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 - the autoreply transport file, log and once options
94 - file names used by the redirect router (including filter files)
96 Previously this was permitted.
102 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
103 side, to decode a TLS 1.3 packet capture.
105 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
106 Previously the default library behaviour applied, sending two, each in
109 JH/03 Debug output for ACL now gives the config file name and line number for
112 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
114 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
116 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
117 buffer overrun for (non-chunking) other transports.
119 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
120 TLS1.3, means that a server rejecting a client certificate is not visible
121 to the client until the first read of encrypted data (typically the
122 response to EHLO). Add detection for that case and treat it as a failed
123 TLS connection attempt, so that the normal retry-in-clear can work (if
124 suitably configured).
126 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
127 and/or domain. Found and fixed by Jason Betts.
129 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
130 configuration). If a CNAME target was not a wellformed name pattern, a
133 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
134 the OS reports them interleaved with other addresses.
136 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
137 used both for input and for a verify callout, both encrypted, SMTP
138 responses being sent by the server could be lost. This resulted in
139 dropped connections and sometimes bounces generated by a peer sending
142 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
143 string. Previously it was possible to cause undefined behaviour in a
144 library routine (usually a crash). Found by "zerons".
146 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
149 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
150 API was removed, so update to use the newer ones.
152 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
153 any timeout set, is taking a long time. Previously we would hang on to a
154 rotated logfile "forever" if the input was arriving with long gaps
155 (a previous attempt to fix addressed lack, for a long time, of initial
158 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
159 shared (NFS) environment. The length of the tempfile name is now
160 4 + 16 ("hdr.$message_exim_id") which might break on file
161 systems which restrict the file name length to lower values.
164 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
165 shared (NFS) environment.
167 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
168 did for all versions <4.90). Notably -M, -m, --invert, -I may be
171 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
172 on some platforms for bit 31.
174 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
175 to changes apparently associated with TLS1.3 handling some of the APIs
176 previously used were either nonfunctional or inappropriate. Strings
177 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
178 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
179 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
180 This affects log line X= elements, the $tls_{in,out}_cipher variables,
181 and the use of specific cipher names in the encrypted= ACL condition.
183 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
185 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
186 verification result was not updated unless hosts_require_ocsp applied.
188 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
189 queue_list_requires_admin set to false, non-admin users were denied the
192 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
193 directory-of-certs mode. Previously they were advertised despite the
196 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
197 A single TCP connection by a client will now hold a TLS connection open
198 for multiple message deliveries, by default. Previoud the default was to
201 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
202 default. If built with the facility, DANE will be used. The facility
203 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
205 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
206 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
207 must be defined and you must still, unless you define DISABLE_TLS, manage
208 the the include-dir and library-file requirements that go with that
209 choice. Non-TLS builds are still supported.
211 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
214 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
215 default. If the platform supports and has the facility enabled, it will
216 be requested on all coneections.
218 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
219 controlled by the build-time option SUPPORT_PIPE_CONNECT.
221 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
223 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
224 success-DSN messages. Previously the From: header was always the default
225 one for these; the option was ignored.
227 JH/28 Fix the timeout on smtp response to apply to the whole response.
228 Previously it was reset for every read, so a teergrubing peer sending
229 single bytes within the time limit could extend the connection for a
230 long time. Credit to Qualsys Security Advisory Team for the discovery.
232 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
233 delivery address, which leaked information of the results of local
234 forwarding. Change to the original envelope recipient address, per
237 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
238 requested. Previously not bounce was generated and a log entry of
239 error ignored was made.
241 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
243 JH/32 Introduce a general tainting mechanism for values read from the input
244 channel, and values derived from them. Refuse to expand any tainted
245 values, to catch one form of exploit.
247 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
248 was unused and the unexpanded text used for the test. Found and
249 fixed by Ruben Jenster.
251 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
252 an attempt to use a TLS library read routine dereffed a nul pointer,
255 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
256 them in our smtp response.
258 JH/36 Have the general router option retry_use_local_part default to true when
259 any of the restrictive preconditions are set (to anything). Previously it
260 was only for check_local user. The change removes one item of manual
261 configuration which is required for proper retries when a remote router
262 handles a subset of addresses for a domain.
264 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
265 link count into consideration.
267 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
268 caused the extension of big_buffer, the following lines were ignored.
270 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
271 accordance with RFC 2308. Previously there was no expiry, so a longlived
272 receive process (eg. due to ACL delays) versus a short SOA value could
275 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
277 JH/39 Promote DMARC support to mainline.
279 JH/40 Bug 2452: Add a References: header to DSNs.
281 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
282 parameters. The relevant library call is documented as "Deprecated: This
283 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
284 3.6.0, DH parameters are negotiated following RFC7919."
286 HS/06 Change the default of dnssec_request_domains to "*"
288 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
289 carried on and emitted a BDAT command, even when PIPELINING was not
292 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
293 buffer was used for the filename, resulting in a trap when tainted
294 arguments (eg. $domain) were used.
296 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
297 recommended to avoid a possible server-load attack. The feature can be
298 re-enabled via the openssl_options main cofiguration option.
300 JH/45 local_scan API: documented the current smtp_printf() call. This changed
301 for version 4.90 - adding a "more data" boolean to the arguments.
302 Bumped the ABI version number also, this having been missed previously;
303 release versions 4.90 to 4.92.3 inclusive were effectively broken in
304 respect of usage of smtp_printf() by either local_scan code or libraries
305 accessed via the ${dlfunc } expansion item. Both will need coding
306 adjustment for any calls to smtp_printf() to match the new function
307 signature; a FALSE value for the new argument is always safe.
309 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
310 the file-offset (which the Linux syscall does, and exim expects); this
311 resulted in an indefinite loop.
313 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
314 to do ARC verification. The Authentication-Results: header line added
315 by the configuration then had no ARC item.
321 JH/01 Remove code calling the customisable local_scan function, unless a new
322 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
324 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
325 non-signal-safe functions being used.
327 JH/03 Bug 2269: When presented with a received message having a stupidly large
328 number of DKIM-Signature headers, disable DKIM verification to avoid
329 a resource-consumption attack. The limit is set at twenty.
331 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
332 report of oldest_pass in ${authres } in consequence, and separate out
333 some descriptions of reasons for verification fail.
335 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
336 files in the spool were present and unlocked. A queue-runner could spot
337 them, resulting in a duplicate delivery. Fix that by doing the unlock
338 after the unlink. Investigation by Tim Stewart. Take the opportunity to
339 add more error-checking on spoolfile handling while that code is being
342 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
343 No known attacks, no CVE, this is defensive hardening.
345 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
346 a queue-runner could start a delivery while other operations were ongoing.
347 Cutthrough delivery was a common victim, resulting in duplicate delivery.
348 Found and investigated by Tim Stewart. Fix by using the open message data
349 file handle rather than opening another, and not locally closing it (which
350 releases a lock) for that case, while creating the temporary .eml format
351 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
353 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
354 $sender_verify_failure/$recipient_verify_failure to "random".
356 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
359 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
360 Previously this would segfault.
362 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
365 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
366 like zero, since the resolver should be doing this for us, But we need one
367 as a CNAME but no MX presence gets the CNAME returned; we need to check
368 that doesn't point to an MX to declare it "no MX returned" rather than
369 "error, loop". A new main option is added so the older capability of
370 following some limited number of chain links is maintained.
372 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
374 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
375 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
376 now-deprecated earlier definitions used only specified the range up to TLS
377 1.2 (in the older-version library docs).
379 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
381 JH/15 Rework TLS client-side context management. Stop using a global, and
382 explicitly pass a context around. This enables future use of TLS for
383 connections to service-daemons (eg. malware scanning) while a client smtp
384 connection is using TLS; with cutthrough connections this is quite likely.
386 JH/16 Fix ARC verification to do AS checks in reverse order.
388 JH/17 Support a "tls" option on the ${readsocket } expansion item.
390 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
391 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
392 Previously the "utf8" would be re-prepended for every additional message.
394 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
395 Previously thery were accepted, resulting in issues when attempting to
396 forward messages to a non-supporting MTA.
398 PP/02 Let -n work with printing macros too, not just options.
400 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
401 one parent address was copied, and bogus data was used at delivery-logging
402 time. Either a crash (after delivery) or bogus log data could result.
403 Discovery and analysis by Tim Stewart.
405 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
406 Previously if the string ended mid-character, we did not insert the
407 promised '?' replacement.
409 PP/04 Documentation: current string operators work on bytes, not codepoints.
411 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
412 should pack well giving a smaller memory footprint so better caching and
413 therefore performance. Group the declarations where this can't be done so
414 that the byte-sized flag variables are not interspersed among pointer
415 variables, giving a better chance of good packing by the compiler.
417 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
418 non-null, to avoid issues with sites running BATV. Previously reports were
419 sent with an empty envelope sender so looked like bounces.
421 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
422 The ignore_error flag wasn't being returned from the filter subprocess so
423 was not set for later routers. Investigation and fix by Matthias Kurz.
425 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
426 and a msg:complete for the whole, when a message is manually removed using
427 -Mrm. Developement by Matthias Kurz, hacked on by JH.
429 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
430 a "Gnu special" function, asprintf() in the DB utility binary builds; I
431 hope that is portable enough.
433 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
434 requiring a known-CA anchor certificate; make it now rely entirely on the
435 TLSA as an anchor. Checking the name on the leaf cert against the name
436 on the A-record for the host is still done for TA (but not for EE mode).
438 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
439 included in delivery lines for non-proxied connections, when compiled with
440 SUPPORT_SOCKS and running with proxy logging enabled.
442 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
443 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
444 move the existing event to fire before the normal logging of message
445 failure so that custom logging is bracketed by normal logging.
447 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
448 msg:fail:internal event. Developement by Matthias Kurz.
450 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
451 far too small for todays use of crypto signatures stored there. Go all
452 the way to the max DNS message size of 64kB, even though this might be
453 overmuch for IOT constrained device use.
455 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
456 copy a string over itself. The library routine is documented as not
457 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
459 JH/32 For main options check_spool_space and check_inode_space, where the
460 platform supports 64b integers, support more than the previous 2^31 kB
461 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
462 the previous G, M, k.
464 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
465 $authenticated_fail_id variable on authentication failure. Previously
468 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
469 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
470 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
471 GNUTLS_SEC_PARAM_MEDIUM.
473 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
474 side. Previously we would continue as if no SNI had been received.
476 JH/36 Harden the handling of string-lists. When a list consisted of a sole
477 "<" character, which should be a list-separator specification, we walked
478 off past the nul-terimation.
480 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
481 causes) even when the retry time is not yet met. Previously they were
482 not, meaning that when (say) an account was over-quota and temp-rejecting,
483 and multiple senders' messages were queued, only one sender would get
484 notified on each configured delay_warning cycle.
486 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
488 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
489 cipher-suites, an error can be left on the stack even for a succeeding
490 accept; this results in impossible error messages when a later operation
493 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
494 return error codes indicating retry. Under TLS1.3 this becomes required.
496 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
497 it only wrote the new authenticators, resulting in a lack of tracking of
498 peer changes of ESMTP extensions until the next cache flush.
500 JH/41 Fix the loop reading a message header line to check for integer overflow,
501 and more-often against header_maxsize. Previously a crafted message could
502 induce a crash of the recive process; now the message is cleanly rejected.
504 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
505 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
511 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
512 When redis_servers is set to a list of > 1 element, and the Redis servers
513 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
514 case of MOVED into a DEFER case instead, thus moving the query onto the
515 next server in the list. For a cluster of N elements, all N servers must
516 be defined in redis_servers.
518 GF/02 Catch and remove uninitialized value warning in exiqsumm
519 Check for existence of @ARGV before looking at $ARGV[0]
521 JH/01 Replace the store_release() internal interface with store_newblock(),
522 which internalises the check required to safely use the old one, plus
523 the allocate and data copy operations duplicated in both (!) of the
524 extant use locations.
526 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
527 modifier. This matches the restriction on the commandline.
529 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
530 Previously only the last row was returned.
532 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
533 we assumed that tags in the header were well-formed, and parsed the
534 element content after inspecting only the first char of the tag.
535 Assumptions at that stage could crash the receive process on malformed
538 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
539 While running the DKIM ACL we operate on the Permanent memory pool so that
540 variables created with "set" persist to the DATA ACL. Also (at any time)
541 DNS lookups that fail create cache records using the Permanent pool. But
542 expansions release any allocations made on the current pool - so a dnsdb
543 lookup expansion done in the DKIM ACL releases the memory used for the
544 DNS negative-cache, and bad things result. Solution is to switch to the
545 Main pool for expansions.
546 While we're in that code, add checks on the DNS cache during store_reset,
547 active in the testsuite.
548 Problem spotted, and debugging aided, by Wolfgang Breyha.
550 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
551 When none of the hosts presented to a transport match an already-open
552 connection, close it and proceed with the list. Previously we would
553 queue the message. Spotted by Lena with Yahoo, probably involving
556 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
557 Previously a spurious "250 OK id=" response was appended to the proper
560 JH/08 The "support for" informational output now, which built with Content
561 Scanning support, has a line for the malware scanner interfaces compiled
562 in. Interface can be individually included or not at build time.
564 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
565 by the template makefile "src/EDITME". The "STREAM" support for an older
566 ClamAV interface method is removed.
568 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
569 rows affected is given instead).
571 JH/11 The runtime Berkeley DB library version is now additionally output by
572 "exim -d -bV". Previously only the compile-time version was shown.
574 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
575 SMTP connection. Previously, when one had more receipients than the
576 first, an abortive onward connection was made. Move to full support for
577 multiple onward connections in sequence, handling cutthrough connection
578 for all multi-message initiating connections.
580 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
581 routers. Previously, a multi-recipient message would fail to match the
582 onward-connection opened for the first recipient, and cause its closure.
584 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
585 a timeout on read on a GnuTLS initiating connection, resulting in the
586 initiating connection being dropped. This mattered most when the callout
587 was marked defer_ok. Fix to keep the two timeout-detection methods
590 JH/15 Relax results from ACL control request to enable cutthrough, in
591 unsupported situations, from error to silently (except under debug)
592 ignoring. This covers use with PRDR, frozen messages, queue-only and
595 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
597 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
598 metadata, resulting in a crash in free().
600 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
601 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
602 Broken also in d185889f4, with init system revamp.
604 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
605 Previously we abruptly closed the connection after reading a malware-
606 found indication; now we go on to read the "scan ok" response line,
609 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
610 ACL. Previously, a crash would result.
612 JH/19 Speed up macro lookups during configuration file read, by skipping non-
613 macro text after a replacement (previously it was only once per line) and
614 by skipping builtin macros when searching for an uppercase lead character.
616 JH/20 DANE support moved from Experimental to mainline. The Makefile control
617 for the build is renamed.
619 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
620 was allocated for every new TLS startup, meaning one per message. Fix
621 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
623 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
624 reported the original. Fix to report (as far as possible) the ACL
625 result replacing the original.
627 JH/23 Fix memory leak during multi-message connections using STARTTLS under
628 OpenSSL. Certificate information is loaded for every new TLS startup,
629 and the resources needed to be freed.
631 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
633 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
636 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
637 DATA response info to the (existing) per-recipient response info for
638 the "C=" log element. It can have useful tracking info from the
639 destination system. Patch from Simon Arlott.
641 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
642 length value. Previously this would segfault.
644 HS/02 Support Avast multiline protoocol, this allows passing flags to
645 newer versions of the scanner.
647 JH/28 Ensure that variables possibly set during message acceptance are marked
648 dead before release of memory in the daemon loop. This stops complaints
649 about them when the debug_store option is enabled. Discovered specifically
650 for sender_rate_period, but applies to a whole set of variables.
651 Do the same for the queue-runner and queue-list loops, for variables set
652 from spool message files. Do the same for the SMTP per-message loop, for
653 certain variables indirectly set in ACL operations.
655 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
656 as a multi-recipient message from a mailinglist manager). The coding had
657 an arbitrary cutoff number of characters while checking for more input;
658 enforced by writing a NUL into the buffer. This corrupted long / fast
659 input. The problem was exposed more widely when more pipelineing of SMTP
660 responses was introduced, and one Exim system was feeding another.
661 The symptom is log complaints of SMTP syntax error (NUL chars) on the
662 receiving system, and refused recipients seen by the sending system
663 (propating to people being dropped from mailing lists).
664 Discovered and pinpointed by David Carter.
666 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
667 replaced by the ${authresults } expansion.
669 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
671 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
672 allows proper process termination in container environments.
674 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
675 Previously the "final dot" had a newline after it; ensure it is CR,LF.
677 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
678 and "err_perm", deprecated since 4.83 when the RFC-defined words
679 "temperror" and "permerror" were introduced.
681 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
682 transport-filters or DKIM-signing. The restriction was lost in the
683 consolidation of verify-callout and delivery SMTP handling.
684 Extend the restriction to also cover ARC-signing.
686 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
687 in defer=pass mode supply a 450 to the initiator. Previously the message
690 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
691 tls_require_ciphers is used as before.
693 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
694 "pass_unscanned". Only tmpfails from the scanner are written to
695 the paniclog, as they may require admin intervention (permission
696 denied, license issues). Other scanner errors (like decompression
697 bombs) do not cause a paniclog entry.
699 JH/36 Fix reinitialisation of DKIM logging variable between messages.
700 Previously it was possible to log spurious information in receive log
703 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
704 triggered odd behaviour from Outlook Express clients.
706 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
709 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
710 since the IETF WG has not yet settled on that versus the original
711 "bare" representation.
713 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
714 Previously the millisecond value corrupted the output.
715 Fix also for syslog_pid=no and log_selector +pid, for which the pid
716 corrupted the output.
722 JH/01 Rework error string handling in TLS interface so that the caller in
723 more cases is responsible for logging. This permits library-sourced
724 string to be attached to addresses during delivery, and collapses
725 pairs of long lines into single ones.
727 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
728 during configuration. Wildcards are allowed and expanded.
730 JH/02 Rework error string handling in DKIM to pass more info back to callers.
731 This permits better logging.
733 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
734 do not close it down and have the child transport start it up again on
735 the passed-on TCP connection. Instead, proxy the child (and any
736 subsequent ones) for TLS via a unix-domain socket channel. Logging is
737 affected: the continued delivery log lines do not have any DNSSEC, TLS
738 Certificate or OCSP information. TLS cipher information is still logged.
740 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
741 identical IP addresses on different listening ports. Will also affect
744 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
745 add noisy ifdef guards to special-case this sillyness.
746 Patch from Bernd Kuhls.
748 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
749 than 255 are no longer allowed.
751 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
752 Disable the session-cache too, which might reduce our load. Since we
753 currrectly use a new context for every connection, both as server and
754 client, there is no benefit for these.
755 GnuTLS appears to not support tickets server-side by default (we don't
756 call gnutls_session_ticket_enable_server()) but client side is enabled
757 by default on recent versions (3.1.3 +) unless the PFS priority string
760 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
761 <https://reproducible-builds.org/specs/source-date-epoch/>.
763 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
764 the check for any unsuccessful recipients did not notice the limit, and
765 erroneously found still-pending ones.
767 JH/08 Pipeline CHUNKING command and data together, on kernels that support
768 MSG_MORE. Only in-clear (not on TLS connections).
770 JH/09 Avoid using a temporary file during transport using dkim. Unless a
771 transport-filter is involved we can buffer the headers in memory for
772 creating the signature, and read the spool data file once for the
773 signature and again for transmission.
775 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
776 4.77 as the kernel support then wasn't solid, having issues in 64bit
777 mode. Now, it's been long enough. Add support for FreeBSD also.
779 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
780 case where the routing stage had gathered several addresses to send to
781 a host before calling the transport for the first, we previously failed
782 to close down TLS in the old transport process before passing the TCP
783 connection to the new process. The new one sent a STARTTLS command
784 which naturally failed, giving a failed delivery and bloating the retry
785 database. Investigation and fix prototype from Wolfgang Breyha.
787 JH/12 Fix check on SMTP command input synchronisation. Previously there were
788 false-negatives in the check that the sender had not preempted a response
789 or prompt from Exim (running as a server), due to that code's lack of
790 awareness of the SMTP input buffering.
792 PP/04 Add commandline_checks_require_admin option.
793 Exim drops privileges sanely, various checks such as -be aren't a
794 security problem, as long as you trust local users with access to their
795 own account. When invoked by services which pass untrusted data to
796 Exim, this might be an issue. Set this option in main configuration
797 AND make fixes to the calling application, such as using `--` to stop
800 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
801 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
802 responses to those, into a single TLS record each way (this usually means
803 a single packet). As a side issue, smtp_enforce_sync now works on TLS
806 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
807 affects you only if you're dancing at the edge of the param size limits.
808 If you are, and this message makes sense to you, then: raise the
809 configured limit or use OpenSSL 1.1. Nothing we can do for older
812 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
813 cmdline element to get the documented default one. Previously it was
816 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
817 are unsuable for non-message-specific reasons - eg. connection timeout,
818 banner-time rejection.
820 JH/16 Fix logging of delivery remote port, when specified by router, under
823 PP/06 Repair manualroute's ability to take options in any order, even if one
824 is the name of a transport.
827 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
829 JH/17 Change the list-building routines interface to use the expanding-string
830 triplet model, for better allocation and copying behaviour.
832 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
833 Previously it was constructed the first time a possibly-matching string
834 was met in the configuration file input during startup; now it is done
837 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
838 compatible one, to avoid the (poorly documented) possibility of a config
839 file in the working directory redirecting the DB files, possibly correpting
840 some existing file. CVE-2017-10140 assigned for BDB.
842 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
843 cache-hot. Previously, although the result was properly cached, the
844 initial verify call returned a defer.
846 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
847 the main verify for receipient in uncached-mode.
849 JH/22 Retire historical build files to an "unsupported" subdir. These are
850 defined as "ones for which we have no current evidence of testing".
852 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
853 if present. Previously it was ignored.
855 JH/24 Start using specified-initialisers in C structure init coding. This is
856 a C99 feature (it's 2017, so now considered safe).
858 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
859 if was a fixed-sized field and bitmask ops via macros; it is now more
862 PP/07 GitHub PR 56: Apply MariaDB build fix.
863 Patch provided by Jaroslav Škarvada.
865 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
866 during Coverity cleanups [4.87 JH/47]
867 Diagnosis and fix provided by Michael Fischer v. Mollard.
869 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
870 the right size to place the terminating semicolon on its own folded
871 line, the header hash was calculated to an incorrect value thanks to
872 the (relaxed) space the fold became.
874 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
875 and confused the parent.
877 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
878 which could crash as a result. This could lead to undeliverable messages.
880 JH/28 Logging: "next input sent too soon" now shows where input was truncated
883 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
884 matters on fast-turnover and PID-randomising systems, which were getting
885 out-of-order delivery.
887 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
888 a possibly-overlapping copy. The symptom was that "Remote host closed
889 connection in response to HELO" was logged instead of the actual 4xx
892 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
893 Previously only that bufferd was discarded, resulting in SYMTP command
896 JH/32 DKIM: when a message has multiple signatures matching an identity given
897 in dkim_verify_signers, run the dkim acl once for each. Previously only
898 one run was done. Bug 2189.
900 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
901 "panic the current process" to "deliberately defer". The panic log is
902 still written with the problem list name; the mail and reject logs now
903 get a temp-reject line for the message that was being handled, saying
904 something like "domains check lookup or other defer". The SMTP 451
905 message is still "Temporary local problem".
907 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
908 A crafted sequence of BDAT commands could result in in-use memory beeing
909 freed. CVE-2017-16943.
911 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
912 from SMTP input. Previously it was always done; now only done for DATA
913 and not BDAT commands. CVE-2017-16944.
915 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
916 to the message (such as an overlong header line). Previously this was
917 not done and we did not exit BDAT mode. Followon from the previous item
918 though a different problem.
924 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
925 than -2003 did; needs libidn2 in addition to libidn.
927 JH/02 The path option on a pipe transport is now expanded before use.
929 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
930 Patch provided by "Björn", documentation fix added too.
932 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
933 missing a wire-to-host endian conversion.
935 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
936 close after a BDAT command line could be taken as a following command,
937 giving a synch failure. Fix by only checking for synch immediately
938 before acknowledging the chunk.
940 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
941 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
942 macro. Patches provided by Josh Soref.
944 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
945 Previously we did not; the RFC seems ambiguous and VRFY is not listed
946 by IANA as a service extension. However, John Klensin suggests that we
949 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
950 the dkim code may be unix-mode line endings rather than smtp wire-format
951 CRLF, so prepend a CR to any bare LF.
953 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
954 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
956 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
957 the first were themselves being wrongly included in the feed into dkim
958 processing; with most chunk sizes in use this resulted in an incorrect
959 body hash calculated value.
961 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
962 DKIM signature block, for verification. Although advised against by
963 standards it is specifically not ruled illegal.
965 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
967 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
968 missing a body hash (the bh= tag).
970 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
971 It seems that HAProxy sends the Proxy Protocol information in clear and
972 only then does a TLS startup, so do the same.
974 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
975 TCP connections (such as for Spamd) unless the daemon successfully set
976 Fast Open mode on its listening sockets. This fixes breakage seen on
977 too-old kernels or those not configured for Fast Open, at the cost of
978 requiring both directions being enabled for TFO, and TFO never being used
979 by non-daemon-related Exim processes.
981 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
982 endings, at least on the first header line. Try to canonify any that get
983 past that check, despite the cost.
985 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
986 now limited to an arbitrary five deep, while parsing addresses with the
987 strip_excess_angle_brackets option enabled.
989 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
990 instead leave the unprompted TLS handshake in socket buffer for the
991 TLS library to consume.
993 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
995 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
997 JH/16 Drop variables when they go out of scope. Memory management drops a whole
998 region in one operation, for speed, and this leaves assigned pointers
999 dangling. Add checks run only under the testsuite which checks all
1000 variables at a store-reset and panics on a dangling pointer; add code
1001 explicitly nulling out all the variables discovered. Fixes one known
1002 bug: a transport crash, where a dangling pointer for $sending_ip_address
1003 originally assigned in a verify callout, is re-used.
1005 PP/06 Drop '.' from @INC in various Perl scripts.
1007 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1009 PP/08 Reduce a number of compilation warnings under clang; building with
1010 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1011 should be warning-free.
1013 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1015 HS/01 Fix portability problems introduced by PP/08 for platforms where
1016 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1018 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1019 chunk. This allows us to accept broken chunked messages. We need a more
1020 general solution here.
1022 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1023 already-broken messages in the queue.
1025 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1027 JH/19 Fix reference counting bug in routing-generated-address tracking.
1033 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1034 supports it and a size is available (ie. the sending peer gave us one).
1036 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1037 years of being deprecated). The replacements are the ACLs
1038 acl_smtp_mime and acl_not_smtp_mime.
1040 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1041 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1042 or even an in-clear connection were permitted. Now, if the host lookup
1043 was dnssec and dane was requested then the host is only used if the
1044 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1045 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1046 if one fails this test.
1047 This means that a poorly-configured remote DNS will make it incommunicado;
1048 but it protects against a DNS-interception attack on it.
1050 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1051 non-noisy when a race steals the message being considered.
1053 JH/05 If main configuration option tls_certificate is unset, generate a
1054 self-signed certificate for inbound TLS connections.
1056 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1057 in rewrites and routers.
1059 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1060 and logged a warning sing 4.83; now they are a configuration file error.
1062 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1063 (lacking @domain). Apply the same qualification processing as RCPT.
1065 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1067 JH/10 Support ${sha256:} applied to a string (as well as the previous
1070 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1071 a cutthrough deliver is pending, as we always want to make a connection.
1072 This also avoids re-routing the message when later placing the cutthrough
1073 connection after a verify cache hit.
1074 Do not update it with the verify result either.
1076 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1077 when routing results in more than one destination address.
1079 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1080 signing (which inhibits the cutthrough capability). Previously only
1081 the presence of an option was tested; now an expansion evaluating as
1082 empty is permissible (obviously it should depend only on data available
1083 when the cutthrough connection is made).
1085 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1086 the relevant preceding SMTP command did not note the pipelining mode.
1088 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1089 Previously they were not counted.
1091 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1092 as one having no matching records. Previously we deferred the message
1093 that needed the lookup.
1095 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1096 distinguished as "(=".
1098 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1099 for missing MX records. Previously it only worked for missing A records.
1101 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1103 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1104 after the data-go-ahead and data-ack. Patch from Jason Betts.
1106 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1107 even for a "none" policy. Patch from Tony Meyer.
1109 JH/22 Fix continued use of a connection for further deliveries. If a port was
1110 specified by a router, it must also match for the delivery to be
1113 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1114 When one of the recipients of a message was unsuitable for the connection
1115 (has no matching addresses), we lost track of needing to mark it
1116 deferred. As a result mail would be lost.
1118 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1120 JH/25 Decoding ACL controls is now done using a binary search; the source code
1121 takes up less space and should be simpler to maintain. Merge the ACL
1122 condition decode tables also, with similar effect.
1124 JH/26 Fix problem with one_time used on a redirect router which returned the
1125 parent address unchanged. A retry would see the parent address marked as
1126 delivered, so not attempt the (identical) child. As a result mail would
1129 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1130 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1131 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1134 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1135 Main config options check_{log,spool}_{inodes,space} are now
1136 100 inodes, 10MB unless set otherwise in the configuration.
1138 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1139 Previously it only applied to the main-section connection policy
1142 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1144 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1145 by me. Added RFC7919 DH primes as an alternative.
1147 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1148 are not in the system include path.
1150 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1151 GnuTLS, when a session startup failed (eg because the client disconnected)
1152 Exim did stdio operations after fclose. This was exposed by a recent
1153 change which nulled out the file handle after the fclose.
1155 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1156 signed directly by the cert-signing cert, rather than an intermediate
1157 OCSP-signing cert. This is the model used by LetsEncrypt.
1159 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1161 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1162 an incoming connection.
1164 HS/02 Bug 1802: Do not half-close the connection after sending a request
1167 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1168 fallback to "prime256v1".
1170 JH/34 SECURITY: Use proper copy of DATA command in error message.
1171 Could leak key material. Remotely exploitable. CVE-2016-9963.
1177 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1178 and 3.4.4 - once the server is enabled to respond to an OCSP request
1179 it does even when not requested, resulting in a stapling non-aware
1180 client dropping the TLS connection.
1182 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1183 support variable-length bit vectors. No functional change.
1185 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1186 The I= interface field on outgoing lines is now after the H= remote
1187 host field, same as incoming lines. There is a separate
1188 outgoing_interface log selector which allows you to disable the
1191 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1192 If not running log_selector +smtp_connection the mainlog would be held
1193 open indefinitely after a "too many connections" event, including to a
1194 deleted file after a log rotate. Leave the per net connection logging
1195 leaving it open for efficiency as that will be quickly detected by the
1196 check on the next write.
1198 HS/01 Bug 1671: Fix post transport crash.
1199 Processing the wait-<transport> messages could crash the delivery
1200 process if the message IDs didn't exist for some reason. When
1201 using 'split_spool_directory=yes' the construction of the spool
1202 file name failed already, exposing the same netto behaviour.
1204 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1205 mime_regex ACL conditions.
1207 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1208 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1209 to HELO, local diagnostic string.
1211 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1212 log line to debug. Even when configured with a tls authenticator many
1213 client connections are expected to not authenticate in this way, so
1214 an authenticate fail is not an error.
1216 HS/02 Add the Exim version string to the process info. This way exiwhat
1217 gives some more detail about the running daemon.
1219 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1220 matter for fast-change records such as DNSBLs.
1222 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1223 retry record, even if constant. There may be multiple transports with
1224 different interface settings and the retry behaviour needs to be kept
1227 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1229 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1231 JH/10 Bug 840: fix log_defer_output option of pipe transport
1233 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1234 hosts_randomize. This matters a lot when combined with mua_wrapper.
1236 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1237 ${quote_pgsql:<string>} operator.
1239 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1241 JH/14 Change header file name for freeradius-client. Relevant if compiling
1242 with Radius support; from the Gentoo tree and checked under Fedora.
1244 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1246 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1247 option transport_filter, meaning no filtering.
1249 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1251 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1252 defaults to "*" (all hosts). The variable is now available when not built
1253 with TLS, default unset, mainly to enable keeping the testsuite sane.
1254 If a server certificate is not supplied (via tls_certificate) an error is
1255 logged, and clients will find TLS connections fail on startup. Presumably
1256 they will retry in-clear.
1257 Packagers of Exim are strongly encouraged to create a server certificate
1258 at installation time.
1260 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1261 with the $config_file variable.
1263 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1264 in transport context, after the attempt, and per-recipient. The latter type
1265 is per host attempted. The event data is the error message, and the errno
1266 information encodes the lookup type (A vs. MX) used for the (first) host,
1267 and the trailing two digits of the smtp 4xx response.
1269 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1270 to write to mainlog (or rejectlog, paniclog) in the window between file
1271 creation and permissions/ownership being changed. Particularly affects
1272 installations where exicyclog is run as root, rather than exim user;
1273 result is that the running daemon panics and dies.
1275 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1277 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1278 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1279 "pri" and "weight". Note that the previous implicit priority given by the
1280 list order is no longer honoured.
1282 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1283 for DKIM processing.
1285 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1286 by defining SUPPORT_SOCKS.
1288 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1289 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1290 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1291 variables are renamed to proxy_{local,external}_{address,port}.
1293 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1294 for a build by defining SUPPORT_I18N
1296 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1297 of the query string, and make ${quote_redis:} do that quoting.
1299 JH/29 Move Events support from Experimental to mainline, enabled by default
1300 and removable for a build by defining DISABLE_EVENT.
1302 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1304 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1305 cached by the daemon.
1307 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1308 by defining LOOKUP_REDIS. The libhiredis library is required.
1310 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1311 keys are given for lookup.
1313 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1314 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1315 only supported when built with TLS support. The PolarSSL SHA routines
1316 are still used when the TLS library is too old for convenient support.
1318 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1319 openssl_options), for security. OpenSSL forces this from version 1.1.0
1320 server-side so match that on older versions.
1322 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1323 allocation for $value could be released as the expansion processing
1324 concluded, but leaving the global pointer active for it.
1326 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1327 and to use the domains and local_parts ACL conditions.
1329 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1330 incorrectly not doubled on cutthrough transmission, hence seen as a
1331 body-termination at the receiving system - resulting in truncated mails.
1332 Commonly the sender saw a TCP-level error, and retransmitted the message
1333 via the normal store-and-forward channel. This could result in duplicates
1334 received - but deduplicating mailstores were liable to retain only the
1335 initial truncated version.
1337 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1339 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1341 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1342 we're in there, support oversigning also; bug 1309.
1344 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1346 HS/04 Add support for keep_environment and add_environment options.
1348 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1349 either intentional arithmetic overflow during PRNG, or testing config-
1352 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1353 delivery resulted in actual delivery. Cancel cutthrough before DATA
1356 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1357 rejects a recipient: pass the reject to the originator.
1359 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1360 Many were false-positives and ignorable, but it's worth fixing the
1363 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1364 for the new environment-manipulation done at startup. Move the routines
1365 from being local to tls.c to being global via the os.c file.
1367 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1368 an extract embedded as result-arg for a map, the first arg for extract
1369 is unavailable so we cannot tell if this is a numbered or keyed
1370 extraction. Accept either.
1376 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1379 JH/02 The smtp transport option "multi_domain" is now expanded.
1381 JH/03 The smtp transport now requests PRDR by default, if the server offers
1384 JH/04 Certificate name checking on server certificates, when exim is a client,
1385 is now done by default. The transport option tls_verify_cert_hostnames
1386 can be used to disable this per-host. The build option
1387 EXPERIMENTAL_CERTNAMES is withdrawn.
1389 JH/05 The value of the tls_verify_certificates smtp transport and main options
1390 default to the word "system" to access the system default CA bundle.
1391 For GnuTLS, only version 3.0.20 or later.
1393 JH/06 Verification of the server certificate for a TLS connection is now tried
1394 (but not required) by default. The verification status is now logged by
1395 default, for both outbound TLS and client-certificate supplying inbound
1398 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1401 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1402 Status Notification (bounce) messages are now MIME format per RFC 3464.
1403 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1404 under the control of the dsn_advertise_hosts option, and routers may
1405 have a dsn_lasthop option.
1407 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1408 default, modifiable by a malware= option. The list separator for
1409 the options can now be changed in the usual way. Bug 68.
1411 JH/10 The smtp_receive_timeout main option is now expanded before use.
1413 JH/11 The incoming_interface log option now also enables logging of the
1414 local interface on delivery outgoing connections.
1416 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1417 if the interface and destination host and port all match.
1419 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1422 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1423 Patch from Andrew Lewis.
1425 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1426 now supports optional time-restrictions, weighting, and priority
1427 modifiers per server. Patch originally by <rommer@active.by>.
1429 JH/16 The spamd_address main option now supports a mixed list of local
1430 and remote servers. Remote servers can be IPv6 addresses, and
1431 specify a port-range.
1433 JH/17 Bug 68: The spamd_address main option now supports an optional
1434 timeout value per server.
1436 JH/18 Bug 1581: Router and transport options headers_add/remove can
1437 now have the list separator specified.
1439 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1442 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1445 JH/21 Support for the A6 type of dns record is withdrawn.
1447 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1448 rather than the verbs used.
1450 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1451 from 255 to 1024 chars.
1453 JH/24 Verification callouts now attempt to use TLS by default.
1455 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1456 are generic router options now. The defaults didn't change.
1458 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1459 Original patch from Alexander Shikoff, worked over by JH.
1461 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1462 files if ZCAT_COMMAND is not executable.
1464 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1466 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1468 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1469 Normally benign, it bites when the pair was led to by a CNAME;
1470 modern usage is to not canonicalize the domain to a CNAME target
1471 (and we were inconsistent anyway for A-only vs AAAA+A).
1473 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1475 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1476 when evaluating $sender_host_dnssec.
1478 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1479 $sender_helo_dnssec variable.
1481 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1483 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1485 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1487 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1488 documented as working, but never had. Support all but $spam_report.
1490 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1491 added for tls authenticator.
1493 HS/03 Add perl_taintmode main config option
1499 TL/01 When running the test suite, the README says that variables such as
1500 no_msglog_check are global and can be placed anywhere in a specific
1501 test's script, however it was observed that placement needed to be near
1502 the beginning for it to behave that way. Changed the runtest perl
1503 script to read through the entire script once to detect and set these
1504 variables, reset to the beginning of the script, and then run through
1505 the script parsing/test process like normal.
1507 TL/02 The BSD's have an arc4random API. One of the functions to induce
1508 adding randomness was arc4random_stir(), but it has been removed in
1509 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1510 function when detected.
1512 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1513 cause callback expansion.
1515 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1516 syntax errors in an expansion can be treated as a string instead of
1517 logging or causing an error, due to the internal use of bool_lax
1518 instead of bool when processing it.
1520 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1521 server certificates when making smtp deliveries.
1523 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1525 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1527 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1529 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1530 Merged patch from Sebastian Wiedenroth.
1532 JH/05 Fix results-pipe from transport process. Several recipients, combined
1533 with certificate use, exposed issues where response data items split
1534 over buffer boundaries were not parsed properly. This eventually
1535 resulted in duplicates being sent. This issue only became common enough
1536 to notice due to the introduction of connection certificate information,
1537 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1539 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1540 size buffer was used, resulting in syntax errors when an expansion
1543 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1544 version 3.3.6 or later.
1546 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1547 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1548 both become "event_action", the variables become $event_name, $event_data
1549 and $event_defer_errno. There is a new variable $verify_mode, usable in
1550 routers, transports and related events. The tls:cert event is now also
1551 raised for inbound connections, if the main configuration event_action
1554 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1555 early OCSP support, but no stapling (appears to be less than 1.0.0).
1557 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1558 server certificate names available under the smtp transport option
1559 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1562 JH/10 Time-related extraction expansions from certificates now use the main
1563 option "timezone" setting for output formatting, and are consistent
1564 between OpenSSL and GnuTLS compilations. Bug 1541.
1566 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1567 encoded parameter in the incoming message. Bug 1558.
1569 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1570 include certificate info, eximon was claiming there were spoolfile
1573 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1575 JH/14 Log delivery-related information more consistently, using the sequence
1576 "H=<name> [<ip>]" wherever possible.
1578 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1579 are problematic for Debian distribution, omit them from the release
1582 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1584 JH/16 Fix string representation of time values on 64bit time_t architectures.
1587 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1588 output list separator was used.
1593 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1594 checkers that were complaining about end of non-void function with no
1597 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1598 This was a regression introduced in 4.83 by another bugfix.
1600 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1602 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1603 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1609 TF/01 Correctly close the server side of TLS when forking for delivery.
1611 When a message was received over SMTP with TLS, Exim failed to clear up
1612 the incoming connection properly after forking off the child process to
1613 deliver the message. In some situations the subsequent outgoing
1614 delivery connection happened to have the same fd number as the incoming
1615 connection previously had. Exim would try to use TLS and fail, logging
1616 a "Bad file descriptor" error.
1618 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1619 utilities have not been installed.
1621 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1622 temporary space as the ACL may create new global variables.
1624 TL/01 LDAP support uses per connection or global context settings, depending
1625 upon the detected version of the libraries at build time.
1627 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1628 to extract and use the src ip:port in logging and expansions as if it
1629 were a direct connection from the outside internet. PPv2 support was
1630 updated based on HAProxy spec change in May 2014.
1632 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1634 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1635 Properly escape header and check for NULL return.
1637 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1640 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1642 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1643 characters in header names, implemented as a verify condition.
1644 Contributed by Michael Fischer v. Mollard.
1646 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1647 results permerror and temperror. Previous values are deprecated but
1648 still accepted. In a future release, err_perm and err_temp will be
1649 completely removed, which will be a backward incompatibility if the
1650 ACL tests for either of these two old results. Patch contributed by
1651 user bes-internal on the mailing list.
1653 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1655 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1656 selectors, in both main and reject logs.
1658 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1661 JH/07 Add malware type "sock" for talking to simple daemon.
1663 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1665 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1666 routers/transports under cutthrough routing.
1668 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1669 numbers. Touch up "bool" conditional to keep the same definition.
1671 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1673 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1675 JH/12 Expand items in router/transport headers_add or headers_remove lists
1676 individually rather than the list as a whole. Bug 1452.
1678 Required for reasonable handling of multiple headers_ options when
1679 they may be empty; requires that headers_remove items with embedded
1680 colons must have them doubled (or the list-separator changed).
1682 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1683 view the policy declared in the DMARC record. Currently, $dmarc_status
1684 is a combined value of both the record presence and the result of the
1687 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1689 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1690 dnslookup router and the smtp transport (applying to the forward
1693 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1694 of ldap servers used for a specific lookup. Patch provided by Heiko
1697 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1698 New variable $lookup_dnssec_authenticated for observability.
1700 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1701 Patch submitted by Lars Timman.
1703 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1705 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1706 Requires trusted mode and valid format message id, aborts otherwise.
1707 Patch contributed by Heiko Schlichting.
1709 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1710 certextract with support for various fields. Bug 1358.
1712 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1713 is requested by default, modifiable by smtp transport option
1716 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1717 operate on certificate variables to give certificate fingerprints
1718 Also new ${sha256:cert_variable}.
1720 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1722 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1723 Christian Aistleitner.
1725 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1727 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1728 file. Patch from Wolfgang Breyha.
1730 JH/25 Expand the coverage of the delivery $host and $host_address to
1731 client authenticators run in verify callout. Bug 1476.
1733 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1734 align with daemon_smtp_ports. Bug 72.
1736 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1737 support and error reporting did not work properly.
1739 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1740 and is readable. Patch from Andrew Colin Kissa.
1742 TL/14 Enhance documentation of ${run expansion and how it parses the
1743 commandline after expansion, particularly in the case when an
1744 unquoted variable expansion results in an empty value.
1746 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1748 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1749 by renaming away from it. Addresses GH issue 15, reported by
1752 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1754 TL/15 SECURITY: prevent double expansion in math comparison functions
1755 (can expand unsanitized data). Not remotely exploitable.
1762 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1764 PP/02 Make -n do something, by making it not do something.
1765 When combined with -bP, the name of an option is not output.
1767 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1770 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1771 $sender_host_name and config options to manage this, and basic check
1774 PP/05 DSCP support for outbound connections and control modifier for inbound.
1776 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1777 (Only plugin which currently uses this is kerberos4, which nobody should
1778 be using, but we should make it available and other future plugins might
1779 conceivably use it, even though it would break NAT; stuff *should* be
1780 using channel bindings instead).
1782 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1783 name; added for Sendmail compatibility; requires admin caller.
1784 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1785 just ignore it); requires trusted caller.
1786 Also parse but ignore: -Ac -Am -X<logfile>
1789 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1791 TL/02 Add +smtp_confirmation as a default logging option.
1793 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1794 Patch by Magnus Holmgren from 2007-02-20.
1796 TL/04 Bugzilla 1281 - Spec typo.
1797 Bugzilla 1283 - Spec typo.
1798 Bugzilla 1290 - Spec grammar fixes.
1800 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1802 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1804 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1805 mailing list by by Dmitry Isaikin.
1807 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1809 JH/02 Support "G" suffix to numbers in ${if comparisons.
1811 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1813 NM/01 Bugzilla 1197 - Spec typo
1814 Bugzilla 1196 - Spec examples corrections
1816 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1818 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1819 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1822 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1823 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1825 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1826 "acl {{name}{arg}...}", and optional args on acl condition
1829 JH/05 Permit multiple router/transport headers_add/remove lines.
1831 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1833 JH/07 Avoid using a waiting database for a single-message-only transport.
1834 Performance patch from Paul Fisher. Bugzilla 1262.
1836 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1839 JH/09 Add $headers_added variable, with content from use of ACL modifier
1840 add_header (but not yet added to the message). Bugzilla 199.
1842 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1843 Pulled from Bugzilla 817 by Wolfgang Breyha.
1845 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1847 (nb: this is the same fix as in Exim 4.80.1)
1849 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1852 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1855 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1857 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1858 advertise SMTP AUTH mechanism to us, instead of a generic
1859 protocol violation error. Also, make Exim more robust to bad
1860 data from the Dovecot auth socket.
1862 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1864 When a queue runner is handling a message, Exim first routes the
1865 recipient addresses, during which it prunes them based on the retry
1866 hints database. After that it attempts to deliver the message to
1867 any remaining recipients. It then updates the hints database using
1870 So if a recipient address works intermittently, it can get repeatedly
1871 deferred at routing time. The retry hints record remains fresh so the
1872 address never reaches the final cutoff time.
1874 This is a fairly common occurrence when a user is bumping up against
1875 their storage quota. Exim had some logic in its local delivery code
1876 to deal with this. However it did not apply to per-recipient defers
1877 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1879 This change adds a proper retry rule check during routing so that the
1880 final cutoff time is checked against the message's age. We only do
1881 this check if there is an address retry record and there is not a
1882 domain retry record; this implies that previous attempts to handle
1883 the address had the retry_use_local_parts option turned on. We use
1884 this as an approximation for the destination being like a local
1885 delivery, as in LMTP.
1887 I suspect this new check makes the old local delivery cutoff check
1888 redundant, but I have not verified this so I left the code in place.
1890 TF/02 Correct gecos expansion when From: is a prefix of the username.
1892 Test 0254 submits a message to Exim with the header
1896 When I ran the test suite under the user fanf2, Exim expanded
1897 the header to contain my full name, whereas it should have added
1898 a Resent-Sender: header. It erroneously treats any prefix of the
1899 username as equal to the username.
1901 This change corrects that bug.
1903 GF/01 DCC debug and logging tidyup
1904 Error conditions log to paniclog rather than rejectlog.
1905 Debug lines prefixed by "DCC: " to remove any ambiguity.
1907 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1909 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1910 Bug spotted by Jeremy Harris; was flawed since initial commit.
1911 Would have resulted in OCSP responses post-SNI triggering an Exim
1912 NULL dereference and crash.
1914 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1916 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1917 Bug detection, analysis and fix by Samuel Thibault.
1918 Bugzilla 1331, Debian bug #698092.
1920 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1922 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1923 Server implementation by Todd Lyons, client by JH.
1924 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1925 config variable "prdr_enable" controls whether the server
1926 advertises the facility. If the client requests PRDR a new
1927 acl_data_smtp_prdr ACL is called once for each recipient, after
1928 the body content is received and before the acl_smtp_data ACL.
1929 The client is controlled by both of: a hosts_try_prdr option
1930 on the smtp transport, and the server advertisement.
1931 Default client logging of deliveries and rejections involving
1932 PRDR are flagged with the string "PRDR".
1934 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1935 fclose(). Diagnosis by Todd Lyons.
1937 PP/17 Update configure.default to handle IPv6 localhost better.
1938 Patch by Alain Williams (plus minor tweaks).
1941 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1942 This is now consistent with GnuTLS, and is now documented: the
1943 previous undocumented portable approach to treating the option as
1944 unset was to force an expansion failure. That still works, and
1945 an empty string is now equivalent.
1947 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1948 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1949 not performing validation itself.
1951 PP/20 Added force_command boolean option to pipe transport.
1952 Patch from Nick Koston, of cPanel Inc.
1954 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1957 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1959 PP/21 Fix eximon continuous updating with timestamped log-files.
1960 Broken in a format-string cleanup in 4.80, missed when I repaired the
1961 other false fix of the same issue.
1962 Report and fix from Heiko Schlichting.
1965 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1966 Report from Prashanth Katuri.
1968 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1969 It's SecureTransport, so affects any MacOS clients which use the
1970 system-integrated TLS libraries, including email clients.
1972 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1973 using a MIME ACL for non-SMTP local injection.
1974 Report and assistance in diagnosis by Warren Baker.
1976 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1978 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1980 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1981 last id that failed so it may be referenced in subsequent ACL's.
1983 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1986 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1987 ldap library initialization, allowing self-signed CA's to be
1988 used. Also properly sets require_cert option later in code by
1989 using NULL (global ldap config) instead of ldap handle (per
1990 session). Bug diagnosis and testing by alxgomz.
1992 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1993 the src/util/ subdirectory.
1995 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1996 renamed to Transport Post Delivery Action by Jeremy Harris, as
1999 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2000 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2001 redis_servers = needs to be configured which will be used by the redis
2002 lookup. Patch from Warren Baker, of The Packet Hub.
2004 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2006 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2007 hostname or reverse DNS when processing a host list. Used suggestions
2008 from multiple comments on this bug.
2010 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2012 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2013 Missed a few lines, added it to make the runtest require no keyboard
2016 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2017 contains upper case chars. Make router use caseful_local_part.
2019 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2020 support when GnuTLS has been built with p11-kit.
2026 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2028 This, or similar/improved, will also be change PP/11 of 4.82.
2034 PP/01 Handle short writes when writing local log-files.
2035 In practice, only affects FreeBSD (8 onwards).
2036 Bugzilla 1053, with thanks to Dmitry Isaikin.
2038 NM/01 Bugzilla 949 - Documentation tweak
2040 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2043 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2045 PP/02 Implemented gsasl authenticator.
2047 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2049 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2050 `pkg-config foo` for cflags/libs.
2052 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2053 with rest of GSASL and with heimdal_gssapi.
2055 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2056 `pkg-config foo` for cflags/libs for the TLS implementation.
2058 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2059 properties get this fed in as external SSF. A number of robustness
2060 and debugging improvements to the cyrus_sasl authenticator.
2062 PP/08 cyrus_sasl server now expands the server_realm option.
2064 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2065 Patch by Jeremy Harris.
2067 PP/10 Added dbmjz lookup type.
2069 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2071 PP/12 MAIL args handles TAB as well as SP, for better interop with
2072 non-compliant senders.
2073 Analysis and variant patch by Todd Lyons.
2075 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2076 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2077 Patch from Dirk Mueller <dmueller@suse.com>
2079 PP/13 tls_peerdn now print-escaped for spool files.
2080 Observed some $tls_peerdn in wild which contained \n, which resulted
2081 in spool file corruption.
2083 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2084 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2085 or write after TLS renegotiation, which otherwise led to messages
2088 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2089 as a tracking header (ie: a signed header comes before the signature).
2090 Patch from Wolfgang Breyha.
2092 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2093 comma-sep list; embedded commas doubled.
2095 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2097 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2099 Report and patch from Dmitry Banschikov.
2101 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2102 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2103 failures. We appear to now support TLS1.1+ with Exim.
2105 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2106 lets Exim select keys and certificates based upon TLS SNI from client.
2107 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2108 before an outbound SMTP session. New log_selector, +tls_sni.
2110 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2111 NULL dereference. Report and patch from Alun Jones.
2113 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2114 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2115 Not seeing resolver debug output on NetBSD, but suspect this is a
2116 resolver implementation change.
2118 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2119 Left warnings. Added "eximon gdb" invocation mode.
2121 PP/21 Defaulting "accept_8bitmime" to true, not false.
2123 PP/22 Added -bw for inetd wait mode support.
2125 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2126 locate the relevant includes and libraries. Made this the default.
2128 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2129 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2131 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2132 This may cause build issues on older platforms.
2134 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2135 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2136 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2137 Added SNI support via GnuTLS too.
2138 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2140 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2142 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2143 Applied second patch from Janne, implementing suggestion to default
2144 multiple-strings-in-record handling to match SPF spec.
2146 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2148 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2149 read-only, out of scope).
2150 Patch from Wolfgang Breyha, report from Stuart Northfield.
2152 PP/29 Fix three issues highlighted by clang analyser static analysis.
2153 Only crash-plausible issue would require the Cambridge-specific
2154 iplookup router and a misconfiguration.
2155 Report from Marcin Mirosław.
2157 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2159 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2160 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2161 As part of this, removing so much warning spew let me fix some minor
2162 real issues in debug logging.
2164 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2165 assignment on my part. Fixed.
2167 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2168 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2169 Janne Snabb (who went above and beyond: thank you).
2171 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2172 string otherwise requires a connection and a bunch more work and it's
2173 relatively easy to get wrong. Should also expose TLS library linkage
2176 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2177 64-bit ${eval} (JH/03).
2179 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2180 GNU libc to support some of the 64-bit stuff, should not lead to
2181 conflicts. Defined before os.h is pulled in, so if a given platform
2182 needs to override this, it can.
2184 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2185 protection layer was required, which is not implemented.
2186 Bugzilla 1254, patch from Wolfgang Breyha.
2188 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2189 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2190 tls_dhparam take prime identifiers. Also unbreak combination of
2191 OpenSSL+DH_params+TLSSNI.
2193 PP/39 Disable SSLv2 by default in OpenSSL support.
2199 PP/01 Solaris build fix for Oracle's LDAP libraries.
2200 Bugzilla 1109, patch from Stephen Usher.
2202 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2204 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2207 TF/02 Fix a couple more cases where we did not log the error message
2208 when unlink() failed. See also change 4.74-TF/03.
2210 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2211 lock up or crash if it happened to be inside a call to libc when it
2212 got a SIGUSR1 from exiwhat.
2214 The SIGUSR1 handler appends the current process status to the process
2215 log which is later printed by exiwhat. It used to use the general
2216 purpose logging code to do this, but several functions it calls are
2217 not safe for signals.
2219 The new output code in the SIGUSR1 handler is specific to the process
2220 log, and simple enough that it's easy to inspect for signal safety.
2221 Removing some special cases also simplifies the general logging code.
2222 Removing the spurious timestamps from the process log simplifies
2225 TF/04 Improved ratelimit ACL condition.
2227 The /noupdate option has been deprecated in favour of /readonly which
2228 has clearer semantics. The /leaky, /strict, and /readonly update modes
2229 are mutually exclusive. The update mode is no longer included in the
2230 database key; it just determines when the database is updated. (This
2231 means that when you upgrade Exim will forget old rate measurements.)
2233 Exim now checks that the per_* options are used with an update mode that
2234 makes sense for the current ACL. For example, when Exim is processing a
2235 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2236 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2237 must specify per_mail/readonly. If you omit the update mode it defaults to
2238 /leaky where that makes sense (as before) or /readonly where required.
2240 The /noupdate option is now undocumented but still supported for
2241 backwards compatibility. It is equivalent to /readonly except that in
2242 ACLs where /readonly is required you may specify /leaky/noupdate or
2243 /strict/noupdate which are treated the same as /readonly.
2245 A useful new feature is the /count= option. This is a generalization
2246 of the per_byte option, so that you can measure the throughput of other
2247 aggregate values. For example, the per_byte option is now equivalent
2248 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2250 The per_rcpt option has been generalized using the /count= mechanism
2251 (though it's more complicated than the per_byte equivalence). When it is
2252 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2253 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2254 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2255 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2256 Note that using per_rcpt with a non-readonly update mode in more than
2257 one ACL will cause the recipients to be double-counted. (The per_mail
2258 and per_byte options don't have this problem.)
2260 The handling of very low rates has changed slightly. If the computed rate
2261 is less than the event's count (usually one) then this event is the first
2262 after a long gap. In this case the rate is set to the same as this event's
2263 count, so that the first message of a spam run is counted properly.
2265 The major new feature is a mechanism for counting the rate of unique
2266 events. The new per_addr option counts the number of different
2267 recipients that someone has sent messages to in the last time period. It
2268 behaves like per_rcpt if all the recipient addresses are different, but
2269 duplicate recipient addresses do not increase the measured rate. Like
2270 the /count= option this is a general mechanism, so the per_addr option
2271 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2272 example, measure the rate that a client uses different sender addresses
2273 with the options per_mail/unique=$sender_address. There are further
2274 details in the main documentation.
2276 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2278 TF/06 Removed a few PCRE remnants.
2280 TF/07 Automatically extract Exim's version number from tags in the git
2281 repository when doing development or release builds.
2283 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2284 Bugzilla 879. Patch from Paul Fisher.
2286 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2287 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2290 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2292 PP/05 Variable $av_failed, true if the AV scanner deferred.
2293 Bugzilla 1078. Patch from John Horne.
2295 PP/06 Stop make process more reliably on build failure.
2296 Bugzilla 1087. Patch from Heiko Schlittermann.
2298 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2299 Bugzilla 1089. Patch from Heiko Schlittermann.
2301 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2302 Bugzilla 1131. Patch from Holger Weiß.
2304 PP/09 Handle IPv6 addresses with SPF.
2305 Bugzilla 860. Patch from Wolfgang Breyha.
2307 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2309 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2312 PP/11 match_* no longer expand right-hand-side by default.
2313 New compile-time build option, EXPAND_LISTMATCH_RHS.
2314 New expansion conditions, "inlist", "inlisti".
2316 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2318 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2320 PP/14 fix log_write() format string regression from TF/03.
2321 Bugzilla 1152. Patch from Dmitry Isaikin.
2327 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2329 PP/02 Harmonised TLS library version reporting; only show if debugging.
2330 Layout now matches that introduced for other libraries in 4.74 PP/03.
2332 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2334 PP/04 New "dns_use_edns0" global option.
2336 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2339 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2340 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2342 TK/01 Updated PolarSSL code to 0.14.2.
2343 Bugzilla 1097. Patch from Andreas Metzler.
2345 PP/07 Catch divide-by-zero in ${eval:...}.
2346 Fixes bugzilla 1102.
2348 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2351 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2352 format-string attack -- SECURITY: remote arbitrary code execution.
2354 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2355 time unintentionally subject to list matching rules, letting the header
2356 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2357 arbitrary string expansion). This allowed for information disclosure.
2359 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2360 INT_MIN/-1 -- value coerced to INT_MAX.
2366 NM/01 Workaround for PCRE version dependency in version reporting
2369 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2370 This fixes portability to compilers other than gcc, notably
2371 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2373 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2374 makefiles for portability to HP-UX and POSIX correctness.
2376 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2377 Also via indented variable definition in the Makefile.
2378 (Debugging by Oliver Heesakkers).
2380 PP/02 Restore caching of spamd results with expanded spamd_address.
2381 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2383 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2384 Improves build reliability. Fix from: Frank Elsner
2386 NM/02 Fix wide character breakage in the rfc2047 coding
2387 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2389 NM/03 Allow underscore in dnslist lookups
2390 Fixes bug 1026. Patch from Graeme Fowler
2392 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2393 Code patches from Adam Ciarcinski of NetBSD.
2395 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2398 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2399 is logged, to avoid truncation. Patch from John Horne.
2401 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2402 Patch from Jakob Hirsch.
2404 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2405 SQL string expansion failure details.
2406 Patch from Andrey Oktyabrski.
2408 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2409 Patch from Simon Arlott.
2411 PP/09 New lookups functionality failed to compile on old gcc which rejects
2412 extern declarations in function scope.
2413 Patch from Oliver Fleischmann
2415 PP/10 Use sig_atomic_t for flags set from signal handlers.
2416 Check getgroups() return and improve debugging.
2417 Fixed developed for diagnosis in bug 927 (which turned out to be
2420 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2421 Patch from Mark Zealey.
2423 PP/12 Bugzilla 1056: Improved spamd server selection.
2424 Patch from Mark Zealey.
2426 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2427 Based on patch from Heiko Schlittermann.
2429 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2430 Patch from Uwe Doering, sign-off by Michael Haardt.
2432 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2433 variable declaration deep within a block. Bug and patch from
2436 PP/15 lookups-Makefile IRIX compatibility coercion.
2438 PP/16 Make DISABLE_DKIM build knob functional.
2440 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2441 Patch by Simon Arlott
2443 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2444 variable argument macros. Our copy now differs from upstream.
2450 TF/01 Failure to get a lock on a hints database can have serious
2451 consequences so log it to the panic log.
2453 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2454 controlled using the smtp_confirmation log selector.
2456 TF/03 Include the error message when we fail to unlink a spool file.
2458 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2459 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2460 for maintaining out-of-tree patches for some time.
2462 PP/01 Bugzilla 139: Documentation and portability issues.
2463 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2464 Handle per-OS dynamic-module compilation flags.
2466 PP/02 Let /dev/null have normal permissions.
2467 The 4.73 fixes were a little too stringent and complained about the
2468 permissions on /dev/null. Exempt it from some checks.
2469 Reported by Andreas M. Kirchwitz.
2471 PP/03 Report version information for many libraries, including
2472 Exim version information for dynamically loaded libraries. Created
2473 version.h, now support a version extension string for distributors
2474 who patch heavily. Dynamic module ABI change.
2476 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2477 privilege escalation vulnerability whereby the Exim run-time user
2478 can cause root to append content of the attacker's choosing to
2481 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2484 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2485 If dropping privileges for untrusted macros, we disabled normal logging
2486 on the basis that it would fail; for the Exim run-time user, this is not
2487 the case, and it resulted in successful deliveries going unlogged.
2488 Fixed. Reported by Andreas Metzler.
2494 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2495 only prepend for the Resent-* case. Fixes regression introduced in
2496 Exim 4.70 by NM/22 for Bugzilla 607.
2498 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2499 increasing numbers of administrators be bitten by this.
2501 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2503 PP/03 Bugzilla 994: added openssl_options main configuration option.
2505 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2507 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2509 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2511 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2512 without a peer certificate, leading to a segfault because of an
2513 assumption that peers always have certificates. Be a little more
2514 paranoid. Problem reported by Martin Tscholak.
2516 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2517 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2518 NB: ClamAV planning to remove STREAM in "middle of 2010".
2519 CL also introduces -bmalware, various -d+acl logging additions and
2520 more caution in buffer sizes.
2522 PP/09 Implemented reverse_ip expansion operator.
2524 PP/10 Bugzilla 937: provide a "debug" ACL control.
2526 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2528 PP/12 Bugzilla 973: Implement --version.
2530 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2532 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2534 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2536 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2537 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2538 ignore trailing whitespace.
2540 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2542 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2545 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2546 Notification from Dr Andrew Aitchison.
2548 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2549 ExtendedDetectionInfo response format.
2550 Notification from John Horne.
2552 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2555 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2556 XSL and documented dependency on system catalogs, with examples of how
2559 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2562 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2563 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2564 configuration file which is writeable by the Exim user or group.
2566 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2567 of configuration files to cover files specified with the -C option if
2568 they are going to be used with root privileges, not just the default
2571 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2572 option (effectively making it always true).
2574 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2575 files to be used while preserving root privileges.
2577 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2578 that rogue child processes cannot use them.
2580 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2581 run-time user, instead of root.
2583 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2584 Exim run-time user without dropping privileges.
2586 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2587 result string, instead of calling string_vformat() twice with the same
2590 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2591 for other users. Others should always drop root privileges if they use
2592 -C on the command line, even for a whitelisted configure file.
2594 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2596 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2602 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2603 $data_path, and $header_path variables; fixed documentation bugs and
2606 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2607 exipick to access non-standard spools, including the "frozen" queue
2610 NM/01 Bugzilla 965: Support mysql stored procedures.
2611 Patch from Alain Williams
2613 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2615 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2616 Patch from Andreas Metzler
2618 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2619 Patch from Kirill Miazine
2621 NM/05 Bugzilla 671: Added umask to procmail example.
2623 JJ/03 installed exipick 20100323.0, fixing doc bug
2625 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2626 directory. Notification and patch from Dan Rosenberg.
2628 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2630 TK/02 Improve log output when DKIM signing operation fails.
2632 MH/01 Treat the transport option dkim_domain as a colon separated
2633 list, not as a single string, and sign the message with each element,
2634 omitting multiple occurences of the same signer.
2636 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2637 Bugzilla 985, 986. Patch by Simon Arlott
2639 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2640 Patch by Simon Arlott
2642 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2643 MBX locking. Notification from Dan Rosenberg.
2649 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2651 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2653 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2655 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2657 NM/04 Bugzilla 915: Flag broken perl installation during build.
2663 TK/01 Added patch by Johannes Berg that expands the main option
2664 "spamd_address" if it starts with a dollar sign.
2666 TK/02 Write list of recipients to X-Envelope-Sender header when building
2667 the mbox-format spool file for content scanning (suggested by Jakob
2670 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2671 (http://www.dcc-servers.net/) support via dccifd. Activated by
2672 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2674 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2675 by Mark Daniel Reidel <mr@df.eu>.
2677 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2678 When building exim an external PCRE library is now needed -
2679 PCRE is a system library on the majority of modern systems.
2680 See entry on PCRE_LIBS in EDITME file.
2682 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2683 conversation. Added nologin parameter to request.
2684 Patch contributed by Kirill Miazine.
2686 TF/01 Do not log submission mode rewrites if they do not change the address.
2688 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2690 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2691 log files in place. Contributed by Roberto Lima.
2693 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2695 TF/03 Bugzilla 615: When checking the local_parts router precondition
2696 after a local_part_suffix or local_part_prefix option, Exim now
2697 does not use the address's named list lookup cache, since this
2698 contains cached lookups for the whole local part.
2700 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2701 Robert Millan. Documentation is in experimental-spec.txt.
2703 TF/04 Bugzilla 668: Fix parallel build (make -j).
2705 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2707 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2708 Patch provided by Jan Srzednicki.
2710 TF/05 Leading white space used to be stripped from $spam_report which
2711 wrecked the formatting. Now it is preserved.
2713 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2714 that they are available at delivery time.
2716 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2718 TF/08 TLS error reporting now respects the incoming_interface and
2719 incoming_port log selectors.
2721 TF/09 Produce a more useful error message if an SMTP transport's hosts
2722 setting expands to an empty string.
2724 NM/06 Bugzilla 744: EXPN did not work under TLS.
2725 Patch provided by Phil Pennock.
2727 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2728 Patch provided by Richard Godbee.
2730 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2731 acl_smtp_notquit, added index entry.
2733 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2734 Patch provided by Eugene Bujak.
2736 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2737 accept(). Patch provided by Maxim Dounin.
2739 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2740 Patch provided by Phil Pennock.
2742 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2744 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2745 Patch provided by Brad "anomie" Jorsch.
2747 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2748 Patch provided by Dean Brooks.
2750 TK/05 Add native DKIM support (does not depend on external libraries).
2752 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2753 Patch provided by Graeme Fowler.
2755 NM/16 Bugzilla 851: Documentation example syntax fix.
2757 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2759 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2762 NM/19 Bugzilla 745: TLS version reporting.
2763 Patch provided by Phil Pennock.
2765 NM/20 Bugzilla 167: bool: condition support.
2766 Patch provided by Phil Pennock.
2768 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2769 clients. Patch provided by Phil Pennock.
2771 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2772 Patch provided by Brad "anomie" Jorsch.
2774 NM/23 Bugzilla 687: Fix misparses in eximstats.
2775 Patch provided by Heiko Schlittermann.
2777 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2778 Patch provided by Heiko Schlittermann.
2780 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2781 plus update to original patch.
2783 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2785 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2786 Patch provided by David Brownlee.
2788 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2790 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2792 NM/30 Bugzilla 888: TLS documentation bugfixes.
2794 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2796 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2797 Unlike the original bugzilla I have changed all shell scripts in src tree.
2799 NM/33 Bugzilla 898: Transport filter timeout fix.
2800 Patch by Todd Rinaldo.
2802 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2803 Patch by Serge Demonchaux.
2805 NM/35 Bugzilla 39: Base64 decode bug fixes.
2806 Patch by Jakob Hirsch.
2808 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2810 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2812 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2814 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2820 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2821 ALT-N's libdkim that I have put here:
2822 http://duncanthrax.net/exim-experimental/
2824 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2825 were called 'true' and it seems that C99 defines that as a reserved
2826 keyword to be used with 'bool' variable types. That means you could
2827 not include C99-style headers which use bools without triggering
2828 build errors in sieve.c.
2830 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2831 as mailq or other aliases. Changed the --help handling significantly
2832 to do whats expected. exim_usage() emits usage/help information.
2834 SC/01 Added the -bylocaldomain option to eximstats.
2836 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2838 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2840 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2846 PH/01 Another patch from the Sieve maintainer.
2848 PH/02 When an IPv6 address is converted to a string for single-key lookup
2849 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2850 dots are used instead of colons so that keys in lsearch files need not
2851 contain colons. This was done some time before quoting was made available
2852 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2853 (notated using the quote facility) so as to distinguish them from IPv4
2854 keys. This meant that lookups for IP addresses in host lists did not work
2855 for iplsearch lookups.
2857 This has been fixed by arranging for IPv6 addresses to be expressed with
2858 colons if the lookup type is iplsearch. This is not incompatible, because
2859 previously such lookups could never work.
2861 The situation is now rather anomalous, since one *can* have colons in
2862 ordinary lsearch keys. However, making the change in all cases is
2863 incompatible and would probably break a number of configurations.
2865 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2868 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2869 conversion specification without a maximum field width, thereby enabling
2870 a rogue spamd server to cause a buffer overflow. While nobody in their
2871 right mind would setup Exim to query an untrusted spamd server, an
2872 attacker that gains access to a server running spamd could potentially
2873 exploit this vulnerability to run arbitrary code as the Exim user.
2875 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2876 $primary_hostname instead of what libspf2 thinks the hosts name is.
2878 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2879 a directory entry by the name of the lookup key. Previously, if a
2880 symlink pointed to a non-existing file or a file in a directory that
2881 Exim lacked permissions to read, a lookup for a key matching that
2882 symlink would fail. Now it is enough that a matching directory entry
2883 exists, symlink or not. (Bugzilla 503.)
2885 PH/03 The body_linecount and body_zerocount variables are now exported in the
2888 PH/04 Added the $dnslist_matched variable.
2890 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2891 This means they are set thereafter only if the connection becomes
2894 PH/06 Added the client_condition to authenticators so that some can be skipped
2895 by clients under certain conditions.
2897 PH/07 The error message for a badly-placed control=no_multiline_responses left
2898 "_responses" off the end of the name.
2900 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2902 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2903 (without spaces) instead of just copying the configuration text.
2905 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2907 PH/11 Added $max_received_linelength.
2909 PH/12 Added +ignore_defer and +include_defer to host lists.
2911 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2912 way in which PCRE > 7.0 is built.
2914 PH/14 Implemented queue_only_load_latch.
2916 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2917 MAIL command. The effect was to mangle the value on 64-bit systems.
2919 PH/16 Another patch from the Sieve maintainer.
2921 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2923 PH/18 If a system quota error occurred while trying to create the file for
2924 a maildir delivery, the message "Mailbox is full" was not appended to the
2925 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2926 only to a quota excession during the actual writing of the file.
2928 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2929 characters?) which causes problems in log lines. The DN values are now
2930 passed through string_printing() before being added to log lines.
2932 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2933 and InterBase are left for another time.)
2935 PH/21 Added message_body_newlines option.
2937 PH/22 Guard against possible overflow in moan_check_errorcopy().
2939 PH/23 POSIX allows open() to be a macro; guard against that.
2941 PH/24 If the recipient of an error message contained an @ in the local part
2942 (suitably quoted, of course), incorrect values were put in $domain and
2943 $local_part during the evaluation of errors_copy.
2949 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2950 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2953 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2954 issue a MAIL command.
2956 PH/02 In an ACL statement such as
2958 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2960 if a client was not listed at all, or was listed with a value other than
2961 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2962 the condition was not true (as it should be), so access was not denied.
2963 The bug was that the ! inversion was incorrectly passed on to the second
2964 item. This has been fixed.
2966 PH/03 Added additional dnslists conditions == and =& which are different from
2967 = and & when the dns lookup returns more than one IP address.
2969 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2970 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2972 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2973 FSYNC, which compiles an option called disable_fsync that allows for
2974 bypassing fsync(). The documentation is heavily laced with warnings.
2976 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2978 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2979 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2980 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2981 including adding "make clean"; (3) Added -fPIC when compiling the test
2982 dynamically loaded module, to get rid of a warning.
2984 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2985 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2986 = 0s. The bug is otherwise harmless.
2988 PH/07 There was a bug in the dovecot authenticator such that the value of
2989 $auth1 could be overwritten, and so not correctly preserved, after a
2990 successful authentication. This usually meant that the value preserved by
2991 the server_setid option was incorrect.
2993 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2995 PH/09 Installed PCRE release 7.0.
2997 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2998 run for batched SMTP input. It is now run at the start of every message
2999 in the batch. While fixing this I discovered that the process information
3000 (output by running exiwhat) was not always getting set for -bs and -bS
3001 input. This is fixed, and it now also says "batched" for BSMTP.
3003 PH/11 Added control=no_pipelining.
3005 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3006 patch, slightly modified), and move the expansion of helo_data till after
3007 the connection is made in the smtp transport (so it can use these
3010 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3012 PH/14 Added log_selector = +pid.
3014 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3016 PH/16 Add ${if forany and ${if forall.
3018 PH/17 Added dsn_from option to vary the From: line in DSNs.
3020 PH/18 Flush SMTP output before performing a callout, unless control =
3021 no_callout_flush is set.
3023 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3024 was true (the default) a successful delivery failed to delete the retry
3025 item, thus causing premature timeout of the address. The bug is now
3028 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3030 PH/21 Long custom messages for fakedefer and fakereject are now split up
3031 into multiline responses in the same way that messages for "deny" and
3032 other ACL rejections are.
3034 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3035 with slight modification.
3037 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3038 draft, changing the syntax and factoring some duplicate code".
3040 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3041 for deliveries of the second and subsequent messages over the same SMTP
3044 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3045 ${reduce, with only minor "tidies".
3047 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3049 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3050 expansion side effects.
3052 PH/27 When a message times out after an over-quota error from an Exim-imposed
3053 quota, the bounce message says "mailbox is full". This message was not
3054 being given when it was a system quota that was exceeded. It now should
3057 MH/03 Made $recipients available in local_scan(). local_scan() already has
3058 better access to the recipient list through recipients_list[], but
3059 $recipients can be useful in postmaster-provided expansion strings.
3061 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3062 in the case of a MAIL command with additional options following the
3063 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3064 were accidentally chopped off.
3066 PH/29 SMTP synchronization checks are implemented when a command is read -
3067 there is a check that no more input is waiting when there shouldn't be
3068 any. However, for some commands, a delay in an ACL can mean that it is
3069 some time before the response is written. In this time, more input might
3070 arrive, invalidly. So now there are extra checks after an ACL has run for
3071 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3072 pipelining has not been advertised.
3074 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3076 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3077 returned in the SMTP response when smtp_return_error_details was false.
3078 This has been fixed.
3080 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3081 instead of the C I/O that was originally supplied, because problems were
3082 reported on Solaris.
3084 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3085 Exim which did not show up earlier: it was assuming that a call to
3086 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3087 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3088 was a macro that became an assignment, so it seemed to work. This has
3089 changed to a proper function call with a void return, hence the compile
3090 error. Exim's code has been fixed.
3092 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3095 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3097 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3099 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3100 because it turns out to be too big or there is a timeout) there is no
3101 "Completed" line in the log. When some messages of this type were
3102 selected by exigrep, they were listed as "not completed". Others were
3103 picked up by some special patterns. I have improved the selection
3104 criteria to be more general.
3106 PH/38 The host_find_failed option in the manualroute router can now be set
3107 to "ignore", to completely ignore a host whose IP address cannot be
3108 found. If all hosts are ignored, the behaviour is controlled by the new
3109 host_all_ignored option.
3111 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3112 homing or because of multiple MX records with /mx) generated more than
3113 one IP address, and the following item turned out to be the local host,
3114 all the secondary addresses of the first item were incorrectly removed
3115 from the list, along with the local host and any following hosts (which
3116 is what is supposed to happen).
3118 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3119 whoever called Exim into the -H file. In the case of the daemon it was
3120 behaving confusingly. When first started, it used values for whoever
3121 started the daemon, but after a SIGHUP it used the Exim user (because it
3122 calls itself on a restart). I have changed the code so that it now always
3125 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3126 message are rejected with the same error (e.g. no authentication or bad
3127 sender address), and a DATA command is nevertheless sent (as can happen
3128 with PIPELINING or a stupid MUA), the error message that was given to the
3129 RCPT commands is included in the rejection of the DATA command. This is
3130 intended to be helpful for MUAs that show only the final error to their
3133 PH/42 Another patch from the Sieve maintainer.
3135 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3136 Eximstats - Fixed some broken HTML links and added missing column headers
3138 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3139 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3141 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3143 PH/43 Yet another patch from the Sieve maintainer.
3145 PH/44 I found a way to check for a TCP/IP connection going away before sending
3146 the response to the final '.' that terminates a message, but only in the
3147 case where the client has not sent further data following the '.'
3148 (unfortunately, this is allowed). However, in many cases there won't be
3149 any further data because there won't be any more messages to send. A call
3150 to select() can be used: if it shows that the input is "ready", there is
3151 either input waiting, or the socket has been closed. An attempt to read
3152 the next input character can distinguish the two cases. Previously, Exim
3153 would have sent an OK response which the client would never have see.
3154 This could lead to message repetition. This fix should cure that, at
3155 least in a lot of common cases.
3157 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3158 advertised in response to EHLO.
3164 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3165 fixed by 4.65/MH/01 (is this a record?) are fixed:
3167 (i) An empty string was always treated as zero by the numeric comparison
3168 operators. This behaviour has been restored.
3170 (ii) It is documented that the numeric comparison operators always treat
3171 their arguments as decimal numbers. This was broken in that numbers
3172 starting with 0 were being interpreted as octal.
3174 While fixing these problems I realized that there was another issue that
3175 hadn't been noticed. Values of message_size_limit (both the global option
3176 and the transport option) were treated as octal if they started with 0.
3177 The documentation was vague. These values are now always treated as
3178 decimal, and I will make that clear in the documentation.
3184 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3185 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3188 MH/01 Don't check that the operands of numeric comparison operators are
3189 integers when their expansion is in "skipping" mode (fixes bug
3190 introduced by 4.64-PH/07).
3192 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3193 child addresses, Exim now panics and dies. Previously, because the count
3194 is held in a short int, deliveries were likely to be lost. As such a
3195 large number of recipients for a single message is ridiculous
3196 (performance will be very, very poor), I have chosen to impose a limit
3197 rather than extend the field.
3203 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3204 leftover -K file (the existence of which was triggered by #402).
3205 While we were at it, introduced process PID as part of the -K
3206 filename. This should rule out race conditions when creating
3209 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3210 processing considerably. Previous code took too long for large mails,
3211 triggering a timeout which in turn triggers #401.
3213 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3214 in the DK code in transports.c. sendfile() is not really portable,
3215 hence the _LINUX specificness.
3217 TF/01 In the add_headers option to the mail command in an Exim filter,
3218 there was a bug that Exim would claim a syntax error in any
3219 header after the first one which had an odd number of characters
3222 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3223 callout verification, Exim cached a "reject" for the entire domain. This
3224 is correct for most verifications, but it is not correct for a recipient
3225 verification with use_sender or use_postmaster set, because in that case
3226 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3227 case of MAIL FROM:<> rejection from other early rejections (e.g.
3228 rejection of HELO). When verifying a recipient using a non-null MAIL
3229 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3230 Whatever the result of the callout, the value of the domain cache is
3231 left unchanged (for any other kind of callout, getting as far as trying
3232 RCPT means that the domain itself is ok).
3234 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3237 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3238 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3239 session, the daemon ignores ECONNECT errors and logs others; it now
3240 ignores EPIPE as well.
3242 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3243 (quoted-printable decoding).
3245 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3246 later the small subsequent patch to fix an introduced bug.
3248 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3250 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3252 PH/08 An error is now given if message_size_limit is specified negative.
3254 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3255 to be given (somewhat) arbitrary names.
3257 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3260 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3261 miscellaneous code fixes
3263 PH/10 Added the log_reject_target ACL modifier to specify where to log
3266 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3267 hostname. This is wrong, because it relates to the incoming message (and
3268 probably the interface on which it is arriving) and not to the outgoing
3269 callout (which could be using a different interface). This has been
3270 changed to use the value of the helo_data option from the smtp transport
3271 instead - this is what is used when a message is actually being sent. If
3272 there is no remote transport (possible with a router that sets up host
3273 addresses), $smtp_active_hostname is used.
3275 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3276 tweaks were necessary in order to get it to work (see also 21 below):
3277 (a) The code assumed that strncpy() returns a negative number on buffer
3278 overflow, which isn't the case. Replaced with Exim's string_format()
3280 (b) There were several signed/unsigned issues. I just did the minimum
3281 hacking in of casts. There is scope for a larger refactoring.
3282 (c) The code used strcasecmp() which is not a standard C function.
3283 Replaced with Exim's strcmpic() function.
3284 (d) The code set only $1; it now sets $auth1 as well.
3285 (e) A simple test gave the error "authentication client didn't specify
3286 service in request". It would seem that Dovecot has changed its
3287 interface. Fortunately there's a specification; I followed it and
3288 changed what the client sends and it appears to be working now.
3290 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3293 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3294 address A is aliased to B and C, where B exists and C does not. Without
3295 -v the output is "A verified" because verification stops after a
3296 successful redirection if more than one address is generated. However,
3297 with -v the child addresses are also verified. Exim was outputting "A
3298 failed to verify" and then showing the successful verification for C,
3299 with its parentage. It now outputs "B failed to verify", showing B's
3300 parentage before showing the successful verification of C.
3302 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3303 look up a TXT record in a specific list after matching in a combined
3306 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3307 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3308 they consult the DNS. I had assumed they would set it the way they
3309 wanted; and indeed my experiments on Linux seem to show that in some
3310 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3311 To be on the safe side, however, I have now made the interface to
3312 host_find_byname() similar to host_find_bydns(), with an argument
3313 containing the DNS resolver options. The host_find_byname() function now
3314 sets these options at its start, just as host_find_bydns() does. The smtp
3315 transport options dns_qualify_single and dns_search_parents are passed to
3316 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3317 of host_find_byname() use the default settings of RES_DEFNAMES
3318 (qualify_single) but not RES_DNSRCH (search_parents).
3320 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3321 spool_read_header() do less string testing, by means of a preliminary
3322 switch on the second character of optional "-foo" lines. (This is
3323 overdue, caused by the large number of possibilities that now exist.
3324 Originally there were few.) While I was there, I also converted the
3325 str(n)cmp tests so they don't re-test the leading "-" and the first
3326 character, in the hope this might squeeze out yet more improvement.
3328 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3329 flag allowing group syntax was set by the header_syntax check but not
3330 turned off, possible causing trouble later; (2) The flag was not being
3331 set at all for the header_verify test, causing "group"-style headers to
3332 be rejected. I have now set it in this case, and also caused header_
3333 verify to ignore an empty address taken from a group. While doing this, I
3334 came across some other cases where the code for allowing group syntax
3335 while scanning a header line wasn't quite right (mostly, not resetting
3336 the flag correctly in the right place). These bugs could have caused
3337 trouble for malformed header lines. I hope it is now all correct.
3339 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3340 with the "reply" argument non-NULL. The code, however (which originally
3341 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3342 but it didn't always do it. This confused somebody who was copying the
3343 code for some other use. I have removed all the tests.
3345 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3346 feature that was used to support insecure browsers during the U.S. crypto
3347 embargo. It requires special client support, and Exim is probably the
3348 only MTA that supported it -- and would never use it because real RSA is
3349 always available. This code has been removed, because it had the bad
3350 effect of slowing Exim down by computing (never used) parameters for the
3351 RSA_EXPORT functionality.
3353 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3354 authenticator to fail if there's a tab character in the incoming data
3355 (there should never be unless someone is messing about, as it's supposed
3356 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3357 if the connection is using TLS or if the remote IP is the same as the
3358 local IP, and the "valid-client-cert option" if a client certificate has
3361 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3362 authenticators. This can be used for authorization after authentication
3363 succeeds. (In the case of plaintext, it servers for both authentication
3366 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3367 if any retry times were supplied.
3369 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3370 connection, where there is no client IP address to check. In this
3371 situation, the verify now always succeeds.
3373 PH/25 Applied John Jetmore's -Mset patch.
3375 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3377 PH/27 In a string expansion for a processed (not raw) header when multiple
3378 headers of the same name were present, leading whitespace was being
3379 removed from all of them, but trailing whitespace was being removed only
3380 from the last one. Now trailing whitespace is removed from each header
3381 before concatenation. Completely empty headers in a concatenation (as
3382 before) are ignored.
3384 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3385 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3387 PH/29 [Removed. This was a change that I later backed out, and forgot to
3388 correct the ChangeLog entry (that I had efficiently created) before
3389 committing the later change.]
3391 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3392 address errors (4xx response to RCPT) over the same connection as other
3393 messages routed to the same hosts. Such deliveries are always "forced",
3394 so retry times are not inspected. This resulted in far too many retries
3395 for the affected addresses. The effect occurred only when there were more
3396 hosts than the hosts_max_try setting in the smtp transport when it had
3397 the 4xx errors. Those hosts that it had tried were not added to the list
3398 of hosts for which the message was waiting, so if all were tried, there
3399 was no problem. Two fixes have been applied:
3401 (i) If there are any address or message errors in an SMTP delivery, none
3402 of the hosts (tried or untried) are now added to the list of hosts
3403 for which the message is waiting, so the message should not be a
3404 candidate for sending over the same connection that was used for a
3405 successful delivery of some other message. This seems entirely
3406 reasonable: after all the message is NOT "waiting for some host".
3407 This is so "obvious" that I'm not sure why it wasn't done
3408 previously. Hope I haven't missed anything, but it can't do any
3409 harm, as the worst effect is to miss an optimization.
3411 (ii) If, despite (i), such a delivery is accidentally attempted, the
3412 routing retry time is respected, so at least it doesn't keep
3413 hammering the server.
3415 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3416 in ${readsocket because some servers need this prod.
3418 PH/32 Added some extra debug output when updating a wait-xxx database.
3420 PH/33 The hint "could be header name not terminated by colon", which has been
3421 given for certain expansion errors for a long time, was not being given
3422 for the ${if def:h_colon_omitted{... case.
3424 PH/34 The spec says: "With one important exception, whenever a domain list is
3425 being scanned, $domain contains the subject domain." There was at least
3426 one case where this was not true.
3428 PH/35 The error "getsockname() failed: connection reset by peer" was being
3429 written to the panic log as well as the main log, but it isn't really
3430 panic-worthy as it just means the connection died rather early on. I have
3431 removed the panic log writing for the ECONNRESET error when getsockname()
3434 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3435 runs only) independently of the message's sender address. This meant
3436 that, if the 4xx error was in fact related to the sender, a different
3437 message to the same recipient with a different sender could confuse
3438 things. In particular, this can happen when sending to a greylisting
3439 server, but other circumstances could also provoke similar problems.
3440 I have changed the default so that the retry time for these errors is now
3441 based a combination of the sender and recipient addresses. This change
3442 can be overridden by setting address_retry_include_sender=false in the
3445 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3446 remote server are returned as part of bounce messages. This was not
3447 happening for LMTP over a pipe (the lmtp transport), but now it is the
3448 same for both kinds of LMTP.
3450 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3451 in header lines that were in fact CNAMEs. This is no longer the case.
3453 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3454 and queue runs started by the daemon processed all messages. This has
3455 been fixed so that -R and -S can now usefully be given with -q<time>.
3457 PH/40 Import PCRE release 6.7 (fixes some bugs).
3459 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3461 PH/42 Give an error if -q is specified more than once.
3463 PH/43 Renamed the variables $interface_address and $interface_port as
3464 $received_ip_address and $received_port, to make it clear that these
3465 values apply to message reception, and not to the outgoing interface when
3466 a message is delivered. (The old names remain recognized, of course.)
3468 PH/44 There was no timeout on the connect() call when using a Unix domain
3469 socket in the ${readsocket expansion. There now is.
3471 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3472 be meaningful with "accept".
3474 SC/01 Eximstats V1.43
3475 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3477 SC/02 Eximstats V1.44
3478 Use a glob alias rather than an array ref in the generated
3479 parser. This improves both readability and performance.
3481 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3482 Collect SpamAssassin and rejection statistics.
3483 Don't display local sender or destination tables unless
3484 there is data to show.
3485 Added average volumes into the top table text output.
3487 SC/04 Eximstats V1.46
3488 Collect data on the number of addresses (recipients)
3489 as well as the number of messages.
3491 SC/05 Eximstats V1.47
3492 Added 'Message too big' to the list of mail rejection
3493 reasons (thanks to Marco Gaiarin).
3495 SC/06 Eximstats V1.48
3496 Mainlog lines which have GMT offsets and are too short to
3497 have a flag are now skipped.
3499 SC/07 Eximstats V1.49 (Alain Williams)
3500 Added the -emptyok flag.
3502 SC/08 Eximstats V1.50
3503 Fixes for obtaining the IP address from reject messages.
3505 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3506 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3507 whitespace changes from 4.64-PH/27
3509 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3512 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3513 are found, allow negative numbers in numeric criteria)
3515 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3517 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3518 to match changes made in 4.64-PH/43
3520 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3522 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3523 "domain" is actually an IP address, and force a failure. This locks out
3524 those revolvers/nameservers that support "A-for-A" lookups, in
3525 contravention of the specifications.
3527 PH/48 When a host name was looked up from an IP address, and the subsequent
3528 forward lookup of the name timed out, the host name was left in
3529 $sender_host_name, contrary to the specification.
3531 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3532 restricted to single-key lookups, Exim was not diagnosing an error if
3533 * or *@ was used with a query-style lookup.
3535 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3537 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3538 long ago, but noone interested enough thought of it. Let's just say that
3539 the "1.1" means that there are some new functions that weren't there at
3540 some point in the past.
3542 PH/51 Error processing for expansion failure of helo_data from an smtp
3543 transport during callout processing was broken.
3545 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3546 tested/used via the -bh/-bhc/-bs options.
3548 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3549 bug, fixed in subsequent PCRE releases).
3551 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3552 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3554 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3560 SC/01 Use a glob alias rather than an array ref in eximstats generated
3561 parser. This improves both readability and performance.
3563 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3564 Don't display local sender or destination tables in eximstats unless
3565 there is data to show.
3566 Added average volumes into the eximstats top table text output.
3568 SC/03 Collect data on the number of addresses (recipients) as well
3569 as the number of messages in eximstats.
3571 TF/01 Correct an error in the documentation for the redirect router. Exim
3572 does (usually) call initgroups() when daemonizing.
3574 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3575 with consistent privilege compared to when running as a daemon.
3577 TF/03 Note in the spec that $authenticated_id is not set for local
3578 submissions from trusted users.
3580 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3581 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3583 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3584 by adding some example configuration directives to the default
3585 configuration file. A little bit of work is required to uncomment the
3586 directives and define how usernames and passwords are checked, but
3587 there is now a framework to start from.
3589 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3590 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3591 without this. I don't know how relevant this is to other LDAP libraries.
3593 PH/02 Add the verb name to the "unknown ACL verb" error.
3595 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3597 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3599 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3600 directory not expanded when it should be if an expanded home directory
3601 was set for the address (which is overridden by the transport).
3603 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3606 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3607 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3608 because it is too late at that time, and has no effect.
3610 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3611 security issue with \' (bugzilla #107). I could not use the
3612 PQescapeStringConn() function, because it needs a PGconn value as one of
3615 PH/08 When testing addresses using -bt, indicate those final addresses that
3616 are duplicates that would not cause an additional delivery. At least one
3617 person was confused, thinking that -bt output corresponded to deliveries.
3618 (Suppressing duplicates isn't a good idea as you lose the information
3619 about possibly different redirections that led to the duplicates.)
3621 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3622 systems where poll() doesn't work, in particular OS X.
3624 PH/10 Added more information to debugging output for retry time not reached.
3626 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3627 operations in malware.c.
3629 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3632 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3633 syslog_duplication set false, log lines that would normally be written
3634 both the the main log and to the reject log were not written to syslog at
3637 PH/14 In the default configuration, change the use of "message" in ACL warn
3638 statements to "add_header".
3640 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3641 not followed by a command (e.g. "seen endif").
3643 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3644 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3647 PH/17 Added extra conditions to the default value of delay_warning_condition
3651 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3652 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3653 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3656 The Auto-Submitted: and various List- headers are standardised, whereas I
3657 don't think Precedence: ever was.
3659 PH/18 Refactored debugging code in route_finduser() to show more information,
3660 in particular, the error code if getpwnam() issues one.
3662 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3663 This is apparently needed in addition to the PH/07 change above to avoid
3664 any possible encoding problems.
3666 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3667 but not after initializing Perl.
3669 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3670 output them only if debugging. By default they are written stderr,
3671 apparently, which is not desirable.
3673 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3676 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3679 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3681 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3682 authenticated or an ident call has been made. Suppress the default
3683 values for $authenticated_id and $authenticated_sender (but permit -oMai
3684 and -oMas) when testing with -bh.
3686 PH/24 Re-jigged the order of the tests in the default configuration so that the
3687 tests for valid domains and recipients precede the DNS black list and CSA
3688 tests, on the grounds that those ones are more expensive.
3690 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3691 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3692 This bug exists in every version of Exim that I still have, right back to
3695 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3696 However, an attempt to turn on case-sensitivity in a regex key by
3697 including (?-i) didn't work because the subject string was already
3698 lowercased, and the effects were non-intuitive. It turns out that a
3699 one-line patch can be used to allow (?-i) to work as expected.
3705 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3706 other effects) broke the use of negated acl sub-conditions.
3708 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3711 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3712 "Deny" causes Exim to reject the incoming connection with a 554 error.
3713 Unfortunately, if there is a major crisis, such as a disk failure,
3714 tcp-wrappers gives "deny", whereas what one would like would be some
3715 kind of temporary error. A kludge has been added to help with this.
3716 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3717 554 error is used if errno is still zero or contains ENOENT (which occurs
3718 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3721 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3723 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3724 errors. Otherwise a message that provokes a temporary error (when other
3725 messages do not) can cause a whole host to time out.
3727 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3728 addresses were routed directly to files or pipes from a redirect router.
3729 File deliveries just didn't batch; pipe deliveries might have suffered
3732 PH/06 A failure to get a lock for a hints database would erroneously always say
3733 "Failed to get write lock", even when it was really a read lock.
3735 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3736 of 0600. This has been changed to use the value of the lockfile_mode
3737 option (which defaults to 0600).
3739 PH/08 Applied small patch from the Sieve maintainer.
3741 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3742 folder from quota calculations, a direct delivery into this folder messed
3743 up the contents of the maildirsize file. This was because the regex was
3744 used only to exclude .Trash (or whatever) when the size of the mailbox
3745 was calculated. There was no check that a delivery was happening into an
3746 excluded directory. This bug has been fixed by ignoring all quota
3747 processing for deliveries into excluded directories.
3749 PH/10 Added the maildirfolder_create_regex option to appendfile.
3755 PH/01 The code for finding all the local interface addresses on a FreeBSD
3756 system running IPv6 was broken. This may well have applied to all BSD
3757 systems, as well as to others that have similar system calls. The broken
3758 code found IPv4 interfaces correctly, but gave incorrect values for the
3759 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3760 that it would not match correctly against @[] and not recognize the IPv6
3763 PH/02 The ipliteral router was not recognizing addresses of the form user@
3764 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3766 PH/03 Added disable_ipv6.
3768 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3769 decoded form, because it is most often used to construct To: headers
3770 lines in autoreplies, and the decoded form may well be syntactically
3771 invalid. However, $reply_address has leading white space removed, and all
3772 newlines turned into spaces so that the autoreply transport does not
3775 PH/05 If group was specified without a user on a router, and no group or user
3776 was specified on a transport, the group from the router was ignored.
3778 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3779 for visible compile-time settings that can be used to change these
3780 numbers, for those that want even more. Backwards compatibility with old
3781 spool files has been maintained. However, going back to a previous Exim
3782 release will lost any variables that are in spool files.
3784 PH/07 Two small changes when running in the test harness: increase delay when
3785 passing a TCP/IP connection to a new process, in case the original
3786 process has to generate a bounce, and remove special handling of
3787 127.0.0.2 (sic), which is no longer necessary.
3789 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3790 be the same on different OS.
3792 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3795 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3796 whether --show-vars was specified or not
3798 JJ/02 exipick: Added support for new ACL variable spool format introduced
3801 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3802 syntactically invalid From: or Reply-to: line, and a filter used this to
3803 generate an autoreply, and therefore failed to obtain an address for the
3804 autoreply, Exim could try to deliver to a non-existent relative file
3805 name, causing unrelated and misleading errors. What now happens is that
3806 it logs this as a hard delivery error, but does not attempt to create a
3809 PH/11 The exinext utility has a -C option for testing purposes, but although
3810 the given file was scanned by exinext itself; it wasn't being passed on
3811 when Exim was called.
3813 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3814 an end-of-file indication when reading a command response.
3816 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3817 compiled. In many other places in Exim, IPv6 addresses are always
3818 recognized, so I have changed this. It also means that IPv4 domain
3819 literals of the form [IPV4:n.n.n.n] are now always recognized.
3821 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3822 used if the router is not running as root, for example, when verifying at
3823 ACL time, or when using -bh. The debugging output from this situation was
3824 non-existent - all you got was a failure to exec. I have made two
3827 (a) Failures to set uid/gid, the current directory, or a process leader
3828 in a subprocess such as that created by queryprogram now generate
3829 suitable debugging output when -d is set.
3831 (b) The queryprogram router detects when it is not running as root,
3832 outputs suitable debugging information if -d is set, and then runs
3833 the subprocess without attempting to change uid/gid.
3835 PH/15 Minor change to Makefile for building test_host (undocumented testing
3838 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3839 additional section of a DNS packet that returns MX or SRV records.
3840 Instead, it always explicitly searches for A/AAAA records. This avoids
3841 major problems that occur when a DNS server includes only records of one
3842 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3843 fixed another bug: if SRV records were looked up and the corresponding
3844 address records were *not* found in the additional section, the port
3845 values from the SRV records were lost.
3847 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3848 using the correct key (the original address) when searching the retry
3849 rules in order to find which one to use for generating the retry hint.
3851 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3852 adding the default one. Similarly, if it contains a Reply-To: header, the
3853 errors_reply_to option, if set, is not used.
3855 PH/19 When calculating a retry time, Exim used to measure the "time since
3856 failure" by looking at the "first failed" field in the retry record. Now
3857 it does not use this if it is later than than the arrival time of the
3858 message. Instead it uses the arrival time. This makes for better
3859 behaviour in cases where some deliveries succeed, thus re-setting the
3860 "first failed" field. An example is a quota failure for a huge message
3861 when small messages continue to be delivered. Without this change, the
3862 "time since failure" will always be short, possible causing more frequent
3863 delivery attempts for the huge message than are intended.
3864 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3866 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3867 $1, $2, $3) because the numerical variables can be reset during some
3868 expansion items (e.g. "match"), thereby losing the authentication data.
3870 PH/21 Make -bV show the size of off_t variables so that the test suite can
3871 decide whether to run tests for quotas > 2G.
3873 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3874 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3875 filecount value is greater than 2G or if a quota value is greater than 2G
3876 on a system where the size of off_t is not greater than 4, a panic error
3879 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3880 never match. The debug and -bh output now contains an explicit error
3881 message indicating a malformed IPv4 address or mask.
3883 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3884 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3885 PH/23 above applies.
3887 PH/25 Do not write to syslog when running in the test harness. The only
3888 occasion when this arises is a failure to open the main or panic logs
3889 (for which there is an explicit test).
3891 PH/26 Added the /no_tell option to "control=freeze".
3893 PH/27 If a host name lookup failed very early in a connection, for example, if
3894 the IP address matched host_lookup and the reverse lookup yielded a name
3895 that did not have a forward lookup, an error message of the form "no IP
3896 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3897 could be logged. Now it outputs the IP address instead of "NULL".
3899 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3900 allows the sender and the authenticated sender to be set when
3901 submitting a message from within Exim. Since child_open_exim() is
3902 documented for local_scan(), the new function should be too.
3904 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3905 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3906 results in an empty string is now treated as unset.
3908 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3910 PH/31 Added sender_verify_fail logging option.
3912 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3913 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3914 tidied the source and removed it altogether.
3916 PH/33 When a queue run was abandoned because the load average was too high, a
3917 log line was always written; now it is written only if the queue_run log
3918 selector is set. In addition, the log line for abandonment now contains
3919 information about the queue run such as the pid. This is always present
3920 in "start" and "stop" lines but was omitted from the "abandon" line.
3922 PH/34 Omit spaces between a header name and the colon in the error message that
3923 is given when verify = headers_syntax fails (if there are lots of them,
3924 the message gets confusing).
3926 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3927 names, as there are now some PTR records that contain slashes. This check
3928 is only to protect against broken name servers that fall over on strange
3929 characters, so the fact that it applies to all lookups doesn't matter.
3931 PH/36 Now that the new test suite is complete, we can remove some of the
3932 special code in Exim that was needed for the old test suite. For example,
3933 sorting DNS records because real resolvers return them in an arbitrary
3934 order. The new test suite's fake resolver always returns records in the
3937 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3938 bounces) except when queue_only is set, to avoid logging races between
3939 the different processes.
3941 PH/38 Panic-die if .include specifies a non-absolute path.
3943 PH/39 A tweak to the "H" retry rule from its user.
3945 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3946 a label. They prevented compilation on older perls.
3948 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3949 a warning to be raised on newish perls.
3951 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3952 on queue. Changes to match documented behaviour of showing count of
3953 messages matching specified criteria.
3955 PH/40 Changed the default ident timeout from 30s to 5s.
3957 PH/41 Added support for the use of login_cap features, on those BSD systems
3958 that have them, for controlling the resources used by pipe deliveries.
3960 PH/42 The content-scanning code uses fopen() to create files in which to put
3961 message data. Previously it was not paying any attention to the mode of
3962 the files. Exim runs with umask(0) because the rest of the code creates
3963 files with open(), and sets the required mode explicitly. Thus, these
3964 files were ending up world-writeable. This was not a big issue, because,
3965 being within the spool directory, they were not world-accessible. I have
3966 created a function called modefopen, which takes an additional mode
3967 argument. It sets umask(777), creates the file, chmods it to the required
3968 mode, then resets the umask. All the relevant calls to fopen() in the
3969 content scanning code have been changed to use this function.
3971 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3972 to 24 hours. This avoids potential overflow problems when processing G
3973 and H retry rules. I suspect nobody ever tinkers with this value.
3975 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3977 PH/45 When the plaintext authenticator is running as a client, the server's
3978 challenges are checked to ensure they are valid base64 strings. By
3979 default, the authentication attempt is cancelled if an invalid string is
3980 received. Setting client_ignore_invalid_base64 true ignores these errors.
3981 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3982 they are received. Thus, the responses can be made to depend on the
3983 challenges. If an invalid string is ignored, an empty string is placed in
3986 PH/46 Messages that are created by the autoreply transport now contains a
3987 References: header, in accordance with RFCs 2822 and 3834.
3989 PH/47 Added authenticated_sender_force to the smtp transport.
3991 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3993 PH/49 Installed latest patch from the Sieve maintainer.
3995 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3996 was also set, the appendfile transport was unnecessarily scanning a
3997 directory of message files (e.g. for maildir delivery) to find the count
3998 of files (along with the size), even though it did not need this
3999 information. It now does the scan only if it needs to find either the
4000 size of the count of files.
4002 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4004 PH/52 Two bugs concerned with error handling when the smtp transport is
4007 (i) Exim was not creating retry information for temporary errors given
4008 for individual recipients after the DATA command when the smtp transport
4009 was used in LMTP mode. This meant that they could be retried too
4010 frequently, and not timed out correctly.
4012 (ii) Exim was setting the flag that allows error details to be returned
4013 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4014 recipients that were returned after the DATA command.
4016 PH/53 This is related to PH/52, but is more general: for any failing address,
4017 when detailed error information was permitted to be returned to the
4018 sender, but the error was temporary, then after the final timeout, only
4019 "retry timeout exceeded" was returned. Now it returns the full error as
4020 well as "retry timeout exceeded".
4022 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4023 do this, and (what is worse) MTAs that accept it.
4025 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4026 will now be deprecated.
4028 PH/56 New os.c-cygwin from the Cygwin maintainer.
4030 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4031 formats (previously only available in exim formats via -bpr, -bpru,
4032 and -bpra. Now also available in native and exiqgrep formats)
4034 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4035 with very large, slow to parse queues
4037 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4039 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4041 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4042 responses to authentication challenges, though it was showing the
4043 challenges; (ii) I've removed the CR characters from the debug output for
4046 PH/58 Allow for the insertion of a newline as well as a space when a string
4047 is turned into more than one encoded-word during RFC 2047 encoding. The
4048 Sieve code now uses this.
4050 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4051 data_4xx, lost_connection, tls_required.
4053 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4054 message was being sent as an SMTP response.
4056 PH/61 Add -l and -k options to exicyclog.
4058 PH/62 When verifying, if an address was redirected to one new address, so that
4059 verification continued, and the new address failed or deferred after
4060 having set something in $address_data, the value of $address_data was not
4061 passed back to the ACL. This was different to the case when no
4062 redirection occurred. The value is now passed back in both cases.
4064 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4065 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4066 use login_cap.h, so on its own it isn't the distinguishing feature. The
4067 new name refers directly to the setclassresources() function.
4069 PH/65 Added configuration files for NetBSD3.
4071 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4073 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4074 is preferred over IPv4.
4076 PH/68 The bounce_return_message and bounce_return_body options were not being
4077 honoured for bounces generated during the reception of non-SMTP messages.
4078 In particular, this applied to messages rejected by the ACL. This bug has
4079 been fixed. However, if bounce_return_message is true and bounce_return_
4080 body is false, the headers that are returned for a non-SMTP message
4081 include only those that have been read before the error was detected.
4082 (In the case of an ACL rejection, they have all been read.)
4084 PH/69 The HTML version of the specification is now built in a directory called
4085 spec_html instead of spec.html, because the latter looks like a path with
4086 a MIME-type, and this confuses some software.
4088 PH/70 Catch two compiler warnings in sieve.c.
4090 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4091 function verify_get_ident() calls ip_connect() to connect a socket, but
4092 if the "connect()" function timed out, ip_connect() used to close the
4093 socket. However, verify_get_ident() also closes the socket later, and in
4094 between Exim writes to the log, which may get opened at this point. When
4095 the socket was closed in ip_connect(), the log could get the same file
4096 descriptor number as the socket. This naturally causes chaos. The fix is
4097 not to close the socket in ip_connect(); the socket should be closed by
4098 the function that creates it. There was only one place in the code where
4099 this was missing, in the iplookup router, which I don't think anybody now
4100 uses, but I've fixed it anyway.
4102 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4103 well as to direct DNS lookups. Otherwise the handling of names in host
4104 lists is inconsistent and therefore confusing.
4110 PH/01 Two changes to the default runtime configuration:
4112 (1) Move the checks for relay_from_hosts and authenticated clients from
4113 after to before the (commented out) DNS black list checks.
4115 (2) Add control=submission to the relay_from_hosts and authenticated
4116 clients checks, on the grounds that messages accepted by these
4117 statements are most likely to be submissions.
4119 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4121 (1) Generate an error if the third argument for the ${prvs expansion is
4124 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4127 (3) Reset the variables that are obtained from the first argument of
4128 ${prvscheck and used in the second argument before leaving the code,
4129 because their memory is reclaimed, so using them afterwards may do
4132 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4133 one (it's much easier than Tom thought :-).
4135 (5) Because of (4), we can now allow for the use of $prvscheck_result
4136 inside the third argument.
4138 PH/03 For some reason, the default setting of PATH when running a command from
4139 a pipe transport was just "/usr/bin". I have changed it to
4142 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4143 anything to be listed in the output from -bV.
4145 PH/05 When a filter generated an autoreply, the entire To: header line was
4146 quoted in the delivery log line, like this:
4148 => >A.N.Other <ano@some.domain> <original@ddress> ...
4150 This has been changed so that it extracts the operative address. There
4151 may be more than one such address. If so, they are comma-separated, like
4154 => >ano@some.domain,ona@other.domain <original@ddress> ...
4156 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4157 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4158 not being looked up in the rDNS to get a host name, Exim was showing the
4159 IP address twice in Received: lines, even though the IP addresses were
4160 identical. For example:
4162 Received: from [1.2.3.4] (helo=[1.2.3.4])
4164 However, if the real host name was known, it was omitting the HELO data
4165 if it matched the actual IP address. This has been tidied up so that it
4166 doesn't show the same IP address twice.
4168 PH/07 When both +timestamp and +memory debugging was on, the value given by
4169 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4170 called by the debug printing, thereby overwriting the timestamp buffer.
4171 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4173 PH/08 When the original message was included in an autoreply transport, it
4174 always said "this is a copy of the message, including all the headers",
4175 even if body_only or headers_only was set. It now gives an appropriate
4178 PH/09 Applied a patch from the Sieve maintainer which:
4180 o fixes some comments
4181 o adds the (disabled) notify extension core
4182 o adds some debug output for the result of if/elsif tests
4183 o points to the current vacation draft in the documentation
4184 and documents the missing references header update
4188 o fixes a bug in processing the envelope test (when testing
4189 multiple envelope elements, the last element determined the
4192 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4193 Electronic Mail") by including:
4195 Auto-submitted: auto-generated
4197 in the messages that it generates (bounce messages and others, such as
4198 warnings). In the case of bounce messages for non-SMTP messages, there was
4199 also a typo: it was using "Auto_submitted" (underscore instead of
4200 hyphen). Since every message generated by Exim is necessarily in response
4201 to another message, thes have all been changed to:
4203 Auto-Submitted: auto-replied
4205 in accordance with these statements in the RFC:
4207 The auto-replied keyword:
4209 - SHOULD be used on messages sent in direct response to another
4210 message by an automatic process,
4212 - MUST NOT be used on manually-generated messages,
4214 - MAY be used on Delivery Status Notifications (DSNs) and Message
4215 Disposition Notifications (MDNs),
4217 - MUST NOT be used on messages generated by automatic or periodic
4218 processes, except for messages which are automatic responses to
4221 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4222 to the default Received: header definition.
4224 PH/12 Added log selector acl_warn_skipped (default on).
4226 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4227 variables because (a) they are in the wrong storage pool and (b) even if
4228 they were copied, it wouldn't work properly because of the caching.
4230 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4231 checking when decoding. Apparently there are clients that generate
4232 overlong encoded strings. Why am I not surprised?
4234 PH/15 If the first argument of "${if match_address" was not empty, but did not
4235 contain an "@" character, Exim crashed. Now it writes a panic log message
4236 and treats the condition as false.
4238 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4240 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4241 extension "envelope-auth". The code is finished and in agreement with
4242 other implementations, but there is no documentation so far and in fact,
4243 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4244 not changing the active code.
4246 Print executed "if" and "elsif" statements when debugging is used. This
4247 helps a great deal to understand what a filter does.
4249 Document more things not specified clearly in RFC3028. I had all this
4250 sorted out, when out of a sudden new issues came to my mind. Oops."
4252 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4255 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4256 canonical form (as documented). However, after a host name lookup from
4257 the IP address, check_host() was doing a simple string comparison with
4258 addresses acquired from the DNS when checking that the found name did
4259 have the original IP as one of its addresses. Since any found IPv6
4260 addresses are likely to be in abbreviated form, the comparison could
4261 fail. Luckily, there already exists a function for doing the comparison
4262 by converting both addresses to binary, so now that is used instead of
4263 the text comparison.
4265 PH/20 There was another similar case to PH/19, when a complete host name was
4266 given in a host list; looking up its IP address could give an abbreviated
4267 form, whereas the current host's name might or might not be abbreviated.
4268 The same fix has been applied.
4274 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4275 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4278 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4279 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4281 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4283 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4284 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4285 Also, for "Auto-Submitted", treat anything other than "no" as
4286 non-personal, in accordance with RFC 3834. (Previously it treated
4287 anything starting "auto-" as non-personal.)
4289 TF/01 The control=submission/name=... option had a problem with syntax
4290 errors if the name included a slash character. The /name= option
4291 now slurps the rest of the string, so it can include any characters
4292 but it must come last in the list of options (after /sender_retain
4295 PH/05 Some modifications to the interface to the fake nameserver for the new
4303 TK/01 Added the "success_on_redirect" address verification option. See
4304 NewStuff for rationale and an example.
4306 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4308 PH/02 Patch to exigrep to allow it to work on syslog lines.
4310 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4311 fread() to skip over the body file's header line, because in Cygwin the
4312 header line is locked and is inaccessible.
4314 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4315 co-exist for some time) to make it clear that it is the Exim ID that is
4316 referenced, not the Message-ID: header line.
4318 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4319 string_format() function, because snprintf() does not exist on all
4322 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4323 ${stat: expansion item.
4325 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4326 protocol synchronization error", to keep the pedants happy.
4328 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4329 well as for IRIX systems, when gcc is being used. See the host.c source
4332 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4334 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4337 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4338 transport and to the smtp transport in LMTP mode.
4340 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4342 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4343 run a filter in a subprocess. This could lead to confusion in subsequent
4344 lookups in the parent process. There should also be a search_tidyup() at
4345 the end of the subprocess.
4347 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4348 only if the host matched helo_try_verify_hosts, which caused the
4349 verification to occur when the EHLO/HELO command was issued. The ACL just
4350 tested the remembered result. Now, if a previous verification attempt has
4351 not happened, "verify = helo" does it there and then.
4353 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4355 TK/03 Fix log output including CR from clamd.
4357 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4358 exist provoked a memory error which could cause a segfault.
4360 PH/15 Installed PCRE 6.2
4362 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4364 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4365 of the problem. Specifically, suggested +O2 rather than +O1 for the
4368 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4370 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4371 record, the port was not correctly logged when the outgoing_port log
4372 selector was set (it logged the transort's default port).
4374 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4375 fallback_hosts, and "hosts" in the smtp transport.
4377 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4378 host errors such as "Connection refused".
4380 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4381 authentication with radiusclient 0.4.9:
4383 - Error returned from rc_read_config was caught wrongly
4384 - Username/password not passed on to radius server due to wrong length.
4386 The presumption is that some radiusclient API changes for 4.51/PH/17
4387 were not taken care of correctly. The code is still untested by me (my
4388 Linux distribution still has 0.3.2 of radiusclient), but it was
4389 contributed by a Radius user.
4391 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4392 expanding the "port" option of the smtp transport.
4394 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4395 while reading a MIME header. Thanks to Tom Hughes for a patch.
4397 PH/24 Include config.h inside local_scan.h so that configuration settings are
4400 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4401 that in an ACL for RCPT (for example), you can examine exactly what was
4404 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4405 commands, but it was not correctly comparing the address with the actual
4406 client host address. Thus, it would show the EHLO address in Received:
4407 header lines when this was not necessary.
4409 PH/27 Added the % operator to ${eval:}.
4411 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4412 it should be ignoring failures (because with -C, for example, it has lost
4413 privilege). It wasn't ignoring creation failures other than "already
4416 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4419 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4420 because an input line was too long, either on its own, or by virtue of
4421 too many continuations, the temporary file was not being removed, and the
4422 return code was incorrect.
4424 PH/31 Missing "BOOL" in function definition in filtertest.c.
4426 PH/32 Applied Sieve patches from the maintainer.
4428 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4430 PH/33 Added "verify = not_blind".
4432 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4433 Local/Makefile (with some defaults set). These are used in built scripts
4434 such as exicyclog, but they have never been used in the exim_install
4435 script (though there are many overriding facilities there). I have
4436 arranged that the exim_install script now takes note of these two
4439 PH/35 Installed configuration files for Dragonfly.
4441 PH/36 When a locally submitted message by a trusted user did not contain a
4442 From: header, and the sender address was obtained from -f or from an SMTP
4443 MAIL command, and the trusted user did not use -F to supply a sender
4444 name, $originator_name was incorrectly used when constructing a From:
4445 header. Furthermore, $originator_name was used for submission mode
4446 messages from external hosts without From: headers in a similar way,
4447 which is clearly wrong.
4449 PH/37 Added control=suppress_local_fixups.
4451 PH/38 When log_selector = +received_sender was set, and the addition of the
4452 sender made the log line's construction buffer exactly full, or one byte
4453 less than full, an overflow happened when the terminating "\n" was
4456 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4457 when the result of a list match is failure because a DNS lookup failed.
4459 PH/40 RM_COMMAND is now used in the building process.
4461 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4462 the "build-* directories that it finds.
4464 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4465 address in a domain literal was a prefix of an interface address.
4467 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4468 when verifying a sender address, unless rewrite_headers is false.
4470 PH/44 Wrote a long comment about why errors_to addresses are verified as
4471 recipients, not senders.
4473 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4474 the ratelimit ACL was added.
4476 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4478 PH/46 Added extra information about PostgreSQL errors to the error string.
4480 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4481 avoiding the need to install special zones in a real server. This is
4482 backwards compatible; if it can't find the fake resolver, it drops back.
4483 Thus, both old and new test suites can be run.
4485 TF/02 Added util/ratelimit.pl
4487 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4488 clock is set back in time.
4490 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4491 Candler <B.Candler@pobox.com>.
4493 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4494 OK for addresses that are the result of redirections.
4496 PH/48 A number of further additions for the benefit of the new test suite,
4497 including a fake gethostbyname() that interfaces to the fake DNS resolver
4500 TF/06 The fix for widen_domains has also been applied to qualify_single and
4501 search_parents which are the other dnslookup options that can cause
4504 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4507 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4509 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4510 service" if it encountered an MX record with an empty target hostname.
4511 The message is now "an MX or SRV record indicated no SMTP service".
4513 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4514 if the DNS of the sending site is misconfigured. This is quite a
4515 common situation. This change restores the behaviour of treating a
4516 helo verification defer as a failure.
4518 PH/51 If self=fail was set on a router, the bounce message did not include the
4519 actual error message.
4525 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4527 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4528 command itself ended in error, the underlying message about the transport
4529 filter timeout was being overwritten with the pipe command error. Now the
4530 underlying error message should be appended to the second error message.
4532 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4534 PH/02 Reduce the amount of output that "make" produces by default. Full output
4535 can still be requested.
4537 PH/03 The warning log line about a condition test deferring for a "warn" verb
4538 was being output only once per connection, rather than after each
4539 occurrence (because it was using the same function as for successful
4540 "warn" verbs). This seems wrong, so I have changed it.
4542 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4543 it should not have, which might have caused a crash in the right
4544 circumstances, but probably never did.
4546 PH/04 Installed a modified version of Tony Finch's patch to make submission
4547 mode fix the return path as well as the Sender: header line, and to
4548 add a /name= option so that you can make the user's friendly name appear
4551 TF/03 Added the control = fakedefer ACL modifier.
4553 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4554 Mark Lowes for thorough testing.
4556 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4558 TK/03 Merged latest SRS patch from Miles Wilton.
4560 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4561 with the definition in sysexits.h (which is #included earlier).
4562 Fortunately, Exim does not actually use EX_OK. The code used to try to
4563 preserve the sysexits.h value, by assuming that macro definitions were
4564 scanned for macro replacements. I have been disabused of this notion,
4565 so now the code just undefines EX_OK before #including unistd.h.
4567 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4568 in the smtp transport. When a block could not be written in a single
4569 write() function, the timeout was being re-applied to each part-write.
4570 This seems wrong - if the receiver was accepting one byte at a time it
4571 would take for ever. The timeout is now adjusted when this happens. It
4572 doesn't have to be particularly precise.
4574 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4575 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4577 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4578 without a domain if the check to <postmaster@domain> fails.
4580 SC/01 Eximstats: added -xls and the ability to specify output files
4581 (patch written by Frank Heydlauf).
4583 SC/02 Eximstats: use FileHandles for outputting results.
4585 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4587 SC/04 Eximstats: fixed display of large numbers with -nvr option
4589 SC/05 Eximstats: fixed merging of reports with empty tables.
4591 SC/06 Eximstats: added the -include_original_destination flag
4593 SC/07 Eximstats: removed tabs and trailing whitespace.
4595 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4597 TK/06 MBOX spool code: Add real "From " MBOX separator line
4598 so the .eml file is really in mbox format (even though
4599 most programs do not really care). Patch from Alex Miller.
4601 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4602 The latter is generated from $received_to and is only set if the
4603 message has one envelope recipient. SA can use these headers,
4604 obviously out-of-the-box. Patch from Alex Miller.
4606 PH/08 The ${def test on a variable was returning false if the variable's
4607 value was "0", contrary to what the specification has always said!
4608 The result should be true unless the variable is empty.
4610 PH/09 The syntax error of a character other than { following "${if
4611 def:variable_name" (after optional whitespace) was not being diagnosed.
4612 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4613 accidental colon was present, for example, could give incorrect results.
4615 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4616 result is used (not including appendfile, where other changes are about
4619 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4620 This involved changing a lot of size variables from int to off_t. It
4621 should work with maildirs and everything.
4623 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4624 spamd dying while we are connected to it.
4626 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4629 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4630 function for BDB 4.3.
4632 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4634 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4635 the sources for PCRE have been re-arranged and more files are now
4638 PH/15 The code I had for printing potentially long long variables in PH/11
4639 above was not the best (it lost precision). The length of off_t variables
4640 is now inspected at build time, and an appropriate printing format (%ld
4641 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4642 to be "long long int" or "long int". This is needed for the internal
4643 formatting function string_vformat().
4645 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4646 the configuration file to be ":syslog", then the script "guesses" where
4647 the logs files are, rather than using the compiled in default. In our
4648 case the guess is not the same as the compiled default, so the script
4649 suddenly stopped working when I started to use syslog. The patch checks
4650 to see if log_file_path is "". If so, it attempts to read it from exim
4651 with no configuration file to get the compiled in version, before it
4652 falls back to the previous guessing code."
4654 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4655 implementing BATV in an Exim configuration. See NewStuff for the gory
4658 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4659 Makefile that are specific to HP-UX.
4661 PH/18 If the "use_postmaster" option was set for a recipient callout together
4662 with the "random" option, the postmaster address was used as the MAIL
4663 FROM address for the random test, but not for the subsequent recipient
4664 test. It is now used for both.
4666 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4667 patch removes a few documentation additions to RFC 3028, because the
4668 latest draft now contains them. It adds the new en;ascii-case comparator
4669 and a new error check for 8bit text in MIME parts. Comparator and
4670 require names are now matched exactly. I enabled the subaddress
4671 extension, but it is not well tested yet (read: it works for me)."
4673 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4674 rework some of the code of TK/09 above to avoid the hardwired use of
4675 "%lld" and "long long". Replaced the call to snprintf() with a call to
4678 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4679 records point to non-existent hosts", "retry timeout exceeded", and
4680 "retry time not reached for any host after a long failure period".
4682 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4683 experimental DomainKeys support:
4685 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4686 (2) On an error such as an illegally used "control", the wrong name for
4687 the control was given.
4689 These problems did NOT occur unless DomainKeys support was compiled.
4691 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4693 PH/24 Added ${if match_ip condition.
4695 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4696 databases so that it will be absolutely obvious if a crash occurs in the
4697 DB library. This is a regular occurrence (often caused by mis-matched
4700 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4701 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4702 picked up on a user's system that detects such things. There doesn't seem
4703 to be a gcc warning option for this - only an attribute that has to be
4704 put on the function's prototype. It seems that in Fedora Core 4 they have
4705 set this on a number of new functions. No doubt there will be more in due
4708 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4709 specify a transport. However, if an address that was verified by such a
4710 router was the subject of a callout, Exim crashed because it tried to
4711 read the rcpt_include_affixes from the non-existent transport. Now it
4712 just assumes that the setting of that option is false. This bug was
4713 introduced by 4.51/PH/31.
4715 PH/28 Changed -d+all to exclude +memory, because that information is very
4716 rarely of interest, but it makes the output a lot bigger. People tend to
4717 do -d+all out of habit.
4719 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4720 code in os-type was giving problems when libc.so lives in lib64, like on
4723 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4724 aren't the modern standard, and it seems that some systems' include files
4725 don't always have them. Exim was already checking for some of the newer
4726 ones like T_AAAA, and defining it itself. I've added checks for all the
4727 record types that Exim uses.
4729 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4730 not automatically generating a new one, as it is supposed to. This
4731 prevented TLS from working. If the file did exist, but contained invalid
4732 data, a new version was generated, as expected. It was only the case of a
4733 non-existent file that was broken.
4735 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4736 with a change in libdomainkeys > 0.64.
4738 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4739 from DNS. If the selector record carries the flag, it now has
4740 precedence over the domain-wide flag.
4742 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4744 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4745 the use of a port name that isn't defined in /etc/services) occurred, the
4746 message was deferred as in a normal delivery, and thus remained on the
4747 spool, instead of being failed because of the mua_wrapper setting. This
4748 is now fixed, and I tidied up some of the mua_wrapper messages at the
4751 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4752 the messages in a hash of arrays rather than using individual hashes.
4753 This is a bit cleaner and results in dramatic memory savings, albeit
4754 at a slight CPU cost.
4756 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4757 as requested by Marc Sherman.
4759 SC/10 Eximstats: added histograms for user specified patterns as requested
4762 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4764 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4765 fopen() in the content-scanning modules that did not already have it.
4771 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4772 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4774 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4776 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4778 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4779 to test Sieve filters that use "vacation".
4781 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4782 that changes the way the GnuTLS parameters are stored in the cache file.
4783 The new format can be generated externally. For backward compatibility,
4784 if the data in the cache doesn't make sense, Exim assumes it has read an
4785 old-format file, and it generates new data and writes a new file. This
4786 means that you can't go back to an older release without removing the
4789 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4790 work if there are any delivery delays because "one_time" forces the
4791 parent to be marked "delivered", so its unseen clone is never tried
4792 again. For this reason, Exim now forbids the simultaneous setting of
4795 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4796 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4797 entry if you want to know the details. The fix, however, neglected to
4798 consider the case when local delivery batching is involved. The test for
4799 "previously delivered" was not happening when checking to see if an
4800 address could be batched with a previous (undelivered) one; under
4801 certain circumstances this could lead to multiple deliveries to the same
4804 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4805 in its include files, and this causes problems building Exim.
4807 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4808 header_syntax) but Exim was just ignoring anything given after a slash.
4809 In particular, this caused confusion with an attempt to use "verify =
4810 reverse_host_lookup/defer_ok". An error is now given when options are
4811 supplied for verify items that do not have them. (Maybe reverse_host_
4812 lookup should have a defer_ok option, but that's a different point.)
4814 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4815 defined by RFC 821) to 2048, because there were problems with some AUTH
4816 commands, and RFC 1869 says the size should be increased for extended
4817 SMTP commands that take arguments.
4819 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4822 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4823 "unknown" error; now it says that the functionality isn't in the binary.
4825 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4826 an address' error message when a string expansion fails (syntax or
4827 whatever). Otherwise the password may appear in the log. Following change
4828 PH/42 below, there is no longer a chance of it appearing in a bounce
4831 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4833 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4834 of its addresses was ever tried. (Bugzilla bug #2.)
4836 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4837 the result incorrectly in the debug output. (It correctly added a newline
4838 to what was transported.)
4840 TF/01 Added $received_time.
4842 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4843 commented out examples of how to interface to a virus scanner and to
4844 SpamAssassin. Also added commented examples of av_scanner and
4845 spamd_address settings.
4847 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4848 and controls are allowed in which ACLs. There were a couple of minor
4849 errors. Some of the entries in the conditions table (which is a table of
4850 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4851 negation of where the condition IS allowed.
4853 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4855 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4856 header file does not have a version number, so I've had to invent a new
4857 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4858 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4859 radiusclient), but it was contributed by a Radius user.
4861 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4862 files or directories, for OpenSSL.
4864 PH/19 When an Exim process that is running as root has to create an Exim log
4865 file, it does so in a subprocess that runs as exim:exim so as to get the
4866 ownership right at creation (otherwise, other Exim processes might see
4867 the file with the wrong ownership). There was no test for failure of this
4868 fork() call, which would lead to the process getting stuck as it waited
4869 for a non-existent subprocess. Forks do occasionally fail when resources
4870 run out. I reviewed all the other calls to fork(); they all seem to check
4873 PH/20 When checking for unexpected SMTP input at connect time (before writing
4874 the banner), Exim was not dealing correctly with a non-positive return
4875 from the read() function. If the client had disconnected by this time,
4876 the result was a log entry for a synchronization error with an empty
4877 string after "input=" when read() returned zero. If read() returned -1
4878 (an event I could not check), uninitialized data bytes were printed.
4879 There were reports of junk text (parts of files, etc) appearing after
4882 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4884 PH/22 Added support for macro redefinition, and (re)definition in between
4885 driver and ACL definitions.
4887 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4888 forgetting to use the resulting value; it was using the unexpanded value.
4890 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4891 hadn't been configured. The fix is from Juergen Kreileder, who
4892 understands it better than I do:
4894 "Here's what I see happening with three configured cyrus_sasl
4895 authenticators configured (plain, login, cram-md5):
4897 On startup auth_cyrus_sasl_init() gets called for each of these.
4898 This means three calls to sasl_listmech() without a specified mech_list.
4899 => SASL tests which mechs of all available mechs actually work
4900 => three warnings about OTP not working
4901 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4903 With the patch, sasl_listmech() also gets called three times. But now
4904 SASL's mech_list option is set to the server_mech specified in the the
4905 authenticator. Or in other words, the answer from sasl_listmech()
4906 gets limited to just the mech you're testing for (which is different
4908 => the return list contains just 'plain' or 'login', 'cram-md5' or
4909 nothing depending on the value of ob->server_mech.
4911 I've just tested the patch: Authentication still works fine,
4912 unavailable mechs specified in the exim configuration are still
4913 caught, and the auth.log warnings about OTP are gone."
4915 PH/25 When debugging is enabled, the contents of the command line are added
4916 to the debugging output, even when log_selector=+arguments is not
4919 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4920 answer is "GNU", and only if the return is "GNU/something" is the answer
4923 PH/27 $acl_verify_message is now set immediately after the failure of a
4924 verification in an ACL, and so is available in subsequent modifiers. In
4925 particular, the message can be preserved by coding like this:
4927 warn !verify = sender
4928 set acl_m0 = $acl_verify_message
4930 Previously, $acl_verify_message was set only while expanding "message"
4931 and "log_message" when a very denied access.
4933 PH/28 Modified OS/os.c-Linux with
4935 -#ifndef OS_LOAD_AVERAGE
4936 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4938 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4939 nomenclature these days.)
4941 PH/29 Installed patch from the Sieve maintainer that adds the options
4942 sieve_useraddress and sieve_subaddress to the redirect router.
4944 PH/30 In these circumstances:
4945 . Two addresses routed to the same list of hosts;
4946 . First host does not offer TLS;
4947 . First host accepts first address;
4948 . First host gives temporary error to second address;
4949 . Second host offers TLS and a TLS session is established;
4950 . Second host accepts second address.
4951 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4952 and peerdn, if requested) that were in fact used only for the second
4955 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4956 attention to any local part prefix or suffix that was matched by the
4957 router that accepted the address. It now behaves in the same way as it
4958 does for delivery: the affixes are removed from the local part unless
4959 rcpt_include_affixes is set on the transport.
4961 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4962 timeout during the DATA phase of an incoming message.
4964 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4965 applied a patch sanctioned by the Sieve maintainer.
4967 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4968 the uid or gid is negative. A case of a negative gid caused this to be
4969 noticed. The fix allows for either to be negative.
4971 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4972 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4975 PH/36 The change PH/12 above was broken. Fixed it.
4977 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4978 the grounds that routing the same address twice would always produce the
4979 same answer. This might have been true once, but it is certainly no
4980 longer true now. Routing a child address may depend on the previous
4981 routing that produced that child. Some complicated redirection strategies
4982 went wrong when messages had multiple recipients, and made Exim's
4983 behaviour dependent on the order in which the addresses were given.
4985 I have moved the duplicate checking until after the routing is complete.
4986 Exim scans the addresses that are assigned to local and remote
4987 transports, and removes any duplicates. This means that more work will be
4988 done, as duplicates will always all be routed, but duplicates are
4989 presumably rare, so I don't expect this is of any significance.
4991 For deliveries to pipes, files, and autoreplies, the duplicate checking
4992 still happens during the routing process, since they are not going to be
4995 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4996 It corrects a timeout issue with spamd. This is Ian's comment: "The
4997 background is that sometimes spamd either never reads data from a
4998 connection it has accepted, or it never writes response data. The exiscan
4999 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5000 blindly assumes that writes won't block so it may never time out."
5002 PH/39 Allow G after quota size as well as K and M.
5004 PH/40 The value set for $authenticated_id in an authenticator may not contain
5005 binary zeroes or newlines because the value is written to log lines and
5006 to spool files. There was no check on this. Now the value is run through
5007 the string_printing() function so that such characters are converted to
5008 printable escape sequences.
5010 PH/41 $message_linecount is a new variable that contains the total number of
5011 lines in the message. Compare $body_linecount, which is the count for the
5014 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5015 bounce and delay warning messages, except in certain special cases, which
5018 (a) An SMTP error message from a remote host;
5019 (b) A message specified in a :fail: redirection;
5020 (c) A message specified in a "fail" command in a system filter;
5021 (d) A message specified in a FAIL return from the queryprogram router;
5022 (e) A message specified by the cannot_route_message router option.
5024 In these cases only, Exim does include the error details in bounce and
5025 warning messages. There are also a few cases where bland messages such
5026 as "unrouteable address" or "local delivery error" are given.
5028 PH/43 $value is now also set for the "else" part of a ${run expansion.
5030 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5031 being worked on, but at least Exim now implements the latest version to
5034 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5035 process to complete was treated as a delivery failure, a timeout while
5036 writing the message to the pipe was logged, but erroneously treated as a
5037 successful delivery. Such timeouts include transport filter timeouts. For
5038 consistency with the overall process timeout, these timeouts are now
5039 treated as errors, giving rise to delivery failures by default. However,
5040 there is now a new Boolean option for the pipe transport called
5041 timeout_defer, which, if set TRUE, converts the failures into defers for
5042 both kinds of timeout. A transport filter timeout is now identified in
5045 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5046 systems where "make" and "gmake" are different, calling "gmake" at top
5047 level broke things. I've arranged for the value of $(MAKE) to be passed
5048 from the Makefile to this script so that it can call the same version of
5052 A note about Exim versions 4.44 and 4.50
5053 ----------------------------------------
5055 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5056 changes of various kinds. As a consequence, a big documentation update was
5057 needed. This delayed the release for rather longer than seemed good, especially
5058 in the light of a couple of (minor) security issues. Therefore, the changes
5059 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5060 release. So 4.44 and 4.50 are in effect two different branches that both start
5063 I have left the 4.50 change log unchanged; it contains all the changes since
5064 4.43. The change log for 4.44 is below; many of its items are identical to
5065 those for 4.50. This seems to be the most sensible way to preserve the
5066 historical information.
5072 1. Minor wording change to the doc/README.SIEVE file.
5074 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5075 computation of the current number of files was incorrect.
5077 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5078 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5081 4. Give more explanation in the error message when the command for a transport
5082 filter fails to execute.
5084 5. There are several places where Exim runs a non-Exim command in a
5085 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5086 was being done only for the command run by the queryprogram router. It is
5087 now done for all such subprocesses. The other cases are: ${run, transport
5088 filters, and the commands run by the lmtp and pipe transports.
5090 6. Added CONFIGURE_GROUP build-time option.
5092 7. Some older OS have a limit of 256 on the maximum number of file
5093 descriptors. Exim was using setrlimit() to set 1000 as a large value
5094 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5095 systems. I've change it so that if it can't get 1000, it tries for 256.
5097 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5098 was an oversight, and furthermore, ever since the addition of extra
5099 controls (e.g. 4.43/32), the checks on when to allow different forms of
5100 "control" were broken. There should now be diagnostics for all cases when a
5101 control that does not make sense is encountered.
5103 9. Added the /retain_sender option to "control=submission".
5105 10. $recipients is now available in the predata ACL (oversight).
5107 11. Tidy the search cache before the fork to do a delivery from a message
5108 received from the command line. Otherwise the child will trigger a lookup
5109 failure and thereby defer the delivery if it tries to use (for example) a
5110 cached ldap connection that the parent has called unbind on.
5112 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5113 of $address_data from the recipient verification was clobbered by the
5114 sender verification.
5116 13. The value of address_data from a sender verification is now available in
5117 $sender_address_data in subsequent conditions in the ACL statement.
5119 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5121 15. Added a new option "connect=<time>" to callout options, to set a different
5124 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5125 was its contents. (It was OK if the option was not defined at all.)
5127 17. A "Completed" log line is now written for messages that are removed from
5128 the spool by the -Mrm option.
5130 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5131 information about exactly what failed.
5133 19. Added -dd to debug only the daemon process.
5135 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5136 handles timeouts, both on the server side and network timeouts. Renamed the
5137 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5139 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5140 It is now set to "smtps".
5142 22. $host_address is now set to the target address during the checking of
5143 ignore_target_hosts.
5145 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5146 being passed; this would have caused $sender_host_name to have been used if
5147 matching the list had actually called for a host name (not very likely,
5148 since this list is usually IP addresses). A host name is now passed as
5151 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5152 code that specifies a non-restarting handler (typically sigaction() in
5153 modern systems) in an attempt to fix a rare and obscure crash bug.
5155 25. Narrowed the window for a race in the daemon that could cause it to ignore
5156 SIGCHLD signals. This is not a major problem, because they are used only to
5157 wake it up if nothing else does.
5159 26. A malformed maildirsize file could cause Exim to calculate negative values
5160 for the mailbox size or file count. Odd effects could occur as a result.
5161 The maildirsize information is now recalculated if the size or filecount
5164 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5165 support for a long time. Removed HAVE_SYS_VFS_H.
5167 28. Installed the latest version of exipick from John Jetmore.
5169 29. In an address list, if the pattern was not a regular expression, an empty
5170 subject address (from a bounce message) matched only if the pattern was an
5171 empty string. Non-empty patterns were not even tested. This was the wrong
5172 because it is perfectly reasonable to use an empty address as part of a
5173 database query. An empty address is now tested by patterns that are
5174 lookups. However, all the other forms of pattern expect the subject to
5175 contain a local part and a domain, and therefore, for them, an empty
5176 address still always fails if the pattern is not itself empty.
5178 30. Exim went into a mad DNS loop when attempting to do a callout where the
5179 host was specified on an smtp transport, and looking it up yielded more
5180 than one IP address.
5182 31. Re-factored the code for checking spool and log partition space into a
5183 function that finds that data and another that does the check. The former
5184 is then used to implement four new variables: $spool_space, $log_space,
5185 $spool_inodes, and $log_inodes.
5187 32. The RFC2047 encoding function was originally intended for short strings
5188 such as real names; it was not keeping to the 75-character limit for
5189 encoded words that the RFC imposes. It now respects the limit, and
5190 generates multiple encoded words if necessary. To be on the safe side, I
5191 have increased the buffer size for the ${rfc2047: expansion operator from
5194 33. It is now permitted to omit both strings after an "if" condition; if the
5195 condition is true, the result is "true". As before, when the second string
5196 is omitted, a false condition yields an empty string. This makes it less
5197 cumbersome to write custom ACL and router conditions.
5199 34. Failure to deliver a bounce message always caused it to be frozen, even if
5200 there was an errors_to setting on the router. The errors_to setting is now
5203 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5204 canonical form (fully expanded) before being placed in
5205 $sender_host_address.
5207 36. The table in the code that translates DNS record types into text (T_A to
5208 "A" for instance) was missing entries for NS and CNAME. It is just possible
5209 that this could have caused confusion if both these types were looked up
5210 for the same domain, because the text type is used as part of Exim's
5211 per-process caching. But the chance of anyone hitting this buglet seems
5214 37. The dnsdb lookup has been extended in a number of ways.
5216 (1) There is a new type, "zns", which walks up the domain tree until it
5217 finds some nameserver records. It should be used with care.
5219 (2) There is a new type, "mxh", which is like "mx" except that it returns
5220 just the host names, not the priorities.
5222 (3) It is now possible to give a list of domains (or IP addresses) to be
5223 looked up. The behaviour when one of the lookups defers can be
5224 controlled by a keyword.
5226 (4) It is now possible to specify the separator character for use when
5227 multiple records are returned.
5229 38. The dnslists ACL condition has been extended: it is now possible to supply
5230 a list of IP addresses and/or domains to be looked up in a particular DNS
5233 39. Added log_selector=+queue_time_overall.
5235 40. When running the queue in the test harness, wait just a tad after forking a
5236 delivery process, to get repeatability of debugging output.
5238 41. Include certificate and key file names in error message when GnuTLS fails
5239 to set them up, because the GnuTLS error message doesn't include the name
5240 of the failing file when there is a problem reading it.
5242 42. Allow both -bf and -bF in the same test run.
5244 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5246 44. The "Exiscan patch" is now merged into the mainline Exim source.
5248 45. Sometimes the final signoff response after QUIT could fail to get
5249 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5250 < 0 before doing a fflush(). This bug looks as though it goes back to the
5251 introduction of TLS in release 3.20, but "sometimes" must have been rare
5252 because the tests only now provoked it.
5254 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5255 (this can affect the format of dates).
5257 47. exim_tidydb, when checking for the continued existence of a message for
5258 which it has found a message-specific retry record, was not finding
5259 messages that were in split spool directories. Consequently, it was
5260 deleting retry records that should have stayed in existence.
5262 48. Steve fixed some bugs in eximstats.
5264 49. The SPA authentication driver was not abandoning authentication and moving
5265 on to the next authenticator when an expansion was forced to fail,
5266 contradicting the general specification for all authenticators. Instead it
5267 was generating a temporary error. It now behaves as specified.
5269 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5270 (the order specifies the preference for clients). The order is now AES256,
5271 AES128, 3DES, ARCFOUR128.
5273 51. Small patch to Sieve code - explicitly set From: when generating an
5276 52. Exim crashed if a remote delivery caused a very long error message to be
5277 recorded - for instance if somebody sent an entire SpamAssassin report back
5278 as a large number of 550 error lines. This bug was coincidentally fixed by
5279 increasing the size of one of Exim's internal buffers (big_buffer) that
5280 happened as part of the Exiscan merge. However, to be on the safe side, I
5281 have made the code more robust (and fixed the comments that describe what
5284 53. Now that there can be additional text after "Completed" in log lines (if
5285 the queue_time_overall log selector is set), a one-byte patch to exigrep
5286 was needed to allow it to recognize "Completed" as not the last thing in
5289 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5290 patch that reportedly fixes this has been added. I am not expert enough to
5291 create a test for it. This is what the patch creator wrote:
5293 "I found a little strange behaviour of ldap code when working with
5294 Windows 2003 AD Domain, where users was placed in more than one
5295 Organization Units. When I tried to give exim partial DN, the exit code
5296 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5297 But simultaneously result of request was absolutely normal ldap result,
5298 so I produce this patch..."
5300 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5301 REFERENCE, so I have modified the code to exclude the patch when that macro
5304 55. Some experimental protocols are using DNS PTR records for new purposes. The
5305 keys for these records are domain names, not reversed IP addresses. The
5306 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5307 leaves it alone. Component reversal etc. now happens only for IP addresses.
5310 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5312 57. Double the size of the debug message buffer (to 2048) so that more of very
5313 long debug lines gets shown.
5315 58. The exicyclog utility now does better if the number of log files to keep
5316 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5318 59. Two changes related to the smtp_active_hostname option:
5320 (1) $smtp_active_hostname is now available as a variable.
5321 (2) The default for smtp_banner uses $smtp_active_hostname instead
5322 of $primary_hostname.
5324 60. The host_aton() function is supposed to be passed a string that is known
5325 to be a valid IP address. However, in the case of IPv6 addresses, it was
5326 not checking this. This is a hostage to fortune. Exim now panics and dies
5327 if the condition is not met. A case was found where this could be provoked
5328 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5329 components; fortuitously, this particular loophole had already been fixed
5330 by change 4.50/55 above.
5332 If there are any other similar loopholes, the new check in host_aton()
5333 itself should stop them being exploited. The report I received stated that
5334 data on the command line could provoke the exploit when Exim was running as
5335 exim, but did not say which command line option was involved. All I could
5336 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5337 running as the user.
5340 61. There was a buffer overflow vulnerability in the SPA authentication code
5341 (which came originally from the Samba project). I have added a test to the
5342 spa_base64_to_bits() function which I hope fixes it.
5345 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5346 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5348 63. The daemon start-up calls getloadavg() while still root for those OS that
5349 need the first call to be done as root, but it missed one case: when
5350 deliver_queue_load_max is set with deliver_drop_privilege. This is
5351 necessary for the benefit of the queue runner, because there is no re-exec
5352 when deliver_drop_privilege is set.
5354 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5355 This has been fixed.
5357 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5358 was in use, was not putting the data itself into the right store pool;
5359 consequently, it could be overwritten for a subsequent message in the same
5360 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5363 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5365 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5366 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5367 the function were treating the return as a boolean value, which happened to
5368 work because 0=false and not-0=true, but is not correct code.
5370 68. The host_aton() function was not handling scoped IPv6 addresses (those
5371 with, for example, "%eth0" on the end) correctly.
5373 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5374 negated items (that is, ~something) in unsigned ints. Some compilers
5375 apparently mutter when there is no cast.
5377 70. If an address verification called from an ACL failed, and did not produce a
5378 user-specific message (i.e. there was only a "system" message), nothing was
5379 put in $acl_verify_message. In this situation, it now puts the system
5382 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5383 session; change 4.31/43 added the unwanted input to the log line - except
5384 that it did not do this in the start of session case. It now does.
5386 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5387 This is wrong and can cause the other end to generate a synchronization
5388 error if it is another Exim or anything else that does the synchronization
5389 check. A QUIT command is no longer sent after a timeout.
5391 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5392 during host lookups.
5394 74. The defer_ok option of callout verification was not working if it was used
5395 when verifying addresses in header lines, that is, for this case:
5397 verify = header_sender/callout=defer_ok
5399 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5400 those file descriptors could be used for SMTP connections. If anything
5401 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5402 could be sent to the SMTP client, causing chaos. The daemon now opens
5403 stdin, stdout, and stderr to /dev/null when it puts itself into the
5406 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5407 log by default. The user can override this with suitable Perl magic.
5409 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5410 the log message when discard triggers, was not working for the DATA ACL or
5411 for the non-SMTP ACL.
5413 78. Error message wording change in sieve.c.
5415 79. If smtp_accept_max_per_host was set, the number of connections could be
5416 restricted to fewer than expected, because the daemon was trying to set up
5417 a new connection before checking whether the processes handling previous
5418 connections had finished. The check for completed processes is now done
5419 earlier. On busy systems, this bug wouldn't be noticed because something
5420 else would have woken the daemon, and it would have reaped the completed
5423 80. If a message was submitted locally by a user whose login name contained one
5424 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5425 It caused a spool format error. I have fixed the spool reading code. A
5426 related problem was that the "from" clause in the Received: line became
5427 illegal because of the space(s). It is now covered by ${quote_local_part.
5429 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5432 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5434 83. Patch from Sieve maintainer to fix unterminated string problem in
5435 "vacation" handling.
5437 84. Some minor changes to the Linux configuration files to help with other
5438 OS variants using glibc.
5440 85. One more patch for Sieve to update vacation handling to latest spec.
5443 ----------------------------------------------------
5444 See the note above about the 4.44 and 4.50 releases.
5445 ----------------------------------------------------
5451 1. Change 4.43/35 introduced a bug that caused file counts to be
5452 incorrectly computed when quota_filecount was set in an appendfile
5455 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5456 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5459 3. Give more explanation in the error message when the command for a transport
5460 filter fails to execute.
5462 4. There are several places where Exim runs a non-Exim command in a
5463 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5464 was being done only for the command run by the queryprogram router. It is
5465 now done for all such subprocesses. The other cases are: ${run, transport
5466 filters, and the commands run by the lmtp and pipe transports.
5468 5. Some older OS have a limit of 256 on the maximum number of file
5469 descriptors. Exim was using setrlimit() to set 1000 as a large value
5470 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5471 systems. I've change it so that if it can't get 1000, it tries for 256.
5473 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5474 was an oversight, and furthermore, ever since the addition of extra
5475 controls (e.g. 4.43/32), the checks on when to allow different forms of
5476 "control" were broken. There should now be diagnostics for all cases when a
5477 control that does not make sense is encountered.
5479 7. $recipients is now available in the predata ACL (oversight).
5481 8. Tidy the search cache before the fork to do a delivery from a message
5482 received from the command line. Otherwise the child will trigger a lookup
5483 failure and thereby defer the delivery if it tries to use (for example) a
5484 cached ldap connection that the parent has called unbind on.
5486 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5487 of $address_data from the recipient verification was clobbered by the
5488 sender verification.
5490 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5491 was its contents. (It was OK if the option was not defined at all.)
5493 11. A "Completed" log line is now written for messages that are removed from
5494 the spool by the -Mrm option.
5496 12. $host_address is now set to the target address during the checking of
5497 ignore_target_hosts.
5499 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5500 being passed; this would have caused $sender_host_name to have been used if
5501 matching the list had actually called for a host name (not very likely,
5502 since this list is usually IP addresses). A host name is now passed as
5505 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5506 code that specifies a non-restarting handler (typically sigaction() in
5507 modern systems) in an attempt to fix a rare and obscure crash bug.
5509 15. Narrowed the window for a race in the daemon that could cause it to ignore
5510 SIGCHLD signals. This is not a major problem, because they are used only to
5511 wake it up if nothing else does.
5513 16. A malformed maildirsize file could cause Exim to calculate negative values
5514 for the mailbox size or file count. Odd effects could occur as a result.
5515 The maildirsize information is now recalculated if the size or filecount
5518 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5519 support for a long time. Removed HAVE_SYS_VFS_H.
5521 18. Updated exipick to current release from John Jetmore.
5523 19. Allow an empty sender to be matched against a lookup in an address list.
5524 Previously the only cases considered were a regular expression, or an
5527 20. Exim went into a mad DNS lookup loop when doing a callout where the
5528 host was specified on the transport, if the DNS lookup yielded more than
5531 21. The RFC2047 encoding function was originally intended for short strings
5532 such as real names; it was not keeping to the 75-character limit for
5533 encoded words that the RFC imposes. It now respects the limit, and
5534 generates multiple encoded words if necessary. To be on the safe side, I
5535 have increased the buffer size for the ${rfc2047: expansion operator from
5538 22. Failure to deliver a bounce message always caused it to be frozen, even if
5539 there was an errors_to setting on the router. The errors_to setting is now
5542 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5543 canonical form (fully expanded) before being placed in
5544 $sender_host_address.
5546 24. Updated eximstats to version 1.33
5548 25. Include certificate and key file names in error message when GnuTLS fails
5549 to set them up, because the GnuTLS error message doesn't include the name
5550 of the failing file when there is a problem reading it.
5552 26. Expand error message when OpenSSL has problems setting up cert/key files.
5555 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5556 (this can affect the format of dates).
5558 28. exim_tidydb, when checking for the continued existence of a message for
5559 which it has found a message-specific retry record, was not finding
5560 messages that were in split spool directories. Consequently, it was
5561 deleting retry records that should have stayed in existence.
5563 29. eximstats updated to version 1.35
5564 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5565 1.35 - bugfix such that pie charts by volume are generated correctly
5567 30. The SPA authentication driver was not abandoning authentication and moving
5568 on to the next authenticator when an expansion was forced to fail,
5569 contradicting the general specification for all authenticators. Instead it
5570 was generating a temporary error. It now behaves as specified.
5572 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5573 (the order specifies the preference for clients). The order is now AES256,
5574 AES128, 3DES, ARCFOUR128.
5576 31. Small patch to Sieve code - explicitly set From: when generating an
5579 32. Exim crashed if a remote delivery caused a very long error message to be
5580 recorded - for instance if somebody sent an entire SpamAssassin report back
5581 as a large number of 550 error lines. This bug was coincidentally fixed by
5582 increasing the size of one of Exim's internal buffers (big_buffer) that
5583 happened as part of the Exiscan merge. However, to be on the safe side, I
5584 have made the code more robust (and fixed the comments that describe what
5587 33. Some experimental protocols are using DNS PTR records for new purposes. The
5588 keys for these records are domain names, not reversed IP addresses. The
5589 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5590 leaves it alone. Component reversal etc. now happens only for IP addresses.
5593 34. The host_aton() function is supposed to be passed a string that is known
5594 to be a valid IP address. However, in the case of IPv6 addresses, it was
5595 not checking this. This is a hostage to fortune. Exim now panics and dies
5596 if the condition is not met. A case was found where this could be provoked
5597 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5598 components; fortuitously, this particular loophole had already been fixed
5599 by change 4.50/55 or 4.44/33 above.
5601 If there are any other similar loopholes, the new check in host_aton()
5602 itself should stop them being exploited. The report I received stated that
5603 data on the command line could provoke the exploit when Exim was running as
5604 exim, but did not say which command line option was involved. All I could
5605 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5606 running as the user.
5609 35. There was a buffer overflow vulnerability in the SPA authentication code
5610 (which came originally from the Samba project). I have added a test to the
5611 spa_base64_to_bits() function which I hope fixes it.
5614 36. The daemon start-up calls getloadavg() while still root for those OS that
5615 need the first call to be done as root, but it missed one case: when
5616 deliver_queue_load_max is set with deliver_drop_privilege. This is
5617 necessary for the benefit of the queue runner, because there is no re-exec
5618 when deliver_drop_privilege is set.
5620 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5621 was in use, was not putting the data itself into the right store pool;
5622 consequently, it could be overwritten for a subsequent message in the same
5623 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5626 38. Sometimes the final signoff response after QUIT could fail to get
5627 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5628 < 0 before doing a fflush(). This bug looks as though it goes back to the
5629 introduction of TLS in release 3.20, but "sometimes" must have been rare
5630 because the tests only now provoked it.
5636 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5637 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5638 Now it returns an integer. A number of calls were still expecting a T/F
5639 return. Fortuitously, in all cases, the tests worked in OK situations,
5640 which is the norm. However, things would have gone wrong on any write
5641 failures on the smtp file descriptor. This function is used when sending
5642 messages over SMTP and also when doing verify callouts.
5644 2. When Exim is called to do synchronous delivery of a locally submitted
5645 message (the -odf or -odi options), it no longer closes stderr before doing
5648 3. Implemented the mua_wrapper option.
5650 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5652 5. Implemented the functions header_remove(), header_testname(),
5653 header_add_at_position(), and receive_remove_recipient(), and exported them
5656 6. If an ACL "warn" statement specified the addition of headers, Exim already
5657 inserted X-ACL-Warn: at the start if there was no header name. However, it
5658 was not making this test for the second and subsequent header lines if
5659 there were newlines in the string. This meant that an invalid header could
5660 be inserted if Exim was badly configured.
5662 7. Allow an ACL "warn" statement to add header lines at the start or after all
5663 the Received: headers, as well as at the end.
5665 8. Added the rcpt_4xx retry error code.
5667 9. Added postmaster_mailfrom=xxx to callout verification option.
5669 10. Added mailfrom=xxxx to the callout verification option, for verify=
5672 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5673 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5675 12. Inserted some casts to stop certain compilers warning when using pointer
5676 differences as field lengths or precisions in printf-type calls (mostly
5677 affecting debugging statements).
5679 13. Added optional readline() support for -be (dynamically loaded).
5681 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5682 same clock tick as a message's arrival, so that its received time was the
5683 same as the "first fail" time on the retry record, and that message
5684 remained on the queue past the ultimate address timeout, every queue runner
5685 would try a delivery (because it was past the ultimate address timeout) but
5686 after another failure, the ultimate address timeout, which should have then
5687 bounced the address, did not kick in. This was a "< instead of <=" error;
5688 in most cases the first failure would have been in the next clock tick
5689 after the received time, and all would be well.
5691 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5692 being recognized when the domain list was tested by the match_domain
5693 condition in an expansion string.
5695 16. Added the ${str2b64: operator.
5697 17. Exim was always calling setrlimit() to set a large limit for the number of
5698 processes, without checking whether the existing limit was already
5699 adequate. (It did check for the limit on file descriptors.) Furthermore,
5700 errors from getrlimit() and setrlimit() were being ignored. Now they are
5701 logged to the main and panic logs, but Exim does carry on, to try to do its
5702 job under whatever limits there are.
5704 18. Imported PCRE 5.0.
5706 19. Trivial typo in log message " temporarily refused connection" (the leading
5709 20. If the log selector return_path_on_delivery was set and an address was
5710 redirected to /dev/null, the delivery process crashed because it assumed
5711 that a return path would always be set for a "successful" delivery. In this
5712 case, the whole delivery is bypassed as an optimization, and therefore no
5715 21. Internal re-arrangement: the function for sending a challenge and reading
5716 a response while authentication was assuming a zero-terminated challenge
5717 string. It's now changed to take a pointer and a length, to allow for
5718 binary data in such strings.
5720 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5722 23. Exim was not respecting finduser_retries when seeking the login of the
5723 uid under which it was called; it was always trying 10 times. (The default
5724 setting of finduser_retries is zero.) Also, it was sleeping after the final
5725 failure, which is pointless.
5727 24. Implemented tls_on_connect_ports.
5729 25. Implemented acl_smtp_predata.
5731 26. If the domain in control=submission is set empty, Exim assumes that the
5732 authenticated id is a complete email address when it generates From: or
5733 Sender: header lines.
5735 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5736 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5737 chown and chgrp in /bin and hostname in /usr/bin.
5739 28. Exim was keeping the "process log" file open after each use, just as it
5740 does for the main log. This opens the possibility of it remaining open for
5741 long periods when the USR1 signal hits a daemon. Occasional processlog
5742 errors were reported, that could have been caused by this. Anyway, it seems
5743 much more sensible not to leave this file open at all, so that is what now
5746 29. The long-running daemon process does not normally write to the log once it
5747 has entered its main loop, and it closes the log before doing so. This is
5748 so that log files can straightforwardly be renamed and moved. However,
5749 there are a couple of unusual error situations where the daemon does write
5750 log entries, and I had neglected to close the log afterwards.
5752 30. The text of an SMTP error response that was received during a remote
5753 delivery was being truncated at 512 bytes. This is too short for some of
5754 the long messages that one sometimes sees. I've increased the limit to
5757 31. It is now possible to make retry rules that apply only when a message has a
5758 specific sender, in particular, an empty sender.
5760 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5761 it possible to be selective about when SMTP synchronization is enforced.
5763 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5765 32. Implemented hosts_connection_nolog.
5767 33. Added an ACL for QUIT.
5769 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5772 35. Added mailbox_size and mailbox_filecount to appendfile.
5774 36. Added control = no_multiline_responses to ACLs.
5776 37. There was a bug in the logic of the code that waits for the clock to tick
5777 in the case where the clock went backwards by a substantial amount such
5778 that the microsecond fraction of "now" was more than the microsecond
5779 fraction of "then" (but the whole seconds number was less).
5781 38. Added support for the libradius Radius client library this is found on
5782 FreeBSD (previously only the radiusclient library was supported).
5788 1. When certain lookups returned multiple values in the form name=value, the
5789 quoting of the values was not always being done properly. Specifically:
5790 (a) If the value started with a double quote, but contained no whitespace,
5792 (b) If the value contained whitespace other than a space character (i.e.
5793 tabs or newlines or carriage returns) it was not quoted.
5794 This fix has been applied to the mysql and pgsql lookups by writing a
5795 separate quoting function and calling it from the lookup code. The fix
5796 should probably also be applied to nisplus, ibase and oracle lookups, but
5797 since I cannot test any of those, I have not disturbed their existing code.
5799 2. A hit in the callout cache for a specific address caused a log line with no
5800 reason for rejecting RCPT. Now it says "Previous (cached) callout
5801 verification failure".
5803 3. There was an off-by-one bug in the queryprogram router. An over-long
5804 return line was truncated at 256 instead of 255 characters, thereby
5805 overflowing its buffer with the terminating zero. As well as fixing this, I
5806 have increased the buffer size to 1024 (and made a note to document this).
5808 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5809 when Exim is waiting for an SMTP response from a remote server, Exim
5810 restarts its select() call on the socket, thereby resetting its timeout.
5811 This is not a problem when such interrupts are rare. Somebody set up a cron
5812 job to run exiwhat every 2 minutes, which is less than the normal select()
5813 timeout (5 or 10 minutes). This meant that the select() timeout never
5814 kicked in because it was always reset. I have fixed this by comparing the
5815 time when an interrupt arrives with the time at the start of the first call
5816 to select(). If more time than the timeout has elapsed, the interrupt is
5817 treated as a timeout.
5819 5. Some internal re-factoring in preparation for the addition of Sieve
5820 extensions (by MH). In particular, the "personal" test is moved to a
5821 separate function, and given an option for scanning Cc: and Bcc: (which is
5822 not set for Exim filters).
5824 6. When Exim created an email address using the login of the caller as the
5825 local part (e.g. when creating a From: or Sender: header line), it was not
5826 quoting the local part when it contained special characters such as @.
5828 7. Installed new OpenBSD configuration files.
5830 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5831 try to make them clearer.
5833 9. Callout options, other than the timeout value, were being ignored when
5834 verifying sender addresses in header lines. For example, when using
5836 verify = header_sender/callout=no_cache
5838 the cache was (incorrectly) being used.
5840 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5841 only the Cygwin environment.
5843 11. When return_path_on_delivery was set as a log selector, if different remote
5844 addresses in the same message used different return paths and parallel
5845 remote delivery occurred, the wrong values would sometimes be logged.
5846 (Whenever a remote delivery process finished, the return path value from
5847 the most recently started remote delivery process was logged.)
5849 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5850 lines when AUTH and/or TLS are in use. This is the "received protocol"
5851 field. Exim used to use "asmtp" for authenticated SMTP, without any
5852 indication (in the protocol name) for TLS use. Now it follows the RFC and
5853 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5854 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5855 names appear in log lines as well as in Received: header lines.
5857 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5858 extensions, and comparison tests, and to fix some bugs.
5860 14. Changes to the "personal" filter test:
5862 (1) The test was buggy in that it was just doing the equivalent of
5863 "contains" tests on header lines. For example, if a user's address was
5864 anne@some.where, the "personal" test would incorrectly be true for
5866 To: susanne@some.where
5868 This test is now done by extracting each address from the header in turn,
5869 and checking the entire address. Other tests that are part of "personal"
5870 are now done using regular expressions (for example, to check local parts
5871 of addresses in From: header lines).
5873 (2) The list of non-personal local parts in From: addresses has been
5874 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5875 taken from the Sieve specification recommendations.
5877 (3) If the message contains any header line starting with "List-" it is
5878 treated as non-personal.
5880 (4) The test for "circular" in the Subject: header line has been removed
5881 because it now seems ill-conceived.
5883 15. Minor typos in src/EDITME comments corrected.
5885 16. Installed latest exipick from John Jetmore.
5887 17. If headers_add on a router specified a text string that was too long for
5888 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5889 of string_sprintf() is now avoided.
5891 18. $message_body_size was not set (it was always zero) when running the DATA
5892 ACL and the local_scan() function.
5894 19. For the "mail" command in an Exim filter, no default was being set for
5895 the once_repeat time, causing a random time value to be used if "once" was
5896 specified. (If the value happened to be <= 0, no repeat happened.) The
5897 default is now 0s, meaning "never repeat". The "vacation" command was OK
5898 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5899 (I found it when inspecting the code).
5901 20. There is now an overall timeout for performing a callout verification. It
5902 defaults to 4 times the callout timeout, which applies to individual SMTP
5903 commands during the callout. The overall timeout applies when there is more
5904 than one host that can be tried. The timeout is checked before trying the
5905 next host. This prevents very long delays if there are a large number of
5906 hosts and all are timing out (e.g. when the network connections are timing
5907 out). The value of the overall timeout can be changed by specifying an
5908 additional sub-option for "callout", called "maxwait". For example:
5910 verify = sender/callout=5s,maxwait=20s
5912 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5913 to the end before writing, but this should make it even safer).
5915 22. Exim was forgetting that it had advertised PIPELINING for the second and
5916 subsequent messages on an SMTP connection. It was also not resetting its
5917 memory on STARTTLS and an internal HELO.
5919 23. When Exim logs an SMTP synchronization error within a session, it now
5920 records whether PIPELINING has been advertised or not.
5922 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5923 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5924 rather than long int.
5926 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5928 26. Added the never_mail option to autoreply.
5934 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5935 crash if the getsockname() call failed; this can happen if a connection is
5936 closed very soon after it is established. The problem was simply in the
5937 order in which certain operations were done, causing Exim to try to write
5938 to the SMTP stream before it had set up the file descriptor. The bug has
5939 been fixed by making things happen in the correct order.
5945 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5946 before the connection was closed, thus losing the rejection response.
5948 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5949 some early Solaris releases, but causes trouble in current releases where
5950 socklen_t is defined.
5952 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5955 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5958 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5959 that releasing the top of it at the end releases what was used for sub-
5960 expansions (unless the block got too big). However, discard this block if
5961 the first thing is a variable or header, so that we can use its block when
5962 it is dynamic (useful for very large $message_headers, for example).
5964 6. Lookups now cache *every* query, not just the most recent. A new, separate
5965 store pool is used for this. It can be recovered when all lookup caches are
5966 flushed. Lookups now release memory at the end of their result strings.
5967 This has involved some general refactoring of the lookup sources.
5969 7. Some code has been added to the store_xxx() functions to reduce the amount
5970 of flapping under certain conditions.
5972 8. log_incoming_interface used to affect only the <= reception log lines. Now
5973 it causes the local interface and port to be added to several more SMTP log
5974 lines, for example "SMTP connection from", and rejection lines.
5976 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5978 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5980 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5981 could be overwritten at the end of the current message (or the start of a
5982 new message if it was set in a HELO ACL). The value is now preserved for
5983 the duration of the SMTP connection.
5985 12. If a transport had a headers_rewrite setting, and a matching header line
5986 contained an unqualified address, that address was qualified, even if it
5987 did not match any rewriting rules. The underlying bug was that the values
5988 of the flags that permit the existence of unqualified sender and recipient
5989 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5990 non-local messages, and by -bnq for local messages) were not being
5991 preserved with the message after it was received.
5993 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5994 "next input=" as part of the text comprising the host identity instead of
5995 the correct text. The code was using the same buffer for two different
5996 strings. However, depending on which order the printing function evaluated
5997 its arguments, the bug did not always show up. Under Linux, for example, my
5998 test suite worked just fine.
6000 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6001 doesn't work with some older versions of Perl. It has been changed to "my",
6002 which in any case is probably the better facility to use.
6004 15. A really picky compiler found some instances of statements for creating
6005 error messages that either had too many or two few arguments for the format
6008 16. The size of the buffer for calls to the DNS resolver has been increased
6009 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6010 for addresses that have a lot of PTR records. This alleviates a problem; it
6011 does not fully solve it.
6013 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6014 buffer now truncates the list and logs the incident, which is the same
6015 action as happens when Exim is looking up a host name and its aliases.
6016 Previously in this situation something unpredictable would happen;
6017 sometimes it was "internal error: store_reset failed".
6019 18. If a server dropped the connection unexpectedly when an Exim client was
6020 using GnuTLS and trying to read a response, the client delivery process
6021 crashed while trying to generate an error log message.
6023 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6024 string, for example:
6026 warn message = H1: something\nH2: something
6028 the text was added as a single header line from Exim's point of view
6029 though it ended up OK in the delivered message. However, searching for the
6030 second and subsequent header lines using $h_h2: did not work. This has been
6031 fixed. Similarly, if a system filter added multiple headers in this way,
6032 the routers could not see them.
6034 20. Expanded the error message when iplsearch is called with an invalid key to
6035 suggest using net-iplsearch in a host list.
6037 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6038 ACLs are no longer actually imposed (and a message to that effect is
6041 22. If a "gecos" field in a passwd entry contained escaped characters, in
6042 particular, if it contained a \" sequence, Exim got it wrong when building
6043 a From: or a Sender: header from that name. A second bug also caused
6044 incorrect handling when an unquoted " was present following a character
6045 that needed quoting.
6047 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6048 was not being matched caselessly.
6050 24. Arranged for all hyphens in the exim.8 source to be escaped with
6053 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6054 properly. Recipient callout cache records were still being keyed to include
6055 the sender, even when use_sender was set false. This led to far more
6056 callouts that were necessary. The sender is no longer included in the key
6057 when use_sender is false.
6059 26. Added "control = submission" modifier to ACLs.
6061 27. Added the ${base62d: operator to decode base 62 numbers.
6063 28. dnsdb lookups can now access SRV records.
6065 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6066 the configuration file.
6068 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6069 (-v) mode. This makes the output for a verbose queue run more intelligible.
6071 31. Added a use_postmaster feature to recipient callouts.
6073 32. Added the $body_zerocount variable, containing the number of binary zero
6074 bytes in the message body.
6076 33. The time of last modification of the "new" subdirectory is now used as the
6077 "mailbox time last read" when there is a quota error for a maildir
6080 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6082 35. Added +ignore_unknown as a special item in host lists.
6084 36. Code for decoding IPv6 addresses in host lists is now included, even if
6085 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6086 address was recognized as an IP address, but was then not correctly decoded
6087 into binary, causing unexpected and incorrect effects when compared with
6094 1. Very minor rewording of debugging text in manualroute to say "list of
6095 hosts" instead of "hostlist".
6097 2. If verify=header_syntax was set, and a header line with an unqualified
6098 address (no domain) and a large number of spaces between the end of the
6099 name and the colon was received, the reception process suffered a buffer
6100 overflow, and (when I tested it) crashed. This was caused by some obsolete
6101 code that should have been removed. The fix is to remove it!
6103 3. When running in the test harness, delay a bit after writing a bounce
6104 message to get a bit more predictability in the log output.
6106 4. Added a call to search_tidyup() just before forking a reception process. In
6107 theory, someone could use a lookup in the expansion of smtp_accept_max_
6108 per_host which, without the tidyup, could leave open a database connection.
6110 5. Added the variables $recipient_data and $sender_data which get set from a
6111 lookup success in an ACL "recipients" or "senders" condition, or a router
6112 "senders" option, similar to $domain_data and $local_part_data.
6114 6. Moved the writing of debug_print from before to after the "senders" test
6117 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6118 problems for message scanning, either using a data ACL, or using
6119 local_scan() because the Received: header was not generated till after they
6120 were called (in order to set the time as the time of reception completion).
6121 I have revised the way this works. The header is now generated after the
6122 body is received, but before the ACL or local_scan() are called. After they
6123 are run, the timestamp in the header is updated.
6129 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6130 before starting a queue runner without re-exec. This happened only when
6131 deliver_drop_privilege was set or when the Exim user was set to root. The
6132 effect of the bug was that timeouts during subsequent deliveries caused
6133 crashes instead of being properly handled. The handler is now left at its
6134 default (and expected) setting.
6136 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6137 message, again when deliver_drop_privilege is set or Exim is run as root.
6138 The bug described in (1) was not present in this case, but the tidying up
6139 of the other signals was missing. I have made the two cases consistent.
6141 3. The ignore_target_hosts setting on a manualroute router was being ignored
6142 for hosts that were looked up using the /MX notation.
6144 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6147 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6148 operated on the sender address. After changing the $sender_address to <>
6149 for the sender address verify, Exim was re-instated it as the original
6150 (before rewriting) address, but remembering that it had rewritten it, so it
6151 wasn't rewriting it again. This bug also had the effect of breaking the
6152 sender address verification caching when the sender address was rewritten.
6154 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6155 This has been changed so that if the ip literal address matches
6156 ignore_target_hosts, the router declines.
6158 7. Added expansion conditions match_domain, match_address, and match_local_
6159 part (NOT match_host).
6161 8. The placeholder for the Received: header didn't have a length field set.
6163 9. Added code to Exim itself and to exim_lock to test for a specific race
6164 condition that could lead to file corruption when using MBX delivery. The
6165 issue is with the lockfile that is created in /tmp. If this file is removed
6166 after a process has opened it but before that process has acquired a lock,
6167 there is the potential for a second process to recreate the file and also
6168 acquire a lock. This could lead to two Exim processes writing to the file
6169 at the same time. The added code performs the same test as UW imapd; it
6170 checks after acquiring the lock that its file descriptor still refers to
6171 the same named file.
6173 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6174 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6177 11. Added the smtp_active_hostname option. If used, this will typically be made
6178 to depend on the incoming interface address. Because $interface_address is
6179 not set up until the daemon has forked a reception process, error responses
6180 that can happen earlier (such as "too many connections") no longer contain
6183 12. If an expansion in a condition on a "warn" statement fails because a lookup
6184 defers, the "warn" statement is abandoned, and the next ACL statement is
6185 processed. Previously this caused the whole ACL to be aborted.
6187 13. Added the iplsearch lookup type.
6189 14. Added ident_timeout as a log selector.
6191 15. Added tls_certificate_verified as a log selector.
6193 16. Added a global option tls_require_ciphers (compare the smtp transport
6194 option of the same name). This controls incoming TLS connections.
6196 17. I finally figured out how to make tls_require_ciphers do a similar thing
6197 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6198 before starting the TLS session.
6200 18. Tabs are now shown as \t in -bP output.
6202 19. If the log selector return_path_on_delivery was set, Exim crashed when
6203 bouncing a message because it had too many Received: header lines.
6205 20. If two routers both had headers_remove settings, and the first one included
6206 a superfluous trailing colon, the final name in the first list and the
6207 first name in the second list were incorrectly joined into one item (with a
6208 colon in the middle).
6214 1. Added -C and -D options to the exinext utility, mainly to make it easier
6215 to include in the automated testing, but these could be helpful when
6216 multiple configurations are in use.
6218 2. The exinext utility was not formatting the output nicely when there was
6219 an alternate port involved in the retry record key, nor when there was a
6220 message id as well (for retries that were specific to a specific message
6221 and a specific host). It was also confused by IPv6 addresses, because of
6222 the additional colons they contain. I have fixed the IPv4 problem, and
6223 patched it up to do a reasonable job for IPv6.
6225 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6226 delivery, the log line now contains "pipelined" if PIPELINING was used.
6228 4. An SMTP transport process used to panic and die if the bind() call to set
6229 an explicit outgoing interface failed. This has been changed; it is now
6230 treated in the same way as a connect() failure.
6232 5. A reference to $sender_host_name in the part of a conditional expansion
6233 that was being skipped was still causing a DNS lookup. This no longer
6236 6. The def: expansion condition was not recognizing references to header lines
6237 that used bh_ and bheader_.
6239 7. Added the _cache feature to named lists.
6241 8. The code for checking quota_filecount in the appendfile transport was
6242 allowing one more file than it should have been.
6244 9. For compatibility with Sendmail, the command line option
6252 and sets the incoming protocol and host name (for trusted callers). The
6253 host name and its colon can be omitted when only the protocol is to be set.
6254 Note the Exim already has two private options, -pd and -ps, that refer to
6255 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6256 "s", but I don't think that's a major issue.
6258 10. A number of refactoring changes to the code, none of which should affect
6261 (a) The number of logging options was getting close to filling up the
6262 32-bit word that was used as a bit map. I have split them into two classes:
6263 those that are passed in the argument to log_write(), and those that are
6264 only ever tested independently outside of that function. These are now in
6265 separate 32-bit words, so there is plenty of room for expansion again.
6266 There is no change in the user interface or the logging behaviour.
6268 (b) When building, for example, log lines, the code previously used a
6269 macro that called string_cat() twice, in order to add two strings. This is
6270 not really sufficiently general. Furthermore, there was one instance where
6271 it was actually wrong because one of the argument was used twice, and in
6272 one call a function was used. (As it happened, calling the function twice
6273 did not affect the overall behaviour.) The macro has been replaced by a
6274 function that can join an arbitrary number of extra strings onto a growing
6277 (c) The code for expansion conditions now uses a table and a binary chop
6278 instead of a serial search (which was left over from when there were very
6279 few conditions). Also, it now recognizes conditions like "pam" even when
6280 the relevant support is not compiled in: a suitably worded error message is
6281 given if an attempt is made to use such a condition.
6283 11. Added ${time_interval:xxxxx}.
6285 12. A bug was causing one of the ddress fields not to be passed back correctly
6286 from remote delivery subprocesses. The field in question was not being
6287 subsequently used, so this caused to problems in practice.
6289 13. Added new log selectors queue_time and deliver_time.
6291 14. Might have fixed a bug in maildirsizefile handling that threw up
6292 "unexpected character" debug warnings, and recalculated the data
6293 unnecessarily. In any case, I expanded the warning message to give more
6296 15. Added the message "Restricted characters in address" to the statements in
6297 the default ACL that block characters like @ and % in local parts.
6299 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6300 Three changes have been made:
6302 (a) There was a serious bug; a negative response to MAIL caused the whole
6303 recipient domain to be cached as invalid, thereby blocking all messages
6304 to all local parts at the same domain, from all senders. This bug has
6305 been fixed. The domain is no longer cached after a negative response to
6306 MAIL if the sender used is not empty.
6308 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6311 (c) A new callout option, "use_sender" has been added for people who want
6312 the modified behaviour.
6318 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6321 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6322 indeed breaks things for older releases.
6324 3. Added additional logging to the case where there is a problem reading data
6325 from a filter that is running in a subprocess using a pipe, in order to
6326 try to track down a specific problem.
6328 4. Testing facility fudge: when running in the test harness and attempting
6329 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6330 getting "No route to host". Convert this to a timeout.
6332 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6335 6. Some OS don't have socklen_t but use size_t instead. This affects the
6336 fifth argument of getsockopt() amongst other things. This is now
6337 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6338 can be set for individual OS. I have set it for SunOS5, OSF1, and
6339 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6340 some earlier ones do not.
6342 7. Change 4.30/15 was not doing the test caselessly.
6344 8. The standard form for an IPv6 address literal was being rejected by address
6345 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6346 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6347 this, as well as the form without the "IPv6" on the front (but only when
6348 address literals are enabled, of course).
6350 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6352 10. Exim crashed if a message with an empty sender address specified by -f
6353 encountered a router with an errors_to setting. This could be provoked only
6354 by a command such as
6358 where an empty string was supplied; "<>" did not hit this bug.
6360 11. Installed PCRE release 4.5.
6362 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6363 remained set. It is now erased.
6365 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6366 times from message ids (which are base 36 rather than the normal 62).
6368 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6369 were being counted as actual protocol errors, and logged if the log
6370 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6371 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6372 and DATA following a set of rejected RCPTs do not count as protocol errors.
6373 In other words, Exim assumes they were pipelined, though this may not
6374 actually be the case. Of course, in all cases the client gets an
6375 appropriate error code.
6377 15. If a lookup fails in an ACL condition, a message about the failure may
6378 be available; it is used if testing the ACL cannot continue, because most
6379 such messages specify what the cause of the deferral is. However, some
6380 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6381 that caused an old message to be retained and used if a later statement
6382 caused a defer, replacing the real cause of the deferral.
6384 16. If an IP address had so many PTR records that the DNS lookup buffer
6385 was not large enough to hold them, Exim could crash while trying to process
6386 the truncated data. It now detects and logs this case.
6388 17. Further to 4.21/58, another change has been made: if (and only if) the
6389 first line of a message (the first header line) ends with CRLF, a bare LF
6390 in a subsequent header line has a space inserted after it, so as not to
6391 terminate the header.
6393 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6394 unnecessarily, used atoi() instead of strtol(), and didn't check the
6395 termination when getting file sizes from file names by regex.
6397 19. Completely re-implemented the support for maildirsize files, in the light
6398 of a number of problems with the previous contributed implementation
6399 (4.30/29). In particular:
6401 . If the quota is zero, the maildirsize file is maintained, but no quota is
6404 . If the maildir directory does not exist, it is created before any attempt
6405 to write a maildirsize file.
6407 . The quota value in the file is just a cache; if the quota is changed in
6408 the transport, the new value overrides.
6410 . A regular expression is available for excluding directories from the
6413 20. The autoreply transport checks the characters in options that define the
6414 message's headers; it allows continued headers, but it was checking with
6415 isspace() after an embedded newline instead of explicitly looking for a
6418 21. If all the "regular" hosts to which an address was routed had passed their
6419 expiry times, and had not reached their retry times, the address was
6420 bounced, even if fallback hosts were defined. Now Exim should go on to try
6423 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6424 equivalent code in the SMTP transport. Some hosts send humungous responses
6425 to HELO/EHLO, more than 1024 it seems.
6427 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6428 gives compiler warnings in some environments. I've now done it "properly",
6431 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6432 (because of problems with the built-in one) was declared to return uschar *
6433 instead of char *, causing compiler failure.
6435 25. Fixed a file descriptor leak when processing alias/forward files.
6437 26. Fixed a minor format string issue in dbfn.c.
6439 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6441 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6442 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6445 29. When a sender address is verified, it is cached, to save repeating the test
6446 when there is more than one recipient in a message. However, when the
6447 verification involves a callout, it is possible for different callout
6448 options to be set for different recipients. It is too complicated to keep
6449 track of this in the cache, so now Exim always runs a verification when a
6450 callout is required, relying on the callout cache for the optimization.
6451 The overhead is duplication of the address routing, but this should not be
6454 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6455 to be verified with callout=postmaster, and the main callout worked but the
6456 postmaster check failed, the verification correctly failed. However, if a
6457 subsequent RCPT command asked for sender verification *without* the
6458 postmaster check, incorrect caching caused this verification also to fail,
6461 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6462 it was not caching the DNS options (qualify_single, search_parents) that
6463 were used when the lookup failed. A subsequent lookup with different
6464 options therefore always gave the same answer, though there were cases
6465 where it should not have. (Example: a "domains = !$mx_any" option on a
6466 dnslookup router: the "domains" option is always processed without any
6467 widening, but the router might have qualify_single set.) Now Exim uses the
6468 cached value only when the same options are set.
6470 32. Added John Jetmore's "exipick" utility to the distribution.
6472 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6473 than a timeout (e.g. a certificate is required, and is not provided), an
6474 Exim server now closes the connection immediately. Previously it waited for
6475 the client to close - but if the client is SSL, it seems that they each
6476 wait for each other, leading to a delay before one of them times out.
6478 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6479 maintained 0.8.x compatibility because I don't think many are using it, and
6480 it is clearly obsolete.
6482 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6485 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6486 client certificate was expired. A simple patch fixes this, though I don't
6487 understand the full logic of why the verify callback is called multiple
6490 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6491 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6492 which causes problems with some clients (such as the Certicom SSL Plus
6493 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6494 disables the coutermeasure allowing Eudora to connect."
6496 38. Exim was not checking that a write() to a log file succeeded. This could
6497 lead to Bad Things if a log got too big, in particular if it hit a file
6498 size limit. Exim now panics and dies if it cannot write to a log file, just
6499 as it does if it cannot open a log file.
6501 39. Modified OS/Makefile-Linux so that it now contains
6503 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6505 The two -D definitions ensure that Exim is compiled with large file
6506 support, which makes it possible to handle log files that are bigger than
6509 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6510 instance) a domain was checked against a named list that involved a lookup,
6511 causing $domain_data to be set, then another domain was checked against the
6512 same list, then the first domain was re-checked, the value of $domain_data
6513 after the final check could be wrong. In particular, if the second check
6514 failed, it could be set empty. This bug probably also applied to
6517 41. The strip_trailing_dot option was not being applied to the address given
6518 with the -f command-line option.
6520 42. The code for reading a message's header from the spool was incrementing
6521 $received_count, but never initializing it. This meant that the value was
6522 incorrect (doubled) while delivering a message in the same process in which
6523 it was received. In the most common configuration of Exim, this never
6524 happens - a fresh exec is done - but it can happen when
6525 deliver_drop_privilege is set.
6527 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6528 it now includes up to 150 characters of the unexpected data in the log
6531 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6532 and building data strings. The size of both of these buffers was 10 000
6533 bytes - far larger than anybody would *ever* want, thought I. Needless to
6534 say, somebody hit the limit. I have increased the maximum line length to
6535 20 000 and the maximum data length of concatenated lines to 100 000. I have
6536 also fixed two bugs, because there was no checking on these buffers. Tsk,
6537 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6538 buffer is too small.
6540 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6541 lsearch lookups. Now it does.
6543 46. When parsing a route_list item in a manualroute router, a fixed-length
6544 buffer was used for the list of hosts. I made this 1024 bytes long,
6545 thinking that nobody would ever have a list of hosts that long. Wrong.
6546 Somebody had a whole pile of complicated expansion conditions, and the
6547 string was silently truncated, leading to an expansion error. It turns out
6548 that it is easier to change to an unlimited length (owing to other changes
6549 that have happened since this code was originally written) than to build
6550 structure for giving a limitation error. The length of the item that
6551 expands into the list of hosts is now unlimited.
6553 47. The lsearch lookup could not handle data where the length of text line was
6554 more than 4095 characters. Such lines were truncated, leading to shortened
6555 data being returned. It should now handle lines of any length.
6557 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6558 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6561 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6562 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6563 no longer happens in comment lines. A list of replacements is now placed
6564 at the head of all of the source files, except those whose only change is
6565 to replace PERL_COMMAND in the very first #! line.
6567 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6568 messages on the queue, with a bottom-up merge sort, using code contributed
6569 by Michael Haardt. This should make operations like -bp somewhat faster on
6570 large queues. It won't affect queue runners, except when queue_run_in_order
6573 51. Installed eximstats 1.31 in the distribution.
6575 52. Added support for SRV lookups to the dnslookup router.
6577 53. If an ACL referred to $message_body or $message_body_end, the value was not
6578 reset for any messages that followed in the same SMTP session.
6580 54. The store-handling optimization for building very long strings was not
6581 differentiating between the different store pools. I don't think this
6582 actually made any difference in practice, but I've tidied it.
6584 55. While running the routers to verify a sender address, $sender_address
6585 was still set to the sender address. This is wrong, because when routing to
6586 send a bounce to the sender, it would be empty. Therefore, I have changed
6587 it so that, while verifying a sender address, $sender_address is set to <>.
6588 (There is no change to what happens when verifying a recipient address.)
6590 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6591 target A or AAAA records (if not already returned) without resetting the
6592 qualify_single or search_parents options of the DNS resolver. These are
6593 inappropriate in this case because the targets of MX and SRV records must
6594 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6595 target that, when qualified, matched something in the local domain. These
6596 two options are now turned off when doing these lookups.
6598 57. It seems that at least some releases of Reiserfs (which does not have the
6599 concept of a fixed number of inodes) returns zero and not -1 for the
6600 number of available inodes. This interacted badly with check_spool_inodes,
6601 which assumed that -1 was the "no such thing" setting. What I have done is
6602 to check that the total number of inodes is greater than zero before doing
6603 the test of how many are available.
6605 58. When a "warn" ACL statement has a log_message modifier, the message is
6606 remembered, and not repeated. This is to avoid a lot of repetition when a
6607 message has many recipients that cause the same warning to be written.
6608 However, Exim was preserving the list of already written lines for an
6609 entire SMTP session, which doesn't seem right. The memory is now reset if a
6610 new message is started.
6612 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6613 debugging output unless log_rewrite was also set.
6615 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6616 of (int)(handle) when we know that handle contains (void *)(-1).
6618 61. The Exim daemon panic-logs an error return when it closes the incoming
6619 connection. However "connection reset by peer" seems to be common, and
6620 isn't really an error worthy of noting specially, so that particular error
6623 62. When Exim is trying to find all the local interfaces, it used to panic and
6624 die if the ioctl to get the interface flags failed. However, it seems that
6625 on at least one OS (Solaris 9) it is possible to have an interface that is
6626 included in the list of interfaces, but for which you get a failure error
6627 for this call. This happens when the interface is not "plumbed" into a
6628 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6629 failure of the "get flags" call assumes that the interface is down.
6631 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6632 makes life easier for people who are doing arithmetic on fields extracted
6633 from dates, where you often get leading zeros that should not be
6634 interpreted as octal.
6636 64. Added qualify_domain to the redirect router, to override the global
6639 65. If a pathologically long header line contained very many addresses (the
6640 report of this problem mentioned 10 000) and each of them was rewritten,
6641 Exim could use up a very large amount of memory. (It kept on making new
6642 copies of the header line as it rewrote, and never released the old ones.)
6643 At the expense of a bit more processing, the header rewriting function has
6644 been changed so that it no longer eats memory in this way.
6646 66. The generation of the Received: header has been moved from the time that a
6647 message starts to be received, to the time that it finishes. The timestamp
6648 in the Received: header should now be very close to that of the <= log
6649 line. There are two side-effects of this change:
6651 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6652 logged header lines no longer include the local Received: line, because
6653 it has not yet been created. The same applies to a copy of the message
6654 that is returned to a non-SMTP sender when a message is rejected.
6656 (b) When a filter file is tested using -bf, no additional Received: header
6657 is added to the test message. After some thought, I decided that this
6660 This change does not affect the value of $received_for. It is still set
6661 after address rewriting, but before local_scan() is called.
6663 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6665 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6666 gave an unhelpful panic error message, and a defer error. I have managed to
6667 change this behaviour so that it now rejects any supplied certificate,
6668 which seems right, as the list of acceptable certificates is empty.
6670 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6671 gave an unhelpful defer error. I have not managed to make this reject any
6672 supplied certificates, but the error message it gives is "no certificate
6673 supplied", which is not helpful.
6675 70. exigrep's output now also includes lines that are not associated with any
6676 message, but which match the given pattern. Implemented by a patch from
6677 Martin Sluka, which also tidied up the Perl a bit.
6679 71. Recipient callout verification, like sender verification, was using <> in
6680 the MAIL FROM command. This isn't really the right thing, since the actual
6681 sender may affect whether the remote host accepts the recipient or not. I
6682 have changed it to use the actual sender in the callout; this means that
6683 the cache record is now keyed on a recipient/sender pair, not just the
6684 recipient address. There doesn't seem to be a real danger of callout loops,
6685 since a callout by the remote host to check the sender would use <>.
6686 [SEE ABOVE: changed after hitting problems.]
6688 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6689 temporary errors. However, in the case of such a code being given after
6690 the end of a data transmission (i.e. after ".") Exim was failing to write
6691 a retry record for the message. (Yes, there was some broken host that was
6692 actually sending 8xx at this point.)
6694 73. An unknown lookup type in a host list could cause Exim to panic-die when
6695 the list was checked. (An example that provoked this was putting <; in the
6696 middle of a list instead of at the start.) If this happened during a DATA
6697 ACL check, a -D file could be left lying around. This kind of configuration
6698 error no longer causes Exim to die; instead it causes a defer error. The
6699 incident is still logged to the main and panic logs.
6701 74. Buglet left over from Exim 3 conversion. The message "too many messages
6702 in one connection" was written to the rejectlog but not the mainlog, except
6703 when address rewriting (yes!) was being logged.
6705 75. Added write_rejectlog option.
6707 76. When a system filter was run not as root (that is, when system_filter_user
6708 was set), the values of the $n variables were not being returned to the
6709 main process; thus, they were not subsequently available in the $sn
6712 77. Added +return_path_on_delivery log selector.
6714 78. A connection timeout was being treated differently from recipients deferred
6715 when testing hosts_max_try with a message that was older than the host's
6716 retry timeout. (The host should not be counted, thus allowing all hosts to
6717 be tried at least once before bouncing.) This may have been the cause of an
6718 occasionally reported bug whereby a message would remain on the queue
6719 longer than the retry timeout, but would be bounced if a delivery was
6720 forced. I say "may" because I never totally pinned down the problem;
6721 setting up timeout/retry tests is difficult. See also the next item.
6723 79. The ultimate address timeout was not being applied to errors that involved
6724 a combination of host plus message (for example, a timeout on a MAIL
6725 command). When an address resolved to a number of possible hosts, and they
6726 were not all tried for each delivery (e.g. because of hosts_max_try), a
6727 message could remain on the queue longer than the retry timeout.
6729 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6732 81. Fixed an obscure SMTP outgoing bug which required at least the following
6733 conditions: (a) there was another message waiting for the same server;
6734 (b) the server returned 5xx to all RCPT commands in the first message so
6735 that the message was not completed; (c) the server dropped the connection
6736 or gave a negative response to the RSET that Exim sends to abort the
6737 transaction. The observed case was a dropped connection after DATA that had
6738 been sent in pipelining mode. That is, the server had advertised PIPELINING
6739 but was not implementing it correctly. The effect of the bug was incorrect
6740 behaviour, such as trying another host, and this could lead to a crash.
6746 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6747 and daemon.c were passed as pointers to ints; they should have been
6748 pointers to socklen_t variables (which are typically unsigned ints).
6750 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6753 3. Fixed a really odd bug that affected only the testing scheme; patching a
6754 certain fixed string in the binary changed the value of another string that
6755 happened to be identical to the end of the original first string.
6757 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6758 name", it returns that address as the IP address. On some operating
6759 systems (e.g. Solaris), it also passes back the IP address string as the
6760 "host name". However, on others (e.g. Linux), it passes back an empty
6761 string. Exim wasn't checking for this, and was changing the host name to an
6762 empty string, assuming it had been canonicalized.
6764 5. Although rare, it is permitted to have more than one PTR record for a given
6765 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6766 all the names associated with an address, because they do in Solaris.
6767 However, it seems that they do not in Linux for data that comes from the
6768 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6769 I found this out when I moved to a new Linux workstation and tried to run
6770 the Exim test suite.
6772 To get round this problem I have changed the code so that it now does its
6773 own call to the DNS to look up PTR records when searching for a host name.
6774 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6775 addresses that are only in /etc/hosts are still found.
6777 This behaviour is, however, controlled by an option called host_lookup_
6778 order, which defaults to "bydns:byaddr". If people want to use the other
6779 order, or indeed, just use one or the other means of lookup, they can
6780 specify it in this variable.
6782 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6783 some operating systems, this comes back from gethostbyaddr() as an empty
6784 string, and this is what Exim used to test for. However, it seems that in
6785 other systems, "." is yielded. Exim now tests for this case too.
6787 7. The values of check_spool_space and check_log_space are now held internally
6788 as a number of kilobytes instead of an absolute number of bytes. If a
6789 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6790 kilobyte. This means that much larger values can be stored.
6792 8. Exim monitor: an attempt to get the action menu when not actually pointing
6793 at a message produces an empty menu entitled "No message selected". This
6794 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6795 no entries in it ("Shell widget menu has zero width and/or height"). So I
6796 have added a single, blank menu entry in this case.
6798 9. Added ${quote_local_part.
6800 10. MIME decoding is now applied to the contents of Subject: header lines when
6803 11. Now that a reference to $sender_host_address automatically causes a reverse
6804 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6805 host lookup before query-style lookups in lists that might use this
6806 variable. This has therefore been abolished, and the "net-" prefix is no
6807 longer necessary for query-style lookups.
6809 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6810 have been a typo for LFLAGS, so it has been changed.
6812 13. The install script calls Exim with "-C /dev/null" in order to find the
6813 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6814 to be output. However, since Exim outputs its version number before the
6815 error, it didn't break the script. It just looked ugly. I fixed this by
6816 always allowing "-C /dev/null" if the caller is root.
6818 14. Ignore overlarge ACL variable number when reading spool file - insurance
6819 against a later release with more variables having written the file.
6821 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6822 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6823 this, as well as the form without the "IPv6" on the front.
6825 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6826 OS/Makefile-Darwin file.
6828 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6829 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6831 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6832 variable. However, some OS use ints for inodes. I've added cast to long int
6833 to get rid of the compiler warning.
6835 19. I had forgotten to lock out "/../" in configuration file names when
6836 ALT_CONFIG_PREFIX was set.
6838 20. Routers used for verification do not need to specify transports. However,
6839 if such a router generated a host list, and callout was configured, Exim
6840 crashed, because it could not find a port number from the (non-existent)
6841 transport. It now assumes port 25 in this circumstance.
6843 21. Added the -t option to exigrep.
6845 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6846 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6847 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6848 them is set, LOOKUP_LSEARCH is forced.
6850 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6852 24. Added sender and host information to the "rejected by local_scan()" log
6853 line; previously there was no indication of these.
6855 25. Added .include_if_exists.
6857 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6858 Linux. It turns out that not all file systems support this. Apparently some
6859 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6860 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6861 not supported on the file descriptor, is now ignored when Exim is trying to
6862 sync a directory. This applies only to Linux.
6864 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6866 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6867 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6870 29. Added support for maildirsize files from supplied patch (modified a bit).
6872 30. The use of :fail: followed by an empty string could lead Exim to respond to
6873 sender verification failures with (e.g.):
6875 550 Verification failed for <xxx>
6876 550 Sender verify failed
6878 where the first response line was missing the '-' that indicates it is not
6879 the final line of the response.
6881 31. The loop for finding the name of the user that called Exim had a hardwired
6882 limit of 10; it now uses the value of finduser_retries, which is used for
6883 all other user lookups.
6885 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6888 33. Exim was neglecting to zero errno before one call of strtol() when
6889 expanding a string and expecting an integer value. On some systems this
6890 resulted in spurious "integer overflow" errors. Also, it was casting the
6891 result into an int without checking.
6893 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6894 did not work. The code looks as if it has *never* worked, though it appears
6895 to have been documented since at least release 1.62. I have made it work.
6897 35. The "timeout_DNS" error in retry rules, also documented since at least
6898 1.62, also never worked. As it isn't clear exactly what this means, and
6899 clearly it isn't a major issue, I have abolished the feature by treating it
6900 as "timeout", and writing a warning to the main and panic logs.
6902 36. The display of retry rules for -brt wasn't always showing the error code
6905 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6906 timeout_connect_A, timeout_connect_MX.
6908 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6909 to the empty sender.
6911 39. The daemon was not analysing the content of -oX till after it had closed
6912 stderr and disconnected from the controlling terminal. This meant that any
6913 syntax errors were only noted on the panic log, and the return code from
6914 the command was 0. By re-arranging the code a little, I've made the
6915 decoding happen first, so such errors now appear on stderr, and the return
6916 code is 1. However, the actual setting up of the sockets still happens in
6917 the disconnected process, so errors there are still only recorded on the
6920 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6921 connections (as happens on some IP stacks) was logged at start up time as
6922 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6923 it from "IPv6 and IPv4", which means that two separate sockets are being
6926 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6927 says whether AF_INET or AF_INET6 was passed as an argument.
6929 42. Exiwhat output was messed up when time zones were included in log
6932 43. Exiwhat now gives more information about the daemon's listening ports,
6933 and whether -tls-on-connect was used.
6935 44. The "port" option of the smtp transport is now expanded.
6937 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6938 silently ignored. Now an error message is written to the main and panic
6941 46. There's a new ACL modifier called "logwrite" which writes to a log file
6942 as soon as it is encountered.
6944 47. Added $local_user_uid and $local_user_gid at routing time.
6946 48. Exim crashed when trying to verify a sender address that was being
6949 49. Exim was recognizing only a space character after ".include". It now also
6950 recognizes a tab character.
6952 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6953 extracting the relevant information from the specification. The man page no
6954 longer contains scrambled data for the -d option, and I've added a section
6955 at the front about calling Exim under different names.
6957 51. Added "extra_headers" argument to the "mail" command in filter files.
6959 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6962 53. Installed eximstats 1.29.
6964 54. Added transport_filter_timeout as a generic transport option.
6966 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6967 Cc: header lines. This was required by RFC 822, but it not required by RFC
6970 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6971 incoming messages that did not have them. Now it does so only if the
6972 message originates locally, that is, if there is no associated remote host
6973 address. When Resent- header lines are present, this applies to the Resent-
6974 lines rather than the non-Resent- lines.
6976 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6977 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6979 58. Messages for configuration errors now include the name of the main
6980 configuration file - useful now that there may be more than one file in a
6981 list (.included file names were always shown).
6983 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6984 for those rare installations that do not start the daemon as root or run it
6985 setuid root. I've cut out the call to initgroups() if the daemon is not
6988 60. The Exim user and group can now be bound into the binary as text strings
6989 that are looked up at the start of Exim's processing.
6991 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6993 62. Added $mailstore_basename variable.
6995 63. Installed patch to sieve.c from Michael Haardt.
6997 64. When Exim failed to open the panic log after failing to open the main log,
6998 the original message it was trying to log was written to stderr and debug
6999 output, but if they were not available (the usual case in production), it
7000 was lost. Now it is written to syslog before the two lines that record the
7001 failures to open the logs.
7003 65. Users' Exim filters run in subprocesses under the user's uid. It is
7004 possible for a "deliver" command or an alias in a "personal" command to
7005 provoke an address rewrite. If logging of address rewriting is configured,
7006 this fails because the process is not running as root or exim. There may be
7007 a better way of dealing with this, but for the moment (because 4.30 needs
7008 to be released), I have disabled address rewrite logging when running a
7009 filter in a non-root, non-exim process.
7015 1. The buildconfig auxiliary program wasn't quoting the value set for
7016 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7017 not defined. This bug was masked in 4.22 by the effect that was fixed in
7020 2. Some messages that were rejected after a message id was allocated were
7021 shown as "incomplete" by exigrep. It no longer does this for messages that
7022 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7024 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7025 have allow_domain_literals set, the ID did not get logged in the <= line.
7026 Domain literals are now always recognized in Message-ID: header lines.
7028 4. The first argument for a ${extract expansion item is the key name or field
7029 number. Leading and trailing spaces in this item were not being ignored,
7030 causing some misleading effects.
7032 5. When deliver_drop_privilege was set, single queue runner processes started
7033 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7034 same command in the process it spins off) were not dropping privilege.
7036 6. When the daemon running as "exim" started a queue runner, it always
7037 re-executed Exim in the spun-off process. This is a waste of effort when
7038 deliver_drop_privilege is set. The new process now just calls the
7039 queue-runner function directly.
7045 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7048 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7049 ignored. Though the use of -oP was forcing the writing of a pid file, it
7050 was always written to the default place.
7052 3. If the message "no IP address found for host xxxx" is generated during
7053 incoming verification, it is now followed by identification of the incoming
7054 connection (so you can more easily find what provoked it).
7056 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7058 5. Added some features to "harden" Exim a bit more against certain attacks:
7060 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7061 be put in Local/Makefile. This is like the never_users runtime option,
7062 but it cannot be overridden. The default setting is "root".
7064 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7065 prefix string with which any file named in a -C command line option
7068 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7069 is retained for -C and -D only if the caller of Exim is root. Without
7070 it, the exim user may also use -C and -D and retain privilege.
7072 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7073 command line option is disabled.
7075 6. Macro names set by the -D option must start with an upper case letter, just
7076 like macro names defined in the configuration file.
7078 7. Added "dereference=" facility to LDAP.
7080 8. Two instances of the typo "uknown" in the source files are fixed.
7082 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7083 the Configure-Makefile script screwed up while processing it.
7085 10. Incorporated PCRE 4.4.
7087 11. The SMTP synchronization check was not operating right at the start of an
7088 SMTP session. For example, it could not catch a HELO sent before the client
7089 waited for the greeting. There is now a check for outstanding input at the
7090 point when the greeting is written. Because of the duplex, asynchronous
7091 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7092 way, but not yet received, when the check is performed.
7094 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7095 on TCP/IP sockets, because this apparently causes some broken clients to
7098 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7099 unchanged) from the Cygwin maintainer.
7101 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7102 was supported instead of testing for mbx. Effectively a typo.
7104 15. The spa authenticator server code was not checking that the input it
7105 received was valid base64.
7107 16. The debug output line for the "set" modifier in ACLs was not showing the
7108 name of the variable that was being set.
7110 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7112 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7113 lookup on its own. Something else was needed to trigger the lookup. For
7114 example, a match in host_lookup or the need for a host name in a host list.
7115 Now, if $sender_host_name is referenced and the host name has not yet been
7116 looked up, a lookup is performed. If the lookup fails, the variable remains
7117 empty, and $host_lookup_failed is set to "1".
7119 19. Added "eqi" as a case-independent comparison operator.
7121 20. The saslauthd authentication condition could segfault if neither service
7122 nor realm was specified.
7124 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7125 error message that was logged was misleading, and incoming SMTP
7126 connections were dropped. The message is now more accurate, and temporary
7127 errors are given to SMTP connections.
7129 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7130 (except RSET) with a 421 error, until QUIT is received. However, it was
7131 failing to send a response to QUIT.
7133 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7134 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7135 ACL is now run after the helo_try_verify_hosts code.
7137 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7138 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7139 case-independent, so other case variants are also recognized). Apparently
7140 some systems use these upper case variants.
7142 25. If more than two messages were waiting for the same host, and a transport
7143 filter was specified for the transport, Exim sent two messages over the
7144 same TCP/IP connection, and then failed with "socket operation on non-
7145 socket" when it tried to send the third.
7147 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7149 27. The extern definition of crypt16() in expand.c was not being excluded when
7150 the OS had its own crypt16() function.
7152 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7153 as a preferred synonym for return_size_limit, both as an option and as an
7156 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7158 30. Changed the default configuration ACL to relax the local part checking rule
7159 for addresses that are not in any local domains. For these addresses,
7160 slashes and pipe symbols are allowed within local parts, but the sequence
7161 /../ is explicitly forbidden.
7163 31. SPA server authentication was not clearing the challenge buffer before
7166 32. log_message in a "warn" ACL statement was writing to the reject log as
7167 well as to the main log, which contradicts the documentation and doesn't
7168 seem right (because no rejection is happening). So I have stopped it.
7170 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7171 However, I am unable to do any testing of this.
7173 34. Fixed an infelicity in the appendfile transport. When checking directories
7174 for a mailbox, to see if any needed to be created, it was accidentally
7175 using path names with one or more superfluous leading slashes; tracing
7176 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7178 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7179 contents are added to the log line that is written for every discarded
7180 recipient. (Previously a log_message setting was ignored.)
7182 36. The ${quote: operator now quotes the string if it is empty.
7184 37. The install script runs exim in order to find its version number. If for
7185 some reason other than non-existence or emptiness, which it checks, it
7186 could not run './exim', it was installing it with an empty version number,
7187 i.e. as "exim-". This error state is now caught, and the installation is
7190 38. An argument was missing from the function that creates an error message
7191 when Exim fails to connect to the socket for saslauthd authentication.
7192 This could cause Exim to crash, or give a corrupted message.
7194 39. Added isip, isip4, and isip6 to ${if conditions.
7196 40. The ACL variables $acl_xx are now saved with the message, and can be
7197 accessed later in routers, transports, and filters.
7199 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7200 strings in the file are not string-expanded.
7202 42. If a MAIL command specified a SIZE value that was too large to fit into an
7203 int variable, the check against message_size_limit failed. Such values are
7204 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7205 day this will have to be increased, but I don't think I want to be around
7206 when emails are that large.
7213 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7214 iconv() is not standard in FreeBSD.
7216 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7217 IPv6 enabled. The observed symptom was a segmentation fault on return from
7218 the function os_common_find_running_interfaces() in src/os.c.
7220 3. In the check_special_case() function in daemon.c I had used "errno" as an
7221 argument name, which causes warnings on some systems. This was basically a
7222 typo, since it was named "eno" in the comments!
7224 4. The code that waits for the clock to tick (at a resolution of some fraction
7225 of a second) so as to ensure message-id uniqueness was always waiting for
7226 at least one whole tick, when it could have waited for less. [This is
7227 almost certainly not relevant at current processor speeds, where it is
7228 unlikely to ever wait at all. But we try to future-proof.]
7230 5. The function that sleeps for a time interval that includes fractions of a
7231 second contained a race. It did not block SIGALRM between setting the
7232 timer, and suspending (a couple of lines later). If the interval was short
7233 and the sigsuspend() was delayed until after it had expired, the suspension
7234 never ended. On busy systems this could lead to processes getting stuck for
7237 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7238 process, before it forks any delivery processes. The open lookup caching
7239 mechanism meant that the open file or database connection was passed into
7240 the delivery process. The problem was that delivery processes always tidy
7241 up cached lookup data. This could cause a problem for the next delivery
7242 process started by the queue runner, because the external queue runner
7243 process does not know about the closure. So the next delivery process
7244 still has data in the lookup cache. In the case of a file lookup, there was
7245 no problem because closing a file descriptor in a subprocess doesn't affect
7246 the parent. However, if the lookup was caching a connection to a database,
7247 the connection was closed, and the second delivery process was likely to
7248 see errors such as "PGSQL: query failed: server closed the connection
7249 unexpectedly". The problem has been fixed by closing all cached lookups
7250 in a queue runner before running a delivery process.
7252 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7253 seem to have the "const" qualifier which it has on other OS. I've
7256 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7257 *of the same type* (client or server) with the same public name that an
7258 error should be diagnosed.
7260 9. When Exim looked up a host name for an IP address, but failed to find the
7261 original IP address when looking up the host name (a safety check), it
7262 output the message "<ip address> does not match any IP for NULL", which was
7263 confusing, to say the least. The bug was that the host name should have
7264 appeared instead of "NULL".
7266 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7267 user that is built into the binary, and the -C or -D options is used, root
7268 privilege is dropped before the configuration file is read. In addition,
7269 logging is switched to stderr instead of the normal log files. If the
7270 configuration then re-defines the Exim user, the unprivileged environment
7271 is probably not what is expected, so Exim logs a panic warning message (but
7274 However, if deliver_drop_privilege is set, the unprivileged state may well
7275 be exactly what is intended, so the warning has been cut out in that case,
7276 and Exim is allowed to try to write to its normal log files.
7282 1. smtp_return_error_details was not giving details for temporary sender
7283 or receiver verification errors.
7285 2. Diagnose a configuration error if two authenticators have the same public
7288 3. Exim used not to create the message log file for a message until the first
7289 delivery attempt. This could be confusing when incoming messages were held
7290 for policy or load reasons. The message log file is now created at the time
7291 the message is received, and an initial "Received" line is written to it.
7293 4. The automatically generated man page for command line options had a minor
7294 bug that caused no ill effects; however, a more serious problem was that
7295 the procedure for building the man page automatically didn't always
7296 operate. Consequently, release 4.20 contains an out-of-date version. This
7297 shouldn't happen again.
7299 5. When building Exim with embedded Perl support, the script that builds the
7300 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7301 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7303 6. The freeze_tell option was not being used for messages that were frozen on
7304 arrival, either by an ACL or by local_scan().
7306 7. Added the smtp_incomplete_transaction log selector.
7308 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7309 was accepting AUTH without a new EHLO.
7311 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7312 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7315 10. Logging of TCP/IP connections (when configured) now happens in the main
7316 daemon process instead of the child process, so that the TCP/IP connection
7317 count is more accurate (but it can never be perfect).
7319 11. The use of "drop" in a nested ACL was not being handled correctly in the
7320 outer ACL. Now, if condition failure induced by the nested "drop" causes
7321 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7322 or "require"), the connection is dropped.
7324 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7325 that yield "discard" can now be used with an "accept" or a "discard" verb,
7326 but an error is generated for any others (because I can't see a useful way
7327 to define what should happen).
7329 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7330 are now processed in the same way as lines in the Exim configuration file.
7331 In particular, continuation lines are supported.
7333 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7335 15. Added -ti meaning -t -i.
7337 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7338 domains, and warn by logging if others are found.
7340 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7341 returned from GIFCONF when Exim is trying to find the list of interfaces on
7342 a host. The code in os.c has been modified to copy each ifreq to an aligned
7343 structure in all cases.
7345 Also, in some cases, the returned ifreq's were being copied to a 'struct
7346 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7347 means the last couple of bytes of an IPv6 address could be chopped if the
7348 ifreq contained only a normal sockaddr (14 bytes storage).
7350 18. Named domain lists were not supported in the hosts_treat_as_local option.
7351 An entry such as +xxxx was not recognized, and was treated as a literal
7354 19. Ensure that header lines added by a DATA ACL are included in the reject log
7355 if the ACL subsequently rejects the message.
7357 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7358 MD5 (which is deprecated).
7360 21. When testing a filter file using -bf, Exim was writing a message when it
7361 took the sender from a "From " line in the message, but it was not doing so
7362 when it took $return_path from a Return-Path: header line. It now does.
7364 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7365 with a valid header line field name (a series of printing characters
7366 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7368 23. Changed "disc" in the source to "disk" to conform to the documentation and
7369 the book and for uniformity.
7371 24. Ignore Sendmail's -Ooption=value command line item.
7373 25. When execve() failed while trying to run a command in a pipe transport,
7374 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7375 could be confused with a return value of 69 from the command itself. This
7376 has been changed to 127, the value the shell returns if it is asked to run
7377 a non-existent command. The wording for the related log line suggests a
7378 non-existent command as the problem.
7380 26. If received_header_text expands to an empty string, do not add a Received:
7381 header line to the message. (Well, it adds a token one on the spool, but
7382 marks it "old" so that it doesn't get used or transmitted.)
7384 27. Installed eximstats 1.28 (addition of -nt option).
7386 28. There was no check for failure on the call to getsockname() in the daemon
7387 code. This can fail if there is a shortage of resources on the system, with
7388 ENOMEM, for example. A temporary error is now given on failure.
7390 29. Contrary to the C standard, it seems that in some environments, the
7391 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7392 program. Exim now does this explicitly; it affects the formatting of
7393 timestamps using strftime().
7395 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7396 complaints. I've now initialized all the variables, to avoid this.
7398 32. Header lines added by a system filter were not being "seen" during
7399 transport-time rewrites.
7401 33. The info_callback() function passed to OpenSSL is set up with type void
7402 (*)(SSL *, int, int), as described somewhere. However, when calling the
7403 function (actually a macro) that sets it up, the type void(*)() is
7404 expected. I've put in a cast to prevent warnings from picky compilers.
7406 34. If a DNS black list lookup found a CNAME record, but there were no A
7407 records associated with the domain it pointed at, Exim crashed.
7409 35. If a DNS black list lookup returned more than one A record, Exim ignored
7410 all but the first. It now scans all returned addresses if a particular IP
7411 value is being sought. In this situation, the contents of the
7412 $dnslist_value variable are a list of all the addresses, separated by a
7415 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7416 to accept a host name and all its aliases if the forward lookup for any of
7417 them yielded the IP address of the incoming connection. Now it accepts only
7418 those names whose forward lookup yields the correct IP address. Any other
7419 names are discarded. This closes a loophole whereby a rogue DNS
7420 administrator could create reverse DNS records to break through a
7421 wildcarded host restriction in an ACL.
7423 37. If a user filter or a system filter that ran in a subprocess used any of
7424 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7425 the wrong values were passed to the pipe command ($thisaddress had the
7426 value of $0, $0 had the value of $1, etc). This bug was introduced by
7427 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7429 38. Improved the line breaking for long SMTP error messages from ACLs.
7430 Previously, if there was no break point between 40 and 75 characters, Exim
7431 left the rest of the message alone. Two changes have been made: (a) I've
7432 reduced the minimum length to 35 characters; (b) if it can't find a break
7433 point between 35 and 75 characters, it looks ahead and uses the first one
7434 that it finds. This may give the occasional overlong line, but at least the
7435 remaining text gets split now.
7437 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7438 file descriptors might be low, and that setting 1000 would always raise it.
7439 It turns out that in some environments, the limit is already over 1000 and
7440 that lowering it causes trouble. So now Exim takes care not to decrease it.
7442 40. When delivering a message, the value of $return_path is set to $sender_
7443 address at the start of routing (routers may change the value). By an
7444 oversight, this default was not being set up when an address was tested by
7445 -bt or -bv, which affected the outcome if any router or filter referred to
7448 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7449 log only when "message" or "log_message" are set. However, if one of the
7450 conditions was an address verification, or a call to a nested ACL, the
7451 messages generated by the underlying test were being passed through. This
7452 no longer happens. The underlying message is available in $acl_verify_
7453 message for both "message" and "log_message" expansions, so it can be
7454 passed through if needed.
7456 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7457 new expansion $bh_ to give the encoded byte string without charset
7458 translation. Translation happens only if iconv() is available; HAVE_ICONV
7459 indicates this at build time. HEADERS_CHARSET gives the charset to
7460 translate to; headers_charset can change it in the configuration, and
7461 "headers charset" can change it in an individual filter file.
7463 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7464 that creates RFC 2047 encoded "words" labels them as that charset instead
7465 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7466 expansion operator; (ii) when Exim creates a From: line for a local
7467 message; (iii) when a header line is rewritten to include a "phrase" part.
7469 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7470 buggy, causing it to skip the first lines of messages whose message ID
7471 ended in 'D'. This would not have bitten before Exim release 4.14, because
7472 message IDs were unlikely to end in 'D' before then. The effect was to have
7473 incorrect size information for certain domains.
7475 45. #include "config.h" was missing at the start of the crypt16.c module. This
7476 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7479 46. If there was a timeout during a "random" callout check, Exim treated it as
7480 a failure of the random address, and carried on sending RSET and the real
7481 address. If the delay was just some slowness somewhere, the response to the
7482 original RCPT would be taken as a response to RSET and so on, causing
7483 mayhem of various kinds.
7485 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7486 when I implemented it. It didn't allow for the fact that some option values
7487 may legitimately be negative (e.g. size_addition), and it didn't even do
7488 the right test for positive values.
7490 48. Domain names in DNS records are case-independent. Exim always looks them up
7491 in lower case. Some resolvers return domain names in exactly the case they
7492 appear in the zone file, that is, they may contain uppercase letters. Not
7493 all resolvers do this - some return always lower case. Exim was treating a
7494 change of case by a resolver as a change of domain, similar to a widening
7495 of a domain abbreviation. This triggered its re-routing code and so it was
7496 trying to route what was effectively the same domain again. This normally
7497 caused routing to fail (because the router wouldn't handle the domain
7498 twice). Now Exim checks for this case specially, and just changes the
7499 casing of the domain that it ultimately uses when it transmits the message
7502 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7505 50. If a filter generated a file delivery with a non-absolute name (possible if
7506 no home directory exists for the router), the forbid_file option was not
7509 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7510 the existing equality matching.
7512 52. Exim was using ints instead of ino_t variables in some places where it was
7513 dealing with inode numbers.
7515 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7516 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7517 called TMPDIR, and if it finds it is different, it changes its value.
7519 54. The smtp_printf() function is now made available to local_scan() so
7520 additional output lines can be written before returning. There is also an
7521 smtp_fflush() function to enable the detection of a dropped connection.
7522 The variables smtp_input and smtp_batched_input are exported to
7525 55. Changed the default runtime configuration: the message "Unknown user"
7526 has been removed from the ACL, and instead placed on the localuser router,
7527 using the cannot_route_message feature. This means that any verification
7528 failures that generate their own messages won't get overridden. Similarly,
7529 the "Unrouteable address" message that was in the ACL for unverifiable
7530 relay addresses has also been removed.
7532 56. Added hosts_avoid_esmtp to the smtp transport.
7534 57. The exicyclog script was not checking for the esoteric option
7535 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7536 will work only if exicyclog is run under the appropriate euid.
7538 58. Following a discussion on the list, the rules by which Exim recognises line
7539 endings on incoming messages have been changed. The -dropcr and drop_cr
7540 options are now no-ops, retained only for backwards compatibility. The
7541 following line terminators are recognized: LF CRLF CR. However, special
7542 processing applies to CR:
7544 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7545 nor a local message in the state where . is a terminator.
7547 (ii) If a bare CR is encountered in a header line, an extra space is added
7548 after the line terminator so as not to end the header. The reasoning
7549 behind this is that bare CRs in header lines are most likely either
7550 to be mistakes, or people trying to play silly games.
7552 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7553 was being incorrectly given as 18 bytes larger than it should have been.
7554 This is a VOB (very old bug).
7556 60. This may never have affected anything current, but just in case it has:
7557 When the local host is found other than at the start of a list of hosts,
7558 the local host, those with the same MX, and any that follow, are discarded.
7559 When the list in question was part of a longer list of hosts, the following
7560 hosts (not currently being processed) were also being discarded. This no
7561 longer happens. I'm not sure if this situation could ever has previously
7564 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7567 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7568 header. This is something that is recommended in a new Internet Draft, and
7569 is something that is documented as being done by Sendmail. There are two
7570 possible values. For messages generated by the autoreply transport, Exim
7573 Auto-Submitted: auto-replied
7575 whereas for all other generated messages (e.g. bounces) it adds
7577 Auto-Submitted: auto-generated
7579 63. The "personal" condition in filters now includes a test for the
7580 Auto-Submitted: header. If it contains the string "auto-" the message it
7581 not considered personal.
7583 64. Added rcpt_include_affixes as a generic transport option.
7585 65. Added queue_only_override (default true).
7587 66. Added the syslog_duplication option.
7589 67. If what should have been the first header line of a message consisted of
7590 a space followed by a colon, Exim was mis-interpreting it as a header line.
7591 It isn't of course - it is syntactically invalid and should therefore be
7592 treated as the start of the message body. The misbehaviour could have
7593 caused a number of strange effects, including loss of data in subsequent
7594 header lines, and spool format errors.
7596 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7597 client host had authenticated. This control can now be exercised by an ACL
7598 for more flexibility.
7600 69. By default, callouts do not happen when testing with -bh. There is now a
7601 variant, -bhc, which does actually run the callout code, including
7602 consulting and updating the callout cache.
7604 70. Added support for saslauthd authentication, courtesy of Alexander
7607 71. If statvfs() failed on the spool or log directories while checking their
7608 size for availability, Exim confusingly gave the error "space shortage".
7609 Furthermore, in debugging mode it crashed with a floating point exception.
7610 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7611 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7612 serious problem, Exim now writes to the main and panic logs when this
7613 happens, with details of the failure. It then refuses to accept the
7614 incoming message, giving the message "spool directory problem" or "log
7615 directory problem" with a 421 code for SMTP messages.
7617 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7618 0, 1, and 2 exist, because some OS complain for execs without them (see
7619 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7620 descriptors. However, the code omitted to check that the open succeeded,
7621 causing mysterious errors if for some reason the permissions on /dev/null
7622 got screwed. Now Exim writes a message to the main and panic logs, and
7623 bombs out if it can't open /dev/null.
7625 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7626 interact so that it is all more flexible. It is supposed to remain
7627 backwards compatible. Also added extra_local_interfaces.
7629 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7630 to bomb out with an assertion failure - to the client this appears as a
7631 connection drop. This problem occurs in the part of the code that was taken
7632 from the Samba project. Fortunately, the assertion is in a very simple
7633 function, so I have fixed this by reproducing the function inline in the
7634 one place where it is called, and arranging for authentication to fail
7635 instead of killing the process with assert().
7637 75. The SPA client code was not working when the server requested OEM rather
7638 than Unicode encoding.
7640 76. Added code to make require_files with a specific uid setting more usable in
7641 the case where statting the file as root fails - usually a non-root-mounted
7642 NFS file system. When this happens and the failure is EACCES, Exim now
7643 forks a subprocess and does the per-uid checking as the relevant uid.
7645 77. Added process_log_path.
7647 78. If log_file_path was not explicitly set, a setting of check_log_space or
7648 check_log_inodes was ignored.
7650 79. If a space check for the spool or log partitions fails, the incident is now
7651 logged. Of course, in the latter case the data may get lost...
7653 80. Added the %p formatting code to string_format() so that it can be used to
7654 print addresses in debug_print(). Adjusted all the address printing in the
7655 debugging in store.c to use %p rather than %d.
7657 81. There was a concern that a line of code in smtp_in.c could overflow a
7658 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7659 initially expressed, the concern was not well-founded, because trailing
7660 spaces are removed early. However, if the trailing spaces were followed by
7661 a NULL, they did not get removed, so the overflow was possible. Two fixes
7664 (a) I re-wrote the offending code in a cleaner fashion.
7665 (b) If an incoming SMTP command contains a NULL character, it is rejected
7668 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7669 runs initgroups(), so that if the Exim user is in any additional groups,
7670 they will be used during message reception.
7676 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob