under certain conditions.
5970 8. log_incoming_interface used to affect only the <= reception log lines. Now
5971 it causes the local interface and port to be added to several more SMTP log
5972 lines, for example "SMTP connection from", and rejection lines.
5974 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5976 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5978 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5979 could be overwritten at the end of the current message (or the start of a
5980 new message if it was set in a HELO ACL). The value is now preserved for
5981 the duration of the SMTP connection.
5983 12. If a transport had a headers_rewrite setting, and a matching header line
5984 contained an unqualified address, that address was qualified, even if it
5985 did not match any rewriting rules. The underlying bug was that the values
5986 of the flags that permit the existence of unqualified sender and recipient
5987 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5988 non-local messages, and by -bnq for local messages) were not being
5989 preserved with the message after it was received.
5991 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5992 "next input=" as part of the text comprising the host identity instead of
5993 the correct text. The code was using the same buffer for two different
5994 strings. However, depending on which order the printing function evaluated
5995 its arguments, the bug did not always show up. Under Linux, for example, my
5996 test suite worked just fine.
5998 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5999 doesn't work with some older versions of Perl. It has been changed to "my",
6000 which in any case is probably the better facility to use.
6002 15. A really picky compiler found some instances of statements for creating
6003 error messages that either had too many or two few arguments for the format
6006 16. The size of the buffer for calls to the DNS resolver has been increased
6007 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6008 for addresses that have a lot of PTR records. This alleviates a problem; it
6009 does not fully solve it.
6011 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6012 buffer now truncates the list and logs the incident, which is the same
6013 action as happens when Exim is looking up a host name and its aliases.
6014 Previously in this situation something unpredictable would happen;
6015 sometimes it was "internal error: store_reset failed".
6017 18. If a server dropped the connection unexpectedly when an Exim client was
6018 using GnuTLS and trying to read a response, the client delivery process
6019 crashed while trying to generate an error log message.
6021 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6022 string, for example:
6024 warn message = H1: something\nH2: something
6026 the text was added as a single header line from Exim's point of view
6027 though it ended up OK in the delivered message. However, searching for the
6028 second and subsequent header lines using $h_h2: did not work. This has been
6029 fixed. Similarly, if a system filter added multiple headers in this way,
6030 the routers could not see them.
6032 20. Expanded the error message when iplsearch is called with an invalid key to
6033 suggest using net-iplsearch in a host list.
6035 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6036 ACLs are no longer actually imposed (and a message to that effect is
6039 22. If a "gecos" field in a passwd entry contained escaped characters, in
6040 particular, if it contained a \" sequence, Exim got it wrong when building
6041 a From: or a Sender: header from that name. A second bug also caused
6042 incorrect handling when an unquoted " was present following a character
6043 that needed quoting.
6045 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6046 was not being matched caselessly.
6048 24. Arranged for all hyphens in the exim.8 source to be escaped with
6051 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6052 properly. Recipient callout cache records were still being keyed to include
6053 the sender, even when use_sender was set false. This led to far more
6054 callouts that were necessary. The sender is no longer included in the key
6055 when use_sender is false.
6057 26. Added "control = submission" modifier to ACLs.
6059 27. Added the ${base62d: operator to decode base 62 numbers.
6061 28. dnsdb lookups can now access SRV records.
6063 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6064 the configuration file.
6066 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6067 (-v) mode. This makes the output for a verbose queue run more intelligible.
6069 31. Added a use_postmaster feature to recipient callouts.
6071 32. Added the $body_zerocount variable, containing the number of binary zero
6072 bytes in the message body.
6074 33. The time of last modification of the "new" subdirectory is now used as the
6075 "mailbox time last read" when there is a quota error for a maildir
6078 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6080 35. Added +ignore_unknown as a special item in host lists.
6082 36. Code for decoding IPv6 addresses in host lists is now included, even if
6083 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6084 address was recognized as an IP address, but was then not correctly decoded
6085 into binary, causing unexpected and incorrect effects when compared with
6092 1. Very minor rewording of debugging text in manualroute to say "list of
6093 hosts" instead of "hostlist".
6095 2. If verify=header_syntax was set, and a header line with an unqualified
6096 address (no domain) and a large number of spaces between the end of the
6097 name and the colon was received, the reception process suffered a buffer
6098 overflow, and (when I tested it) crashed. This was caused by some obsolete
6099 code that should have been removed. The fix is to remove it!
6101 3. When running in the test harness, delay a bit after writing a bounce
6102 message to get a bit more predictability in the log output.
6104 4. Added a call to search_tidyup() just before forking a reception process. In
6105 theory, someone could use a lookup in the expansion of smtp_accept_max_
6106 per_host which, without the tidyup, could leave open a database connection.
6108 5. Added the variables $recipient_data and $sender_data which get set from a
6109 lookup success in an ACL "recipients" or "senders" condition, or a router
6110 "senders" option, similar to $domain_data and $local_part_data.
6112 6. Moved the writing of debug_print from before to after the "senders" test
6115 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6116 problems for message scanning, either using a data ACL, or using
6117 local_scan() because the Received: header was not generated till after they
6118 were called (in order to set the time as the time of reception completion).
6119 I have revised the way this works. The header is now generated after the
6120 body is received, but before the ACL or local_scan() are called. After they
6121 are run, the timestamp in the header is updated.
6127 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6128 before starting a queue runner without re-exec. This happened only when
6129 deliver_drop_privilege was set or when the Exim user was set to root. The
6130 effect of the bug was that timeouts during subsequent deliveries caused
6131 crashes instead of being properly handled. The handler is now left at its
6132 default (and expected) setting.
6134 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6135 message, again when deliver_drop_privilege is set or Exim is run as root.
6136 The bug described in (1) was not present in this case, but the tidying up
6137 of the other signals was missing. I have made the two cases consistent.
6139 3. The ignore_target_hosts setting on a manualroute router was being ignored
6140 for hosts that were looked up using the /MX notation.
6142 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6145 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6146 operated on the sender address. After changing the $sender_address to <>
6147 for the sender address verify, Exim was re-instated it as the original
6148 (before rewriting) address, but remembering that it had rewritten it, so it
6149 wasn't rewriting it again. This bug also had the effect of breaking the
6150 sender address verification caching when the sender address was rewritten.
6152 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6153 This has been changed so that if the ip literal address matches
6154 ignore_target_hosts, the router declines.
6156 7. Added expansion conditions match_domain, match_address, and match_local_
6157 part (NOT match_host).
6159 8. The placeholder for the Received: header didn't have a length field set.
6161 9. Added code to Exim itself and to exim_lock to test for a specific race
6162 condition that could lead to file corruption when using MBX delivery. The
6163 issue is with the lockfile that is created in /tmp. If this file is removed
6164 after a process has opened it but before that process has acquired a lock,
6165 there is the potential for a second process to recreate the file and also
6166 acquire a lock. This could lead to two Exim processes writing to the file
6167 at the same time. The added code performs the same test as UW imapd; it
6168 checks after acquiring the lock that its file descriptor still refers to
6169 the same named file.
6171 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6172 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6175 11. Added the smtp_active_hostname option. If used, this will typically be made
6176 to depend on the incoming interface address. Because $interface_address is
6177 not set up until the daemon has forked a reception process, error responses
6178 that can happen earlier (such as "too many connections") no longer contain
6181 12. If an expansion in a condition on a "warn" statement fails because a lookup
6182 defers, the "warn" statement is abandoned, and the next ACL statement is
6183 processed. Previously this caused the whole ACL to be aborted.
6185 13. Added the iplsearch lookup type.
6187 14. Added ident_timeout as a log selector.
6189 15. Added tls_certificate_verified as a log selector.
6191 16. Added a global option tls_require_ciphers (compare the smtp transport
6192 option of the same name). This controls incoming TLS connections.
6194 17. I finally figured out how to make tls_require_ciphers do a similar thing
6195 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6196 before starting the TLS session.
6198 18. Tabs are now shown as \t in -bP output.
6200 19. If the log selector return_path_on_delivery was set, Exim crashed when
6201 bouncing a message because it had too many Received: header lines.
6203 20. If two routers both had headers_remove settings, and the first one included
6204 a superfluous trailing colon, the final name in the first list and the
6205 first name in the second list were incorrectly joined into one item (with a
6206 colon in the middle).
6212 1. Added -C and -D options to the exinext utility, mainly to make it easier
6213 to include in the automated testing, but these could be helpful when
6214 multiple configurations are in use.
6216 2. The exinext utility was not formatting the output nicely when there was
6217 an alternate port involved in the retry record key, nor when there was a
6218 message id as well (for retries that were specific to a specific message
6219 and a specific host). It was also confused by IPv6 addresses, because of
6220 the additional colons they contain. I have fixed the IPv4 problem, and
6221 patched it up to do a reasonable job for IPv6.
6223 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6224 delivery, the log line now contains "pipelined" if PIPELINING was used.
6226 4. An SMTP transport process used to panic and die if the bind() call to set
6227 an explicit outgoing interface failed. This has been changed; it is now
6228 treated in the same way as a connect() failure.
6230 5. A reference to $sender_host_name in the part of a conditional expansion
6231 that was being skipped was still causing a DNS lookup. This no longer
6234 6. The def: expansion condition was not recognizing references to header lines
6235 that used bh_ and bheader_.
6237 7. Added the _cache feature to named lists.
6239 8. The code for checking quota_filecount in the appendfile transport was
6240 allowing one more file than it should have been.
6242 9. For compatibility with Sendmail, the command line option
6250 and sets the incoming protocol and host name (for trusted callers). The
6251 host name and its colon can be omitted when only the protocol is to be set.
6252 Note the Exim already has two private options, -pd and -ps, that refer to
6253 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6254 "s", but I don't think that's a major issue.
6256 10. A number of refactoring changes to the code, none of which should affect
6259 (a) The number of logging options was getting close to filling up the
6260 32-bit word that was used as a bit map. I have split them into two classes:
6261 those that are passed in the argument to log_write(), and those that are
6262 only ever tested independently outside of that function. These are now in
6263 separate 32-bit words, so there is plenty of room for expansion again.
6264 There is no change in the user interface or the logging behaviour.
6266 (b) When building, for example, log lines, the code previously used a
6267 macro that called string_cat() twice, in order to add two strings. This is
6268 not really sufficiently general. Furthermore, there was one instance where
6269 it was actually wrong because one of the argument was used twice, and in
6270 one call a function was used. (As it happened, calling the function twice
6271 did not affect the overall behaviour.) The macro has been replaced by a
6272 function that can join an arbitrary number of extra strings onto a growing
6275 (c) The code for expansion conditions now uses a table and a binary chop
6276 instead of a serial search (which was left over from when there were very
6277 few conditions). Also, it now recognizes conditions like "pam" even when
6278 the relevant support is not compiled in: a suitably worded error message is
6279 given if an attempt is made to use such a condition.
6281 11. Added ${time_interval:xxxxx}.
6283 12. A bug was causing one of the ddress fields not to be passed back correctly
6284 from remote delivery subprocesses. The field in question was not being
6285 subsequently used, so this caused to problems in practice.
6287 13. Added new log selectors queue_time and deliver_time.
6289 14. Might have fixed a bug in maildirsizefile handling that threw up
6290 "unexpected character" debug warnings, and recalculated the data
6291 unnecessarily. In any case, I expanded the warning message to give more
6294 15. Added the message "Restricted characters in address" to the statements in
6295 the default ACL that block characters like @ and % in local parts.
6297 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6298 Three changes have been made:
6300 (a) There was a serious bug; a negative response to MAIL caused the whole
6301 recipient domain to be cached as invalid, thereby blocking all messages
6302 to all local parts at the same domain, from all senders. This bug has
6303 been fixed. The domain is no longer cached after a negative response to
6304 MAIL if the sender used is not empty.
6306 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6309 (c) A new callout option, "use_sender" has been added for people who want
6310 the modified behaviour.
6316 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6319 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6320 indeed breaks things for older releases.
6322 3. Added additional logging to the case where there is a problem reading data
6323 from a filter that is running in a subprocess using a pipe, in order to
6324 try to track down a specific problem.
6326 4. Testing facility fudge: when running in the test harness and attempting
6327 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6328 getting "No route to host". Convert this to a timeout.
6330 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6333 6. Some OS don't have socklen_t but use size_t instead. This affects the
6334 fifth argument of getsockopt() amongst other things. This is now
6335 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6336 can be set for individual OS. I have set it for SunOS5, OSF1, and
6337 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6338 some earlier ones do not.
6340 7. Change 4.30/15 was not doing the test caselessly.
6342 8. The standard form for an IPv6 address literal was being rejected by address
6343 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6344 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6345 this, as well as the form without the "IPv6" on the front (but only when
6346 address literals are enabled, of course).
6348 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6350 10. Exim crashed if a message with an empty sender address specified by -f
6351 encountered a router with an errors_to setting. This could be provoked only
6352 by a command such as
6356 where an empty string was supplied; "<>" did not hit this bug.
6358 11. Installed PCRE release 4.5.
6360 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6361 remained set. It is now erased.
6363 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6364 times from message ids (which are base 36 rather than the normal 62).
6366 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6367 were being counted as actual protocol errors, and logged if the log
6368 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6369 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6370 and DATA following a set of rejected RCPTs do not count as protocol errors.
6371 In other words, Exim assumes they were pipelined, though this may not
6372 actually be the case. Of course, in all cases the client gets an
6373 appropriate error code.
6375 15. If a lookup fails in an ACL condition, a message about the failure may
6376 be available; it is used if testing the ACL cannot continue, because most
6377 such messages specify what the cause of the deferral is. However, some
6378 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6379 that caused an old message to be retained and used if a later statement
6380 caused a defer, replacing the real cause of the deferral.
6382 16. If an IP address had so many PTR records that the DNS lookup buffer
6383 was not large enough to hold them, Exim could crash while trying to process
6384 the truncated data. It now detects and logs this case.
6386 17. Further to 4.21/58, another change has been made: if (and only if) the
6387 first line of a message (the first header line) ends with CRLF, a bare LF
6388 in a subsequent header line has a space inserted after it, so as not to
6389 terminate the header.
6391 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6392 unnecessarily, used atoi() instead of strtol(), and didn't check the
6393 termination when getting file sizes from file names by regex.
6395 19. Completely re-implemented the support for maildirsize files, in the light
6396 of a number of problems with the previous contributed implementation
6397 (4.30/29). In particular:
6399 . If the quota is zero, the maildirsize file is maintained, but no quota is
6402 . If the maildir directory does not exist, it is created before any attempt
6403 to write a maildirsize file.
6405 . The quota value in the file is just a cache; if the quota is changed in
6406 the transport, the new value overrides.
6408 . A regular expression is available for excluding directories from the
6411 20. The autoreply transport checks the characters in options that define the
6412 message's headers; it allows continued headers, but it was checking with
6413 isspace() after an embedded newline instead of explicitly looking for a
6416 21. If all the "regular" hosts to which an address was routed had passed their
6417 expiry times, and had not reached their retry times, the address was
6418 bounced, even if fallback hosts were defined. Now Exim should go on to try
6421 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6422 equivalent code in the SMTP transport. Some hosts send humungous responses
6423 to HELO/EHLO, more than 1024 it seems.
6425 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6426 gives compiler warnings in some environments. I've now done it "properly",
6429 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6430 (because of problems with the built-in one) was declared to return uschar *
6431 instead of char *, causing compiler failure.
6433 25. Fixed a file descriptor leak when processing alias/forward files.
6435 26. Fixed a minor format string issue in dbfn.c.
6437 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6439 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6440 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6443 29. When a sender address is verified, it is cached, to save repeating the test
6444 when there is more than one recipient in a message. However, when the
6445 verification involves a callout, it is possible for different callout
6446 options to be set for different recipients. It is too complicated to keep
6447 track of this in the cache, so now Exim always runs a verification when a
6448 callout is required, relying on the callout cache for the optimization.
6449 The overhead is duplication of the address routing, but this should not be
6452 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6453 to be verified with callout=postmaster, and the main callout worked but the
6454 postmaster check failed, the verification correctly failed. However, if a
6455 subsequent RCPT command asked for sender verification *without* the
6456 postmaster check, incorrect caching caused this verification also to fail,
6459 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6460 it was not caching the DNS options (qualify_single, search_parents) that
6461 were used when the lookup failed. A subsequent lookup with different
6462 options therefore always gave the same answer, though there were cases
6463 where it should not have. (Example: a "domains = !$mx_any" option on a
6464 dnslookup router: the "domains" option is always processed without any
6465 widening, but the router might have qualify_single set.) Now Exim uses the
6466 cached value only when the same options are set.
6468 32. Added John Jetmore's "exipick" utility to the distribution.
6470 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6471 than a timeout (e.g. a certificate is required, and is not provided), an
6472 Exim server now closes the connection immediately. Previously it waited for
6473 the client to close - but if the client is SSL, it seems that they each
6474 wait for each other, leading to a delay before one of them times out.
6476 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6477 maintained 0.8.x compatibility because I don't think many are using it, and
6478 it is clearly obsolete.
6480 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6483 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6484 client certificate was expired. A simple patch fixes this, though I don't
6485 understand the full logic of why the verify callback is called multiple
6488 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6489 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6490 which causes problems with some clients (such as the Certicom SSL Plus
6491 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6492 disables the coutermeasure allowing Eudora to connect."
6494 38. Exim was not checking that a write() to a log file succeeded. This could
6495 lead to Bad Things if a log got too big, in particular if it hit a file
6496 size limit. Exim now panics and dies if it cannot write to a log file, just
6497 as it does if it cannot open a log file.
6499 39. Modified OS/Makefile-Linux so that it now contains
6501 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6503 The two -D definitions ensure that Exim is compiled with large file
6504 support, which makes it possible to handle log files that are bigger than
6507 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6508 instance) a domain was checked against a named list that involved a lookup,
6509 causing $domain_data to be set, then another domain was checked against the
6510 same list, then the first domain was re-checked, the value of $domain_data
6511 after the final check could be wrong. In particular, if the second check
6512 failed, it could be set empty. This bug probably also applied to
6515 41. The strip_trailing_dot option was not being applied to the address given
6516 with the -f command-line option.
6518 42. The code for reading a message's header from the spool was incrementing
6519 $received_count, but never initializing it. This meant that the value was
6520 incorrect (doubled) while delivering a message in the same process in which
6521 it was received. In the most common configuration of Exim, this never
6522 happens - a fresh exec is done - but it can happen when
6523 deliver_drop_privilege is set.
6525 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6526 it now includes up to 150 characters of the unexpected data in the log
6529 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6530 and building data strings. The size of both of these buffers was 10 000
6531 bytes - far larger than anybody would *ever* want, thought I. Needless to
6532 say, somebody hit the limit. I have increased the maximum line length to
6533 20 000 and the maximum data length of concatenated lines to 100 000. I have
6534 also fixed two bugs, because there was no checking on these buffers. Tsk,
6535 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6536 buffer is too small.
6538 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6539 lsearch lookups. Now it does.
6541 46. When parsing a route_list item in a manualroute router, a fixed-length
6542 buffer was used for the list of hosts. I made this 1024 bytes long,
6543 thinking that nobody would ever have a list of hosts that long. Wrong.
6544 Somebody had a whole pile of complicated expansion conditions, and the
6545 string was silently truncated, leading to an expansion error. It turns out
6546 that it is easier to change to an unlimited length (owing to other changes
6547 that have happened since this code was originally written) than to build
6548 structure for giving a limitation error. The length of the item that
6549 expands into the list of hosts is now unlimited.
6551 47. The lsearch lookup could not handle data where the length of text line was
6552 more than 4095 characters. Such lines were truncated, leading to shortened
6553 data being returned. It should now handle lines of any length.
6555 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6556 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6559 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6560 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6561 no longer happens in comment lines. A list of replacements is now placed
6562 at the head of all of the source files, except those whose only change is
6563 to replace PERL_COMMAND in the very first #! line.
6565 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6566 messages on the queue, with a bottom-up merge sort, using code contributed
6567 by Michael Haardt. This should make operations like -bp somewhat faster on
6568 large queues. It won't affect queue runners, except when queue_run_in_order
6571 51. Installed eximstats 1.31 in the distribution.
6573 52. Added support for SRV lookups to the dnslookup router.
6575 53. If an ACL referred to $message_body or $message_body_end, the value was not
6576 reset for any messages that followed in the same SMTP session.
6578 54. The store-handling optimization for building very long strings was not
6579 differentiating between the different store pools. I don't think this
6580 actually made any difference in practice, but I've tidied it.
6582 55. While running the routers to verify a sender address, $sender_address
6583 was still set to the sender address. This is wrong, because when routing to
6584 send a bounce to the sender, it would be empty. Therefore, I have changed
6585 it so that, while verifying a sender address, $sender_address is set to <>.
6586 (There is no change to what happens when verifying a recipient address.)
6588 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6589 target A or AAAA records (if not already returned) without resetting the
6590 qualify_single or search_parents options of the DNS resolver. These are
6591 inappropriate in this case because the targets of MX and SRV records must
6592 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6593 target that, when qualified, matched something in the local domain. These
6594 two options are now turned off when doing these lookups.
6596 57. It seems that at least some releases of Reiserfs (which does not have the
6597 concept of a fixed number of inodes) returns zero and not -1 for the
6598 number of available inodes. This interacted badly with check_spool_inodes,
6599 which assumed that -1 was the "no such thing" setting. What I have done is
6600 to check that the total number of inodes is greater than zero before doing
6601 the test of how many are available.
6603 58. When a "warn" ACL statement has a log_message modifier, the message is
6604 remembered, and not repeated. This is to avoid a lot of repetition when a
6605 message has many recipients that cause the same warning to be written.
6606 However, Exim was preserving the list of already written lines for an
6607 entire SMTP session, which doesn't seem right. The memory is now reset if a
6608 new message is started.
6610 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6611 debugging output unless log_rewrite was also set.
6613 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6614 of (int)(handle) when we know that handle contains (void *)(-1).
6616 61. The Exim daemon panic-logs an error return when it closes the incoming
6617 connection. However "connection reset by peer" seems to be common, and
6618 isn't really an error worthy of noting specially, so that particular error
6621 62. When Exim is trying to find all the local interfaces, it used to panic and
6622 die if the ioctl to get the interface flags failed. However, it seems that
6623 on at least one OS (Solaris 9) it is possible to have an interface that is
6624 included in the list of interfaces, but for which you get a failure error
6625 for this call. This happens when the interface is not "plumbed" into a
6626 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6627 failure of the "get flags" call assumes that the interface is down.
6629 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6630 makes life easier for people who are doing arithmetic on fields extracted
6631 from dates, where you often get leading zeros that should not be
6632 interpreted as octal.
6634 64. Added qualify_domain to the redirect router, to override the global
6637 65. If a pathologically long header line contained very many addresses (the
6638 report of this problem mentioned 10 000) and each of them was rewritten,
6639 Exim could use up a very large amount of memory. (It kept on making new
6640 copies of the header line as it rewrote, and never released the old ones.)
6641 At the expense of a bit more processing, the header rewriting function has
6642 been changed so that it no longer eats memory in this way.
6644 66. The generation of the Received: header has been moved from the time that a
6645 message starts to be received, to the time that it finishes. The timestamp
6646 in the Received: header should now be very close to that of the <= log
6647 line. There are two side-effects of this change:
6649 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6650 logged header lines no longer include the local Received: line, because
6651 it has not yet been created. The same applies to a copy of the message
6652 that is returned to a non-SMTP sender when a message is rejected.
6654 (b) When a filter file is tested using -bf, no additional Received: header
6655 is added to the test message. After some thought, I decided that this
6658 This change does not affect the value of $received_for. It is still set
6659 after address rewriting, but before local_scan() is called.
6661 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6663 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6664 gave an unhelpful panic error message, and a defer error. I have managed to
6665 change this behaviour so that it now rejects any supplied certificate,
6666 which seems right, as the list of acceptable certificates is empty.
6668 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6669 gave an unhelpful defer error. I have not managed to make this reject any
6670 supplied certificates, but the error message it gives is "no certificate
6671 supplied", which is not helpful.
6673 70. exigrep's output now also includes lines that are not associated with any
6674 message, but which match the given pattern. Implemented by a patch from
6675 Martin Sluka, which also tidied up the Perl a bit.
6677 71. Recipient callout verification, like sender verification, was using <> in
6678 the MAIL FROM command. This isn't really the right thing, since the actual
6679 sender may affect whether the remote host accepts the recipient or not. I
6680 have changed it to use the actual sender in the callout; this means that
6681 the cache record is now keyed on a recipient/sender pair, not just the
6682 recipient address. There doesn't seem to be a real danger of callout loops,
6683 since a callout by the remote host to check the sender would use <>.
6684 [SEE ABOVE: changed after hitting problems.]
6686 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6687 temporary errors. However, in the case of such a code being given after
6688 the end of a data transmission (i.e. after ".") Exim was failing to write
6689 a retry record for the message. (Yes, there was some broken host that was
6690 actually sending 8xx at this point.)
6692 73. An unknown lookup type in a host list could cause Exim to panic-die when
6693 the list was checked. (An example that provoked this was putting <; in the
6694 middle of a list instead of at the start.) If this happened during a DATA
6695 ACL check, a -D file could be left lying around. This kind of configuration
6696 error no longer causes Exim to die; instead it causes a defer error. The
6697 incident is still logged to the main and panic logs.
6699 74. Buglet left over from Exim 3 conversion. The message "too many messages
6700 in one connection" was written to the rejectlog but not the mainlog, except
6701 when address rewriting (yes!) was being logged.
6703 75. Added write_rejectlog option.
6705 76. When a system filter was run not as root (that is, when system_filter_user
6706 was set), the values of the $n variables were not being returned to the
6707 main process; thus, they were not subsequently available in the $sn
6710 77. Added +return_path_on_delivery log selector.
6712 78. A connection timeout was being treated differently from recipients deferred
6713 when testing hosts_max_try with a message that was older than the host's
6714 retry timeout. (The host should not be counted, thus allowing all hosts to
6715 be tried at least once before bouncing.) This may have been the cause of an
6716 occasionally reported bug whereby a message would remain on the queue
6717 longer than the retry timeout, but would be bounced if a delivery was
6718 forced. I say "may" because I never totally pinned down the problem;
6719 setting up timeout/retry tests is difficult. See also the next item.
6721 79. The ultimate address timeout was not being applied to errors that involved
6722 a combination of host plus message (for example, a timeout on a MAIL
6723 command). When an address resolved to a number of possible hosts, and they
6724 were not all tried for each delivery (e.g. because of hosts_max_try), a
6725 message could remain on the queue longer than the retry timeout.
6727 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6730 81. Fixed an obscure SMTP outgoing bug which required at least the following
6731 conditions: (a) there was another message waiting for the same server;
6732 (b) the server returned 5xx to all RCPT commands in the first message so
6733 that the message was not completed; (c) the server dropped the connection
6734 or gave a negative response to the RSET that Exim sends to abort the
6735 transaction. The observed case was a dropped connection after DATA that had
6736 been sent in pipelining mode. That is, the server had advertised PIPELINING
6737 but was not implementing it correctly. The effect of the bug was incorrect
6738 behaviour, such as trying another host, and this could lead to a crash.
6744 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6745 and daemon.c were passed as pointers to ints; they should have been
6746 pointers to socklen_t variables (which are typically unsigned ints).
6748 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6751 3. Fixed a really odd bug that affected only the testing scheme; patching a
6752 certain fixed string in the binary changed the value of another string that
6753 happened to be identical to the end of the original first string.
6755 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6756 name", it returns that address as the IP address. On some operating
6757 systems (e.g. Solaris), it also passes back the IP address string as the
6758 "host name". However, on others (e.g. Linux), it passes back an empty
6759 string. Exim wasn't checking for this, and was changing the host name to an
6760 empty string, assuming it had been canonicalized.
6762 5. Although rare, it is permitted to have more than one PTR record for a given
6763 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6764 all the names associated with an address, because they do in Solaris.
6765 However, it seems that they do not in Linux for data that comes from the
6766 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6767 I found this out when I moved to a new Linux workstation and tried to run
6768 the Exim test suite.
6770 To get round this problem I have changed the code so that it now does its
6771 own call to the DNS to look up PTR records when searching for a host name.
6772 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6773 addresses that are only in /etc/hosts are still found.
6775 This behaviour is, however, controlled by an option called host_lookup_
6776 order, which defaults to "bydns:byaddr". If people want to use the other
6777 order, or indeed, just use one or the other means of lookup, they can
6778 specify it in this variable.
6780 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6781 some operating systems, this comes back from gethostbyaddr() as an empty
6782 string, and this is what Exim used to test for. However, it seems that in
6783 other systems, "." is yielded. Exim now tests for this case too.
6785 7. The values of check_spool_space and check_log_space are now held internally
6786 as a number of kilobytes instead of an absolute number of bytes. If a
6787 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6788 kilobyte. This means that much larger values can be stored.
6790 8. Exim monitor: an attempt to get the action menu when not actually pointing
6791 at a message produces an empty menu entitled "No message selected". This
6792 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6793 no entries in it ("Shell widget menu has zero width and/or height"). So I
6794 have added a single, blank menu entry in this case.
6796 9. Added ${quote_local_part.
6798 10. MIME decoding is now applied to the contents of Subject: header lines when
6801 11. Now that a reference to $sender_host_address automatically causes a reverse
6802 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6803 host lookup before query-style lookups in lists that might use this
6804 variable. This has therefore been abolished, and the "net-" prefix is no
6805 longer necessary for query-style lookups.
6807 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6808 have been a typo for LFLAGS, so it has been changed.
6810 13. The install script calls Exim with "-C /dev/null" in order to find the
6811 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6812 to be output. However, since Exim outputs its version number before the
6813 error, it didn't break the script. It just looked ugly. I fixed this by
6814 always allowing "-C /dev/null" if the caller is root.
6816 14. Ignore overlarge ACL variable number when reading spool file - insurance
6817 against a later release with more variables having written the file.
6819 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6820 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6821 this, as well as the form without the "IPv6" on the front.
6823 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6824 OS/Makefile-Darwin file.
6826 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6827 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6829 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6830 variable. However, some OS use ints for inodes. I've added cast to long int
6831 to get rid of the compiler warning.
6833 19. I had forgotten to lock out "/../" in configuration file names when
6834 ALT_CONFIG_PREFIX was set.
6836 20. Routers used for verification do not need to specify transports. However,
6837 if such a router generated a host list, and callout was configured, Exim
6838 crashed, because it could not find a port number from the (non-existent)
6839 transport. It now assumes port 25 in this circumstance.
6841 21. Added the -t option to exigrep.
6843 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6844 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6845 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6846 them is set, LOOKUP_LSEARCH is forced.
6848 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6850 24. Added sender and host information to the "rejected by local_scan()" log
6851 line; previously there was no indication of these.
6853 25. Added .include_if_exists.
6855 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6856 Linux. It turns out that not all file systems support this. Apparently some
6857 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6858 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6859 not supported on the file descriptor, is now ignored when Exim is trying to
6860 sync a directory. This applies only to Linux.
6862 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6864 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6865 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6868 29. Added support for maildirsize files from supplied patch (modified a bit).
6870 30. The use of :fail: followed by an empty string could lead Exim to respond to
6871 sender verification failures with (e.g.):
6873 550 Verification failed for <xxx>
6874 550 Sender verify failed
6876 where the first response line was missing the '-' that indicates it is not
6877 the final line of the response.
6879 31. The loop for finding the name of the user that called Exim had a hardwired
6880 limit of 10; it now uses the value of finduser_retries, which is used for
6881 all other user lookups.
6883 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6886 33. Exim was neglecting to zero errno before one call of strtol() when
6887 expanding a string and expecting an integer value. On some systems this
6888 resulted in spurious "integer overflow" errors. Also, it was casting the
6889 result into an int without checking.
6891 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6892 did not work. The code looks as if it has *never* worked, though it appears
6893 to have been documented since at least release 1.62. I have made it work.
6895 35. The "timeout_DNS" error in retry rules, also documented since at least
6896 1.62, also never worked. As it isn't clear exactly what this means, and
6897 clearly it isn't a major issue, I have abolished the feature by treating it
6898 as "timeout", and writing a warning to the main and panic logs.
6900 36. The display of retry rules for -brt wasn't always showing the error code
6903 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6904 timeout_connect_A, timeout_connect_MX.
6906 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6907 to the empty sender.
6909 39. The daemon was not analysing the content of -oX till after it had closed
6910 stderr and disconnected from the controlling terminal. This meant that any
6911 syntax errors were only noted on the panic log, and the return code from
6912 the command was 0. By re-arranging the code a little, I've made the
6913 decoding happen first, so such errors now appear on stderr, and the return
6914 code is 1. However, the actual setting up of the sockets still happens in
6915 the disconnected process, so errors there are still only recorded on the
6918 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6919 connections (as happens on some IP stacks) was logged at start up time as
6920 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6921 it from "IPv6 and IPv4", which means that two separate sockets are being
6924 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6925 says whether AF_INET or AF_INET6 was passed as an argument.
6927 42. Exiwhat output was messed up when time zones were included in log
6930 43. Exiwhat now gives more information about the daemon's listening ports,
6931 and whether -tls-on-connect was used.
6933 44. The "port" option of the smtp transport is now expanded.
6935 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6936 silently ignored. Now an error message is written to the main and panic
6939 46. There's a new ACL modifier called "logwrite" which writes to a log file
6940 as soon as it is encountered.
6942 47. Added $local_user_uid and $local_user_gid at routing time.
6944 48. Exim crashed when trying to verify a sender address that was being
6947 49. Exim was recognizing only a space character after ".include". It now also
6948 recognizes a tab character.
6950 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6951 extracting the relevant information from the specification. The man page no
6952 longer contains scrambled data for the -d option, and I've added a section
6953 at the front about calling Exim under different names.
6955 51. Added "extra_headers" argument to the "mail" command in filter files.
6957 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6960 53. Installed eximstats 1.29.
6962 54. Added transport_filter_timeout as a generic transport option.
6964 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6965 Cc: header lines. This was required by RFC 822, but it not required by RFC
6968 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6969 incoming messages that did not have them. Now it does so only if the
6970 message originates locally, that is, if there is no associated remote host
6971 address. When Resent- header lines are present, this applies to the Resent-
6972 lines rather than the non-Resent- lines.
6974 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6975 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6977 58. Messages for configuration errors now include the name of the main
6978 configuration file - useful now that there may be more than one file in a
6979 list (.included file names were always shown).
6981 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6982 for those rare installations that do not start the daemon as root or run it
6983 setuid root. I've cut out the call to initgroups() if the daemon is not
6986 60. The Exim user and group can now be bound into the binary as text strings
6987 that are looked up at the start of Exim's processing.
6989 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6991 62. Added $mailstore_basename variable.
6993 63. Installed patch to sieve.c from Michael Haardt.
6995 64. When Exim failed to open the panic log after failing to open the main log,
6996 the original message it was trying to log was written to stderr and debug
6997 output, but if they were not available (the usual case in production), it
6998 was lost. Now it is written to syslog before the two lines that record the
6999 failures to open the logs.
7001 65. Users' Exim filters run in subprocesses under the user's uid. It is
7002 possible for a "deliver" command or an alias in a "personal" command to
7003 provoke an address rewrite. If logging of address rewriting is configured,
7004 this fails because the process is not running as root or exim. There may be
7005 a better way of dealing with this, but for the moment (because 4.30 needs
7006 to be released), I have disabled address rewrite logging when running a
7007 filter in a non-root, non-exim process.
7013 1. The buildconfig auxiliary program wasn't quoting the value set for
7014 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7015 not defined. This bug was masked in 4.22 by the effect that was fixed in
7018 2. Some messages that were rejected after a message id was allocated were
7019 shown as "incomplete" by exigrep. It no longer does this for messages that
7020 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7022 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7023 have allow_domain_literals set, the ID did not get logged in the <= line.
7024 Domain literals are now always recognized in Message-ID: header lines.
7026 4. The first argument for a ${extract expansion item is the key name or field
7027 number. Leading and trailing spaces in this item were not being ignored,
7028 causing some misleading effects.
7030 5. When deliver_drop_privilege was set, single queue runner processes started
7031 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7032 same command in the process it spins off) were not dropping privilege.
7034 6. When the daemon running as "exim" started a queue runner, it always
7035 re-executed Exim in the spun-off process. This is a waste of effort when
7036 deliver_drop_privilege is set. The new process now just calls the
7037 queue-runner function directly.
7043 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7046 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7047 ignored. Though the use of -oP was forcing the writing of a pid file, it
7048 was always written to the default place.
7050 3. If the message "no IP address found for host xxxx" is generated during
7051 incoming verification, it is now followed by identification of the incoming
7052 connection (so you can more easily find what provoked it).
7054 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7056 5. Added some features to "harden" Exim a bit more against certain attacks:
7058 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7059 be put in Local/Makefile. This is like the never_users runtime option,
7060 but it cannot be overridden. The default setting is "root".
7062 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7063 prefix string with which any file named in a -C command line option
7066 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7067 is retained for -C and -D only if the caller of Exim is root. Without
7068 it, the exim user may also use -C and -D and retain privilege.
7070 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7071 command line option is disabled.
7073 6. Macro names set by the -D option must start with an upper case letter, just
7074 like macro names defined in the configuration file.
7076 7. Added "dereference=" facility to LDAP.
7078 8. Two instances of the typo "uknown" in the source files are fixed.
7080 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7081 the Configure-Makefile script screwed up while processing it.
7083 10. Incorporated PCRE 4.4.
7085 11. The SMTP synchronization check was not operating right at the start of an
7086 SMTP session. For example, it could not catch a HELO sent before the client
7087 waited for the greeting. There is now a check for outstanding input at the
7088 point when the greeting is written. Because of the duplex, asynchronous
7089 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7090 way, but not yet received, when the check is performed.
7092 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7093 on TCP/IP sockets, because this apparently causes some broken clients to
7096 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7097 unchanged) from the Cygwin maintainer.
7099 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7100 was supported instead of testing for mbx. Effectively a typo.
7102 15. The spa authenticator server code was not checking that the input it
7103 received was valid base64.
7105 16. The debug output line for the "set" modifier in ACLs was not showing the
7106 name of the variable that was being set.
7108 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7110 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7111 lookup on its own. Something else was needed to trigger the lookup. For
7112 example, a match in host_lookup or the need for a host name in a host list.
7113 Now, if $sender_host_name is referenced and the host name has not yet been
7114 looked up, a lookup is performed. If the lookup fails, the variable remains
7115 empty, and $host_lookup_failed is set to "1".
7117 19. Added "eqi" as a case-independent comparison operator.
7119 20. The saslauthd authentication condition could segfault if neither service
7120 nor realm was specified.
7122 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7123 error message that was logged was misleading, and incoming SMTP
7124 connections were dropped. The message is now more accurate, and temporary
7125 errors are given to SMTP connections.
7127 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7128 (except RSET) with a 421 error, until QUIT is received. However, it was
7129 failing to send a response to QUIT.
7131 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7132 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7133 ACL is now run after the helo_try_verify_hosts code.
7135 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7136 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7137 case-independent, so other case variants are also recognized). Apparently
7138 some systems use these upper case variants.
7140 25. If more than two messages were waiting for the same host, and a transport
7141 filter was specified for the transport, Exim sent two messages over the
7142 same TCP/IP connection, and then failed with "socket operation on non-
7143 socket" when it tried to send the third.
7145 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7147 27. The extern definition of crypt16() in expand.c was not being excluded when
7148 the OS had its own crypt16() function.
7150 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7151 as a preferred synonym for return_size_limit, both as an option and as an
7154 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7156 30. Changed the default configuration ACL to relax the local part checking rule
7157 for addresses that are not in any local domains. For these addresses,
7158 slashes and pipe symbols are allowed within local parts, but the sequence
7159 /../ is explicitly forbidden.
7161 31. SPA server authentication was not clearing the challenge buffer before
7164 32. log_message in a "warn" ACL statement was writing to the reject log as
7165 well as to the main log, which contradicts the documentation and doesn't
7166 seem right (because no rejection is happening). So I have stopped it.
7168 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7169 However, I am unable to do any testing of this.
7171 34. Fixed an infelicity in the appendfile transport. When checking directories
7172 for a mailbox, to see if any needed to be created, it was accidentally
7173 using path names with one or more superfluous leading slashes; tracing
7174 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7176 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7177 contents are added to the log line that is written for every discarded
7178 recipient. (Previously a log_message setting was ignored.)
7180 36. The ${quote: operator now quotes the string if it is empty.
7182 37. The install script runs exim in order to find its version number. If for
7183 some reason other than non-existence or emptiness, which it checks, it
7184 could not run './exim', it was installing it with an empty version number,
7185 i.e. as "exim-". This error state is now caught, and the installation is
7188 38. An argument was missing from the function that creates an error message
7189 when Exim fails to connect to the socket for saslauthd authentication.
7190 This could cause Exim to crash, or give a corrupted message.
7192 39. Added isip, isip4, and isip6 to ${if conditions.
7194 40. The ACL variables $acl_xx are now saved with the message, and can be
7195 accessed later in routers, transports, and filters.
7197 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7198 strings in the file are not string-expanded.
7200 42. If a MAIL command specified a SIZE value that was too large to fit into an
7201 int variable, the check against message_size_limit failed. Such values are
7202 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7203 day this will have to be increased, but I don't think I want to be around
7204 when emails are that large.
7211 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7212 iconv() is not standard in FreeBSD.
7214 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7215 IPv6 enabled. The observed symptom was a segmentation fault on return from
7216 the function os_common_find_running_interfaces() in src/os.c.
7218 3. In the check_special_case() function in daemon.c I had used "errno" as an
7219 argument name, which causes warnings on some systems. This was basically a
7220 typo, since it was named "eno" in the comments!
7222 4. The code that waits for the clock to tick (at a resolution of some fraction
7223 of a second) so as to ensure message-id uniqueness was always waiting for
7224 at least one whole tick, when it could have waited for less. [This is
7225 almost certainly not relevant at current processor speeds, where it is
7226 unlikely to ever wait at all. But we try to future-proof.]
7228 5. The function that sleeps for a time interval that includes fractions of a
7229 second contained a race. It did not block SIGALRM between setting the
7230 timer, and suspending (a couple of lines later). If the interval was short
7231 and the sigsuspend() was delayed until after it had expired, the suspension
7232 never ended. On busy systems this could lead to processes getting stuck for
7235 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7236 process, before it forks any delivery processes. The open lookup caching
7237 mechanism meant that the open file or database connection was passed into
7238 the delivery process. The problem was that delivery processes always tidy
7239 up cached lookup data. This could cause a problem for the next delivery
7240 process started by the queue runner, because the external queue runner
7241 process does not know about the closure. So the next delivery process
7242 still has data in the lookup cache. In the case of a file lookup, there was
7243 no problem because closing a file descriptor in a subprocess doesn't affect
7244 the parent. However, if the lookup was caching a connection to a database,
7245 the connection was closed, and the second delivery process was likely to
7246 see errors such as "PGSQL: query failed: server closed the connection
7247 unexpectedly". The problem has been fixed by closing all cached lookups
7248 in a queue runner before running a delivery process.
7250 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7251 seem to have the "const" qualifier which it has on other OS. I've
7254 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7255 *of the same type* (client or server) with the same public name that an
7256 error should be diagnosed.
7258 9. When Exim looked up a host name for an IP address, but failed to find the
7259 original IP address when looking up the host name (a safety check), it
7260 output the message "<ip address> does not match any IP for NULL", which was
7261 confusing, to say the least. The bug was that the host name should have
7262 appeared instead of "NULL".
7264 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7265 user that is built into the binary, and the -C or -D options is used, root
7266 privilege is dropped before the configuration file is read. In addition,
7267 logging is switched to stderr instead of the normal log files. If the
7268 configuration then re-defines the Exim user, the unprivileged environment
7269 is probably not what is expected, so Exim logs a panic warning message (but
7272 However, if deliver_drop_privilege is set, the unprivileged state may well
7273 be exactly what is intended, so the warning has been cut out in that case,
7274 and Exim is allowed to try to write to its normal log files.
7280 1. smtp_return_error_details was not giving details for temporary sender
7281 or receiver verification errors.
7283 2. Diagnose a configuration error if two authenticators have the same public
7286 3. Exim used not to create the message log file for a message until the first
7287 delivery attempt. This could be confusing when incoming messages were held
7288 for policy or load reasons. The message log file is now created at the time
7289 the message is received, and an initial "Received" line is written to it.
7291 4. The automatically generated man page for command line options had a minor
7292 bug that caused no ill effects; however, a more serious problem was that
7293 the procedure for building the man page automatically didn't always
7294 operate. Consequently, release 4.20 contains an out-of-date version. This
7295 shouldn't happen again.
7297 5. When building Exim with embedded Perl support, the script that builds the
7298 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7299 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7301 6. The freeze_tell option was not being used for messages that were frozen on
7302 arrival, either by an ACL or by local_scan().
7304 7. Added the smtp_incomplete_transaction log selector.
7306 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7307 was accepting AUTH without a new EHLO.
7309 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7310 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7313 10. Logging of TCP/IP connections (when configured) now happens in the main
7314 daemon process instead of the child process, so that the TCP/IP connection
7315 count is more accurate (but it can never be perfect).
7317 11. The use of "drop" in a nested ACL was not being handled correctly in the
7318 outer ACL. Now, if condition failure induced by the nested "drop" causes
7319 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7320 or "require"), the connection is dropped.
7322 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7323 that yield "discard" can now be used with an "accept" or a "discard" verb,
7324 but an error is generated for any others (because I can't see a useful way
7325 to define what should happen).
7327 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7328 are now processed in the same way as lines in the Exim configuration file.
7329 In particular, continuation lines are supported.
7331 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7333 15. Added -ti meaning -t -i.
7335 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7336 domains, and warn by logging if others are found.
7338 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7339 returned from GIFCONF when Exim is trying to find the list of interfaces on
7340 a host. The code in os.c has been modified to copy each ifreq to an aligned
7341 structure in all cases.
7343 Also, in some cases, the returned ifreq's were being copied to a 'struct
7344 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7345 means the last couple of bytes of an IPv6 address could be chopped if the
7346 ifreq contained only a normal sockaddr (14 bytes storage).
7348 18. Named domain lists were not supported in the hosts_treat_as_local option.
7349 An entry such as +xxxx was not recognized, and was treated as a literal
7352 19. Ensure that header lines added by a DATA ACL are included in the reject log
7353 if the ACL subsequently rejects the message.
7355 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7356 MD5 (which is deprecated).
7358 21. When testing a filter file using -bf, Exim was writing a message when it
7359 took the sender from a "From " line in the message, but it was not doing so
7360 when it took $return_path from a Return-Path: header line. It now does.
7362 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7363 with a valid header line field name (a series of printing characters
7364 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7366 23. Changed "disc" in the source to "disk" to conform to the documentation and
7367 the book and for uniformity.
7369 24. Ignore Sendmail's -Ooption=value command line item.
7371 25. When execve() failed while trying to run a command in a pipe transport,
7372 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7373 could be confused with a return value of 69 from the command itself. This
7374 has been changed to 127, the value the shell returns if it is asked to run
7375 a non-existent command. The wording for the related log line suggests a
7376 non-existent command as the problem.
7378 26. If received_header_text expands to an empty string, do not add a Received:
7379 header line to the message. (Well, it adds a token one on the spool, but
7380 marks it "old" so that it doesn't get used or transmitted.)
7382 27. Installed eximstats 1.28 (addition of -nt option).
7384 28. There was no check for failure on the call to getsockname() in the daemon
7385 code. This can fail if there is a shortage of resources on the system, with
7386 ENOMEM, for example. A temporary error is now given on failure.
7388 29. Contrary to the C standard, it seems that in some environments, the
7389 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7390 program. Exim now does this explicitly; it affects the formatting of
7391 timestamps using strftime().
7393 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7394 complaints. I've now initialized all the variables, to avoid this.
7396 32. Header lines added by a system filter were not being "seen" during
7397 transport-time rewrites.
7399 33. The info_callback() function passed to OpenSSL is set up with type void
7400 (*)(SSL *, int, int), as described somewhere. However, when calling the
7401 function (actually a macro) that sets it up, the type void(*)() is
7402 expected. I've put in a cast to prevent warnings from picky compilers.
7404 34. If a DNS black list lookup found a CNAME record, but there were no A
7405 records associated with the domain it pointed at, Exim crashed.
7407 35. If a DNS black list lookup returned more than one A record, Exim ignored
7408 all but the first. It now scans all returned addresses if a particular IP
7409 value is being sought. In this situation, the contents of the
7410 $dnslist_value variable are a list of all the addresses, separated by a
7413 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7414 to accept a host name and all its aliases if the forward lookup for any of
7415 them yielded the IP address of the incoming connection. Now it accepts only
7416 those names whose forward lookup yields the correct IP address. Any other
7417 names are discarded. This closes a loophole whereby a rogue DNS
7418 administrator could create reverse DNS records to break through a
7419 wildcarded host restriction in an ACL.
7421 37. If a user filter or a system filter that ran in a subprocess used any of
7422 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7423 the wrong values were passed to the pipe command ($thisaddress had the
7424 value of $0, $0 had the value of $1, etc). This bug was introduced by
7425 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7427 38. Improved the line breaking for long SMTP error messages from ACLs.
7428 Previously, if there was no break point between 40 and 75 characters, Exim
7429 left the rest of the message alone. Two changes have been made: (a) I've
7430 reduced the minimum length to 35 characters; (b) if it can't find a break
7431 point between 35 and 75 characters, it looks ahead and uses the first one
7432 that it finds. This may give the occasional overlong line, but at least the
7433 remaining text gets split now.
7435 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7436 file descriptors might be low, and that setting 1000 would always raise it.
7437 It turns out that in some environments, the limit is already over 1000 and
7438 that lowering it causes trouble. So now Exim takes care not to decrease it.
7440 40. When delivering a message, the value of $return_path is set to $sender_
7441 address at the start of routing (routers may change the value). By an
7442 oversight, this default was not being set up when an address was tested by
7443 -bt or -bv, which affected the outcome if any router or filter referred to
7446 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7447 log only when "message" or "log_message" are set. However, if one of the
7448 conditions was an address verification, or a call to a nested ACL, the
7449 messages generated by the underlying test were being passed through. This
7450 no longer happens. The underlying message is available in $acl_verify_
7451 message for both "message" and "log_message" expansions, so it can be
7452 passed through if needed.
7454 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7455 new expansion $bh_ to give the encoded byte string without charset
7456 translation. Translation happens only if iconv() is available; HAVE_ICONV
7457 indicates this at build time. HEADERS_CHARSET gives the charset to
7458 translate to; headers_charset can change it in the configuration, and
7459 "headers charset" can change it in an individual filter file.
7461 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7462 that creates RFC 2047 encoded "words" labels them as that charset instead
7463 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7464 expansion operator; (ii) when Exim creates a From: line for a local
7465 message; (iii) when a header line is rewritten to include a "phrase" part.
7467 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7468 buggy, causing it to skip the first lines of messages whose message ID
7469 ended in 'D'. This would not have bitten before Exim release 4.14, because
7470 message IDs were unlikely to end in 'D' before then. The effect was to have
7471 incorrect size information for certain domains.
7473 45. #include "config.h" was missing at the start of the crypt16.c module. This
7474 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7477 46. If there was a timeout during a "random" callout check, Exim treated it as
7478 a failure of the random address, and carried on sending RSET and the real
7479 address. If the delay was just some slowness somewhere, the response to the
7480 original RCPT would be taken as a response to RSET and so on, causing
7481 mayhem of various kinds.
7483 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7484 when I implemented it. It didn't allow for the fact that some option values
7485 may legitimately be negative (e.g. size_addition), and it didn't even do
7486 the right test for positive values.
7488 48. Domain names in DNS records are case-independent. Exim always looks them up
7489 in lower case. Some resolvers return domain names in exactly the case they
7490 appear in the zone file, that is, they may contain uppercase letters. Not
7491 all resolvers do this - some return always lower case. Exim was treating a
7492 change of case by a resolver as a change of domain, similar to a widening
7493 of a domain abbreviation. This triggered its re-routing code and so it was
7494 trying to route what was effectively the same domain again. This normally
7495 caused routing to fail (because the router wouldn't handle the domain
7496 twice). Now Exim checks for this case specially, and just changes the
7497 casing of the domain that it ultimately uses when it transmits the message
7500 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7503 50. If a filter generated a file delivery with a non-absolute name (possible if
7504 no home directory exists for the router), the forbid_file option was not
7507 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7508 the existing equality matching.
7510 52. Exim was using ints instead of ino_t variables in some places where it was
7511 dealing with inode numbers.
7513 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7514 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7515 called TMPDIR, and if it finds it is different, it changes its value.
7517 54. The smtp_printf() function is now made available to local_scan() so
7518 additional output lines can be written before returning. There is also an
7519 smtp_fflush() function to enable the detection of a dropped connection.
7520 The variables smtp_input and smtp_batched_input are exported to
7523 55. Changed the default runtime configuration: the message "Unknown user"
7524 has been removed from the ACL, and instead placed on the localuser router,
7525 using the cannot_route_message feature. This means that any verification
7526 failures that generate their own messages won't get overridden. Similarly,
7527 the "Unrouteable address" message that was in the ACL for unverifiable
7528 relay addresses has also been removed.
7530 56. Added hosts_avoid_esmtp to the smtp transport.
7532 57. The exicyclog script was not checking for the esoteric option
7533 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7534 will work only if exicyclog is run under the appropriate euid.
7536 58. Following a discussion on the list, the rules by which Exim recognises line
7537 endings on incoming messages have been changed. The -dropcr and drop_cr
7538 options are now no-ops, retained only for backwards compatibility. The
7539 following line terminators are recognized: LF CRLF CR. However, special
7540 processing applies to CR:
7542 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7543 nor a local message in the state where . is a terminator.
7545 (ii) If a bare CR is encountered in a header line, an extra space is added
7546 after the line terminator so as not to end the header. The reasoning
7547 behind this is that bare CRs in header lines are most likely either
7548 to be mistakes, or people trying to play silly games.
7550 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7551 was being incorrectly given as 18 bytes larger than it should have been.
7552 This is a VOB (very old bug).
7554 60. This may never have affected anything current, but just in case it has:
7555 When the local host is found other than at the start of a list of hosts,
7556 the local host, those with the same MX, and any that follow, are discarded.
7557 When the list in question was part of a longer list of hosts, the following
7558 hosts (not currently being processed) were also being discarded. This no
7559 longer happens. I'm not sure if this situation could ever has previously
7562 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7565 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7566 header. This is something that is recommended in a new Internet Draft, and
7567 is something that is documented as being done by Sendmail. There are two
7568 possible values. For messages generated by the autoreply transport, Exim
7571 Auto-Submitted: auto-replied
7573 whereas for all other generated messages (e.g. bounces) it adds
7575 Auto-Submitted: auto-generated
7577 63. The "personal" condition in filters now includes a test for the
7578 Auto-Submitted: header. If it contains the string "auto-" the message it
7579 not considered personal.
7581 64. Added rcpt_include_affixes as a generic transport option.
7583 65. Added queue_only_override (default true).
7585 66. Added the syslog_duplication option.
7587 67. If what should have been the first header line of a message consisted of
7588 a space followed by a colon, Exim was mis-interpreting it as a header line.
7589 It isn't of course - it is syntactically invalid and should therefore be
7590 treated as the start of the message body. The misbehaviour could have
7591 caused a number of strange effects, including loss of data in subsequent
7592 header lines, and spool format errors.
7594 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7595 client host had authenticated. This control can now be exercised by an ACL
7596 for more flexibility.
7598 69. By default, callouts do not happen when testing with -bh. There is now a
7599 variant, -bhc, which does actually run the callout code, including
7600 consulting and updating the callout cache.
7602 70. Added support for saslauthd authentication, courtesy of Alexander
7605 71. If statvfs() failed on the spool or log directories while checking their
7606 size for availability, Exim confusingly gave the error "space shortage".
7607 Furthermore, in debugging mode it crashed with a floating point exception.
7608 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7609 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7610 serious problem, Exim now writes to the main and panic logs when this
7611 happens, with details of the failure. It then refuses to accept the
7612 incoming message, giving the message "spool directory problem" or "log
7613 directory problem" with a 421 code for SMTP messages.
7615 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7616 0, 1, and 2 exist, because some OS complain for execs without them (see
7617 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7618 descriptors. However, the code omitted to check that the open succeeded,
7619 causing mysterious errors if for some reason the permissions on /dev/null
7620 got screwed. Now Exim writes a message to the main and panic logs, and
7621 bombs out if it can't open /dev/null.
7623 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7624 interact so that it is all more flexible. It is supposed to remain
7625 backwards compatible. Also added extra_local_interfaces.
7627 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7628 to bomb out with an assertion failure - to the client this appears as a
7629 connection drop. This problem occurs in the part of the code that was taken
7630 from the Samba project. Fortunately, the assertion is in a very simple
7631 function, so I have fixed this by reproducing the function inline in the
7632 one place where it is called, and arranging for authentication to fail
7633 instead of killing the process with assert().
7635 75. The SPA client code was not working when the server requested OEM rather
7636 than Unicode encoding.
7638 76. Added code to make require_files with a specific uid setting more usable in
7639 the case where statting the file as root fails - usually a non-root-mounted
7640 NFS file system. When this happens and the failure is EACCES, Exim now
7641 forks a subprocess and does the per-uid checking as the relevant uid.
7643 77. Added process_log_path.
7645 78. If log_file_path was not explicitly set, a setting of check_log_space or
7646 check_log_inodes was ignored.
7648 79. If a space check for the spool or log partitions fails, the incident is now
7649 logged. Of course, in the latter case the data may get lost...
7651 80. Added the %p formatting code to string_format() so that it can be used to
7652 print addresses in debug_print(). Adjusted all the address printing in the
7653 debugging in store.c to use %p rather than %d.
7655 81. There was a concern that a line of code in smtp_in.c could overflow a
7656 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7657 initially expressed, the concern was not well-founded, because trailing
7658 spaces are removed early. However, if the trailing spaces were followed by
7659 a NULL, they did not get removed, so the overflow was possible. Two fixes
7662 (a) I re-wrote the offending code in a cleaner fashion.
7663 (b) If an incoming SMTP command contains a NULL character, it is rejected
7666 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7667 runs initgroups(), so that if the Exim user is in any additional groups,
7668 they will be used during message reception.
7674 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob